Open Collections

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Nursing practice, continuity of care, and maternal-infant health outcomes in a remote First Nations community Tarlier, Denise Susan 2006

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Notice for Google Chrome users:
If you are having trouble viewing or searching the PDF with Google Chrome, please download it here instead.

Item Metadata


831-ubc_2006-199138.pdf [ 15.43MB ]
JSON: 831-1.0092763.json
JSON-LD: 831-1.0092763-ld.json
RDF/XML (Pretty): 831-1.0092763-rdf.xml
RDF/JSON: 831-1.0092763-rdf.json
Turtle: 831-1.0092763-turtle.txt
N-Triples: 831-1.0092763-rdf-ntriples.txt
Original Record: 831-1.0092763-source.json
Full Text

Full Text

N U R S I N G P R A C T I C E , C O N T I N U I T Y O F C A R E , A N D M A T E R N A L - I N F A N T H E A L T H O U T C O M E S EST A R E M O T E FIRST N A T I O N S C O M M U N I T Y by D E N I S E S U S A N T A R L I E R General Nursing Diploma, British Columbia Institute of Technology, 1982 B . S .N., University of British Columbia, 1992 Diploma in Outpost and Community Health Nursing, Dalhousie University, 1996 M . S . N . , University of British Columbia, 2001 A THESIS S U B M I T T E D I N P A R T I A L F U L F I L L M E N T O F T H E R E Q U I R E M E N T S F O R T H E D E G R E E O F D O C T O R O F P H I L O S O P H Y in T H E F A C U L T Y O F G R A D U A T E S T U D I E S (Nursing) T H E U N I V E R S I T Y O F B R I T I S H C O L U M B I A Apr i l 2006 © Denise Susan Tarlier, 2006 ABSTRACT The purpose of this ethnographic study was to explore nurses' primary care practice, continuity of patient care, and clinical health outcomes in one remote First Nations community in Canada. This work was underpinned by previous conceptual work in (a) primary health care, (b) The Nursing Role Effectiveness Model (i.e., a structure-process-outcomes model), and (c) continuity of care. Continuity was conceptualized in terms of relational, informational and management continuity. Cl inical health outcomes were explored through quantitative data abstracted from patient chart review on two selected health indicator conditions: (a) prenatal care and (b) infant care through the first year. Focusing on maternal-infant outcomes offered a window through which to explore nurses' practice and examine continuity at the individual level and importantly, to extrapolate findings related to continuity and fragmentation beyond the individual level to implications at the community and population levels. Indicator condition criteria, originally developed for the Burlington Randomized Trial o f nurse practitioner practice, were revised and updated to reflect current clinical practice. Additional data from patient chart reviews were collected to provide context for the quantitative outcomes data. Ethnographic data collection strategies included interviews with nurses, other health care professionals and paraprofessional staff, field observations (including observations of nurse-patient encounters), and document and artefact review. Four over-arching themes emerged from analysis and interpretation of the data: (a) nurses' work, influenced by a broader context of inequity and marginalization, occurred at the margins of mainstream nursing practice, (b) health outcomes and continuity of care were characterized by fragmentation, (c) nurse-patient encounters were suggestive of relational disengagement and (d) dissonance existed between perceptions of health care and demonstrated health outcomes. Nurses' practice was subjected to and shaped by the same ii contextual issues that were associated with inequity and marginalization in this study. Unti l this context, specifically the social determinants that exert such profound influence on the health of aboriginal Canadians, begins to shift, shifts in nurses' practice are unlikely to be either effective or sustainable. i i i TABLE OF CONTENTS A B S T R A C T i i T A B L E OF C O N T E N T S iv LIST OF T A B L E S vi i A C K N O W L E D G M E N T S vi i i C H A P T E R O N E - I N T R O D U C T I O N 1 Background and Rationale for the Research 1 Research Purpose and Objectives 6 Organization of the Dissertation 8 C H A P T E R T W O - C O N C E P T U A L F O U N D A T I O N S 10 Introduction 10 Primary Health Care 10 The Nursing Role Effectiveness Model 13 Continuity o f Care 16 Summary : 24 C H A P T E R T H R E E - A C R I T I C A L S Y N O P S I S OF T H E R E L E V A N T L I T E R A T U R E 26 Introduction 26 Context 27 The Implications of Remoteness 27 The Historical and Socio-Political Context of Aboriginal Health and Health Services 34 Outpost Nursing L iterature 41 Organizational and Structural Issues 43 Isolation 43 Understanding the Context of Aboriginal Health 47 Nursing Turnover 48 Worklife Issues 49 Nursing Practice Issues 54 Literature Relevant to the Evaluation of Primary Care 61 The Burlington Randomized Trial: Appraising Quality of Primary Care 62 Appraising Nurse Practitioner Practice 66 Appraising Outpost Nursing Practice 69 Summary 74 C H A P T E R F O U R - R E S E A R C H D E S I G N A N D I M P L E M E N T A T I O N 76 Introduction 76 Method of Inquiry: Ethnography 77 The Study Site 80 Gaining Entry 81 Units of Analysis 82 Overview of Data Collection Strategies 82 The Data Set 84 Patient Chart Review 84 Chart Data Abstraction Instruments 84 iv Reliability and Val idi ty of the Chart Data Review Instruments 87 Chart Review Data 88 Interview Data 90 Observational Data 92 Demographic Data 93 Artefacts 93 Personal Journal and Notebook 95 Overview of Participants 95 Protecting the Anonymity of Participants 98 Data Analysis 98 Validity 101 Ethical Considerations 103 Summary 104 C H A P T E R F I V E - T H E C O N T E X T A N D S T R U C T U R E OF N U R S E S ' W O R K : W O R K I N G A T T H E M A R G I N S 105 Introduction to the Research Findings 105 Stepping into the Community: First Impressions 105 It's a Different World Here 108 The Local Context of Health Services: Structure and Infrastructure 113 It's a Demanding Job: Contextual and Structural Issues That Influenced Nurses' Practice 118 Being Too Busy: High Nursing Turnover and Working Short Staffed 119 Being Prepared: Nurses' Readiness for the Nursing Station Role 125 The Interface of Being Too Busy and Being Unprepared: Inefficiency and Ineffectiveness 134 Case Scenario 5.1 136 Summary 138 C H A P T E R SIX - D A T A F R O M C H A R T R E V I E W : A W I N D O W I N T O H E A L T H O U T C O M E S A N D C O N T I N U I T Y O F C A R E 140 Introduction 140 Exploring Health Outcomes in the Study Site 141 Description of the Sample Population 141 Perinatal Health Outcomes 143 Infant Health Outcomes 148 Linking Health Outcomes to Context and Continuity o f Care 150 Case Scenario 6.1 153 Case Scenario 6.2 157 Case Scenario 6.3 160 Summary 163 C H A P T E R S E V E N - E X P L O R I N G T H E P R O C E S S OF E N G A G I N G IN N U R S E -P A T I E N T E N C O U N T E R S 165 Introduction 165 The Process of Othering: Issues Influencing Distance in Nurse-Patient Relationships 170 Normalization of Risk: Perpetuating Complacency 180 v Case Scenario 7.1 185 Role Disorientation: Whose Job is it Anyway? 186 Summary 193 C H A P T E R E I G H T - D I S S O N A N C E : F A L L I N G T H R O U G H T H E G A P S 195 Introduction 195 There're N o Problems Here 196 Case Scenario 8.1 199 The Fragmentation of Primary Health Care 203 Summary 208 Thematic Summarization of the Research Findings 208 C H A P T E R N I N E - D I S C U S S I O N , R E C O M M E N D A T I O N S A N D R E F L E C T I O N S . . . . 212 Introduction 212 Reflections on Study Methods 213 Clinical Implications of Maternal-Infant Health Outcomes 216 Extremes o f Age and Parity 217 Birth Weight and Gestational Age at Birth 219 Complications of Pregnancy, Labour and Delivery 222 Postpartum Visits 224 Anemia in Infants 225 Immunization 226 Infant Weight Gain in the First 12 Months 228 The Microcosm of Care: Nurse-Patient Encounters and Continuity of Care 229 Vulnerable Groups and Normalization of Risk 230 Barriers to Nurse-Patient Engagement 232 Assisted Evolution: Shifting Nurses' Practice 238 Seeing the Bigger Picture 239 Preparing Nurses for Practice 244 Relief Nurses: Turning Liabili ty into Asset 248 Putting "Health" Back into Primary Health Care 252 Recommendations 259 Significance o f This Study and Future Research 263 A Closing Note 265 R E F E R E N C E S 267 Appendix A : Nursing Role Effectiveness Model 287 Appendix B : Adapted Conceptual Model 288 Appendix C: Research Design 289 Appendix D : Data Collection Protocol Questions 290 Appendix E : Data Abstraction Instruments 293 Appendix F: Interview Guiding Questions 301 Appendix G : Framework for Analyzing Continuity of Care 302 Appendix H : U B C Behavioural Research Ethics Board Certificates of Approval 303 Appendix I: Consent Form 305 vi LIST OF TABLES Table 1: Nurse Demographics 97 Table 2: Maternal Age and Gravida Status 143 Table 3: Perinatal Health Clinical Outcomes 144 Table 4: Gestational Age at Birth and Birth Weight 145 Table 5: Complications o f Pregnancy, Labour and Delivery 147 Table 6: Infant Clinical Health Outcomes 149 v i i ACKNOWLEDGMENTS This research project represents several years of study, during which time I experienced an enormous amount of personal as well as professional growth. It has been a privilege to have this opportunity to develop the knowledge and skills needed to apply research and scholarship to questions that arise through clinical practice. I feel privileged also in having had the opportunity to engage in this process under the tutelage of the members of my supervisory committee: Dr. Joy Johnson, Dr. Annette Browne, and Dr. Sam Sheps. I thank each of them for their mentorship: for sharing their time and expertise so generously, for their enthusiasm, encouragement and support throughout this project, and most o f all, for trusting that I could carry out the project in the way I conceptualized it, and in which I believed it needed to be done. I thank Dr. Johnson especially for her patience and guidance in also helping me to better understand one of the more challenging aspects of doctoral studies: the "culture" of academia. I would also like to acknowledge and thank my parents, Ronald and Amelia Tarlier, for their unconditional love and enthusiastic support of (almost) all my adventures, and in particular, those that took me to remote northern communities, and eventually "back to school" (yet again). M y parents instilled in me the confidence to follow my own path. Knowing they've been there to back me up me has allowed me to take advantage of opportunities that at times, have also carried a degree of risk. Thank-you, M o m and Dad! I love you. The support of many friends and colleagues sustained me through the often-isolating process of writing this dissertation. I want to acknowledge Flo Rublee, Karin Rolfes-Kubik, and Dianne Bi rd especially, for their constant encouragement, wise advisement, and unconditional friendship. I apologize for all the times I was "running late" and you had to v i i i wait, the coffee dates I broke at the last minute, the kayaking trip I cancelled out on, and for all the whining. Thank-you for listening, and thank-you for your friendship. At the risk of exposing my eccentric side, I am compelled to acknowledge the loyal support of my most constant companion through the long days and nights of study: Thank-you, Gussie. Y o u were the one who looked in on me regularly, reminded me when it was time to eat, time to play, and time to go to bed, and who, when I refused to listen to good sense, curled up on top of my computer monitor, satisfied to purr and keep me company. Finally, I would like to acknowledge the support of First Nations and Inuit Health Branch, Manitoba Region, and the First Nations community of Snowy Lake*. I would like to thank all o f the individuals who participated in this study. This project would not have been possible without the support and participation of these key players. I hope that in return, the findings and insights derived from this study may contribute in some small way to better understanding the complex dynamics that shape both the health o f First Nations people, and the practice of the nurses who choose to work in remote aboriginal communities. * a pseudonym for the study site community ix CHAPTER ONE INTRODUCTION Background and Rationale of the Research Anecdotal evidence from both patients and practitioners, along with my own observations from clinical practice and recent health services literature (Biem, Hadjistavropoulos, Morgan, Biem, & Pong, 2003; Reid, Haggerty, & McKendry, 2002) suggests that fragmentation of patient care represents a profound and prevalent problem throughout the Canadian health care system. Patients who experience fragmented health care are also variously described as being "lost to follow-up", or having "fallen between the cracks". These patients typically receive care from multiple health care providers, often at all three levels o f care (i.e., primary, secondary and tertiary), and are often seen by providers at frequent intervals, yet they fail to achieve either optimal clinical outcomes or optimal continuity o f care. Instead, these patients experience "revolving door" health care, presenting and re-presenting for exacerbations of the same health care conditions. Fragmentation o f care may have even more potentially disastrous consequences for patients who live in rural and remote communities than it does for patients who reside in urban or suburban centres. Residents of rural and remote communities are further removed (both in time and distance) from the level o f definitive treatment that is only available in larger population centres. Patients in these communities lack the safety net implied by, for example, proximity to a tertiary level emergency department, or full medical specialist services. These patients are also more likely to be affected by health care delivery issues such as high rates o f staff turnover (both nursing and medical staff), lack of supports and resources within the community, and difficulty in recruiting and retaining highly qualified personnel (Fontaine, 2005). These issues challenge the delivery of quality health services in all settings 1 but their effects may be more pronounced in rural and remote settings due to the overall sparse nature of services (Romanow, 2002). The challenges inherent in delivering effective health services to residents of rural and remote communities are compounded further when these communities are also aboriginal communities. 1 Health status indicators confirm the poor health status of aboriginal Canadians relative to the non-aboriginal Canadian population (Canadian Institutes of Health Information [CIHI] 2004a; Indian and Northern Affairs Canada [ INAC] , 2003; Martens, Sanderson, & Jebamani, 2005). Life expectancy rates are on average almost 7 years lower in the aboriginal population, aboriginal infant mortality rates in 2000 were 22% higher than the 2001 Canadian rate ( F N I H B , 2005), and the prevalence o f chronic debilitating illness such as diabetes and cardiovascular disease is increasing (CIHI; National Forum on Health, 1997; I N A C ; Shah, Hux, & Zinman, 2000; Young, Reading, Elias, & O ' N e i l , 2000). Despite acknowledgment by major health care system stakeholders that such unacceptably poor health status is rooted in the social inequities that prevail in most remote aboriginal communities (CIHI; Romanow, 2002; Royal Commission on Aboriginal Peoples [ R C A P ] , 1996), little has changed either in the way health services are delivered in these settings, or in the social, economic, political and enculturated social conditions 2 that predominate and are Throughout this work the term aboriginal is used to refer in an inclusive sense to indigenous Canadian populations (RCAP, 1996). These are people who identify or are recognized as First Nations, Inuit or Metis. The study site community was a First Nations community where the majority of residents held "registered" First Nations status; therefore, the term First Nations is used in this work when referring specifically to the study site, and to First Nations patients and / or residents of the study site. The more inclusive term aboriginal will be used in this work when I refer more broadly to the wider context of aboriginal health or social issues, or to issues that are relevant to aboriginal people generally. No differentiation is made or implied in this work between aboriginal people with "registered Indian status" versus those without status. The use of the term aboriginal in this work is not meant to be inclusive of aboriginal or indigenous people of countries other than Canada. 21 am using the term "enculturated social conditions" in this sentence to indicate the cultures of poverty, unemployment, and abuse that are the legacy of colonialism, have developed as a result of historical social inequities, and that now serve to perpetuate the cycle of inequity within aboriginal communities (Gregory, 1992; O'Neil, 1986; RCAP, 1996). These cultures are differentiated from aboriginal and First Nations cultures per se. 2 barriers to improving health. For example, the recent evacuation in October 2005 of several hundred residents from the First Nations community of Kashechewan in northern Ontario was prompted by health problems resulting from the community's contaminated water supply. Yet concerns regarding water safety in the community had been recognized for at least two years prior to the large-scale evacuation that finally focused media and political attention on the social, economic and political conditions that had fostered this community health crisis (Bailey, 2005). Inevitably, the context in which people live and in which health care is delivered, and the structure of health services delivery exert strong influences on patient health outcomes and continuity of care. This may seem to be a rather obvious assumption; conversely, I submit that we persistently fail to examine these influences in studies of health care and health outcomes (MacMillan, MacMillan, Offord, & Dingle, 1996). Previous research suggests that the structure and context of health service delivery in remote indigenous communities influences the practice of outpost nurses3 (Vukic & Keddy, 2002). However, little information exists informing us as to how the various aspects of the "bigger picture" of health care influence the process and outcomes of care, and especially continuity of care. While such speculation is likely relevant to health care in all settings, it is particularly critical in the context of health care delivery in remote aboriginal communities. Residents of these communities are not only more vulnerable because their already-poorer health status places them at increased risk of falling through the cracks of health care delivery, but also because of disparities in accessing health services in northern and remote settings (Romanow, 2002). Thus, when these patients fall through the cracks, they are less likely to land in a safety net. 3 Nurses employed in an expanded role in Nursing Stations in remote aboriginal communities in Canada have traditionally been identified as outpost nurses. There is no standard usage of the term outpost nurse; this is not a credential but a job descriptor, thus any nurse employed in an outpost setting may be considered to be an outpost nurse. 3 The problem addressed by this research is the potential for discontinuity of health care in remote aboriginal communities. Better understanding the relationships between the context and structure of health services delivery, the process of nursing care, and health outcomes could potentially improve continuity of patient care and health outcomes. Health care in remote aboriginal communities continues to be delivered chiefly by nurses. Nurses are responsible for not only the provision of primary care but also for community health nursing, public health care and often, the non-nursing health services that are generally provided by an interdisciplinary team of health providers in less remote settings. The team often includes (but is not limited to) physicians, pharmacists, physiotherapists, dieticians, social workers and mental health workers. In remote communities where visits from other interdisciplinary team members may occur infrequently (or not at all), nurses' responsibilities often also encompass role aspects of these other health providers. Thus, as the principal and often sole providers of health services in remote communities, any exploration of health services delivery in these settings necessarily implies an exploration of the role played by nurses. There has been a long-standing recognition of the value of a primary health care (PHC) approach to address the social determinants that influence the health of aboriginal people (O 'Nei l , 1986). The expectation that outpost nurses (i.e., nurses employed in Nursing Stations in remote communities), provide care within a framework and philosophy of primary health care is similarly long-standing (Doucette, 1989). M y previous research suggested that while competency in providing primary care is fundamental to outpost nursing practice, experienced outpost nurses recognized the futility of health service delivery that fails to address the bigger picture of health determinants (Tarlier, Johnson, & Whyte, 2003). Yet nurses working in this diverse and demanding role often receive little or no additional 4 educational preparation to help them meet the expectations o f a scope o f nursing practice that is now broadly acknowledged as being both expanded and advanced, 4 and increasingly identified as a primary health care nurse practitioner role (Kul ig et al., 2003; Stewart & MacLeod, 2005; Tarlier et al.; Vuk ic & Keddy, 2002). Outpost nurses have long been recognized for the quality of healthcare they provide to northern residents (Robertson, 1973, as cited in Spitzer et al., 1974), despite a lack o f evaluative outcome studies to substantiate this claim. Outpost nurses themselves have historically viewed themselves as functioning in a nurse practitioner role (Chaytor, 1994). The model of outpost nursing that developed originally to meet the health needs of residents of remote communities has provided a template for contemporary conceptualizations of the primary health care nurse practitioner role in Canada. But paradoxically, whereas the nurse practitioner role is evolving as an advanced nursing role requiring graduate level education in many Canadian jurisdictions, including at the national level (Canadian Nurses Association [ C N A ] , 2003), most nurses employed in outpost nursing roles continue to be prepared at the diploma or baccalaureate level (Kul ig , 2005; Stewart & MacLeod, 2005). Outpost nurses require a broad base o f knowledge encompassing public health and community development as well as primary care (Health Canada, 2001c, 2001d; Tarlier et al., 2003). Yet many, i f not the majority of nurses employed in Nursing Station settings receive little i f any additional education to better prepare them to meet the expectations of an implicitly advanced and 4 In this dissertation, the use of the terms "advanced" and "expanded" nursing practice is consistent with the definitions provided in the Canadian Nurses Association document, Advanced Nursing Practice: A National Framework (2000). Advanced Nursing Practice is defined as "an umbrella term [that] describes an advanced level of nursing practice that maximizes the use of in-depth nursing knowledge and skill" (p. 4). Expanded practice is defined as "practice that is characterized by the use of competencies required to perform activities usually deemed to be outside the current scope of nursing practice" (p. 5). Thus, the practice of an individual nurse may represent: (a) advanced practice (e.g., Clinical Nurse Specialist); (b) expanded practice (e.g., a clinic nurse who performs a specific skill, such as collecting Pap smears, but who may lack the knowledge base necessary to support a broader primary care practice), or (c) advanced and expanded practice (e.g., Nurse Practitioner [CNA, 2003]). 5 explicitly expanded scope of practice (Stewart, D ' A r c y , Pitblado, Forbes, et al., 2005; Stewart & MacLeod; Vukic & Keddy, 2002). There is currently a gap in our knowledge regarding what, i f any, additional knowledge and skills are necessary to support effective continuity o f care and safe, competent primary care delivery by nurses employed in remote communities. In their study of continuity of care in three remote First Nations communities, Minore, Boone, Katt, Kinch , and Birch (2002) identified the need for research "to identify the core knowledge which all practitioners - even those going into the communities on a short-term basis - must have: the procedural, cultural and clinical essentials [emphasis added]" (p. 24). Commissioner Romanow (2002) similarly noted the need to adequately prepare health care practitioners for practice in both rural / remote and aboriginal communities. This study explored nurses' enactment of core knowledge in providing primary care services to women and children in one remote First Nations community. To identify the core knowledge needed by health practitioners in these settings demanded a better understanding of how core knowledge influenced continuity of patient care and patient health outcomes. How do we judge what knowledge is important for nurses to have without first understanding how that knowledge contributes to positive health outcomes? Moreover, nurses do not work in a vacuum, but within a community context and a structure of health services delivery (Vukic & Keddy, 2002) that also predictably influences patient outcomes directly, as well as indirectly by influencing nurses' work (Sidani, Doran, & Mitchell , 2004). Research Purpose and Objectives The purpose of this study was to address the identified gap in health services delivery knowledge by extending our understanding of how nurses, as the chief providers o f primary care (delivered within a primary health care model) influenced patient care and health 6 outcomes in one remote First Nations community. Continuity of care (conceptualized as the inverse of fragmented care) was the research phenomena of ultimate interest. Process-specific outcomes that included informational, management and relational continuity o f care (Reid et al., 2002) and explicit clinical health outcomes were evaluated through the use of two specific clinical health indicator conditions: (a) the prenatal health of women and (b) infant health. These two indicator conditions were selected because they are recognized to be essential elements of primary health care ( W H O , 1978). A s well , several of the common key indicators of population-based health are related to maternal-infant health (e.g., infant mortality, premature births, birth weights, immunization rates). Additionally, aspects related to the context, structure and processes o f the local health services system were explored and described. This information contributed to a better understanding of how these aspects influenced continuity of care both directly, and indirectly by influencing nurses' work. The main objective o f this study was to develop a big-picture understanding o f how nurses, as the principal providers o f primary care in a remote First Nations community, yet recognized as just one part o f a complex health care system, influenced continuity o f care and clinical outcomes. The following secondary research objectives further focused and guided this study: 1. To explore how specific aspects (e.g., structure and organization of the Nursing Station, staffing, nursing workload, experience and preparation of nurses to function as a primary care provider, availability of resources, communication and referral patterns within and external to the Nursing Station) o f the health service delivery system in the selected remote community influenced continuity of care; 2. To explore how these aspects supported or hindered continuity of care; 7 3. To explore how these aspects affected the nurses' ability to influence clinical outcomes and informational, relational and management continuity of care; 4. To explore the clinical, procedural and cultural 5 core knowledge nurses demonstrated in the process of providing primary care. Organization of the Dissertation The dissertation has been organized into nine chapters. Chapter One has presented a brief introduction to the research by providing the background and rationale for the study, the problem statement, purpose, and research objectives. Chapter Two presents the conceptual underpinnings of the research. The three conceptual frameworks that provided structure for the research design and helped to focus data collection and analysis were: (a) the primary health care model, (b) the Nursing Role Effectiveness Model (Irvine, Sidani, & M c G i l l i s Ha l l , 1998), and (c) the continuity of care conceptual framework (Reid et al., 2002). Chapter Three offers a critical synopsis of the relevant literature, organized in sections that pertain to the context of health and health services in remote aboriginal settings, outpost nursing and the evaluation of primary care. Chapter Four describes the research methods, data collection, analysis and ethical considerations of this ethnographic, mixed methods project. Chapters Five through Eight present an integrated interpretation o f study findings. Chapter Five describes the context o f the study site, both in terms of the community, the organization of health services, and the structure of nurses' work. Chapter Six presents the 51 am defining nurses' cultural core knowledge in an inclusive sense; to imply the knowledge and analytical skills that support an individual nurse's ability to critically reflect on her / his interactions with people who are perceived as being "different" to oneself or to the mainstream (Browne, 2005; Browne & Fiske, 2001; Papps & Ramsden, 1996). That is, cultural core knowledge implies knowledge of the "culture" that is attributable to social inequities and the historical socio-political context of aboriginal health, the unique ethnic heritage of a community of aboriginal people that a nurse may be providing care to, as well as the "culture" implied by living and working in a remote northern community. 8 data from review of patient charts. This data includes quantitative patient health outcomes data on the two selected clinical health indicator conditions, and narrative data related to continuity of patient care that were collected from patient charts as contextual notes. Chapter Seven presents findings related to the process of nursing care and continuity of care, derived primarily from observations, interviews, and artefacts. Chapter Eight presents an integrated interpretation of findings from chart review, observations, and interviews. The intersection of health outcomes, the organization and structure of health services, the process of nursing care, and continuity o f care are described. In Chapter Nine, I reflect on study methods and the significance of this study, and present a discussion of the implications of study findings for nursing practice, health and health services in remote aboriginal communities. Four recommendations relevant to the organization and structure of health services are offered, as well as suggestions for future research. 9 CHAPTER TWO CONCEPTUAL FOUNDATIONS Introduction This study is built on conceptual work in three domains of health services research: (a) primary health care research, (b) The Nursing Role Effectiveness Model , and (c) continuity of care. I w i l l start by offering a contemporary conceptual definition of primary health care that is relevant to health services research, followed by a discussion of the Nursing Role Effectiveness Model , and an overview of recent and relevant conceptual work on continuity of care. Primary Health Care Primary health care provides both a philosophical perspective and a broad conceptual framework within which to situate the proposed study. In 1978 the World Health Organization (WHO) identified P H C as the optimal approach to improving health on a global basis, and identified access, equity, essentiality, community participation and empowerment, and multisectoral collaboration as the fundamental principles of P H C . The Canadian health care system has long espoused P H C as an approach to improving the health of Canadians, to the extent that health care in Canada is now largely characterized as being a P H C system (CIHI, 2003). Thus, the Canada Health Act (1984) is based on the five principles of access, universality, portability, comprehensiveness, and public administration. Two additional principles prohibit user fees or extra billing by health providers. The principles underpinning the Canada Health Act and P H C may together be summarized as representing a participative, community driven system of health care that ensures access to equitable, essential health care for al l . 10 A significant point of difference between the Canada Health Act (1984) and a P H C approach is that the Canada Health Ac t primarily addresses primary care health needs and hospital-based acute care, but does not specifically acknowledge the influence of the social determinants of health. In contrast, P H C clearly recognizes the important role of broader socio-historical, political, economic, environmental and cultural factors in determining the health of individuals and populations. Despite early recognition of these broader dimensions in the Lalonde report (1974) and the Ottawa Charter for Health Promotion ( W H O , Health & Welfare Canada, & Canadian Public Health Association, 1986), as Romanow (2002) recognized, this disparity in focus has presented a barrier to health care reform; that is, in achieving a full transformation to a P H C system. Conceptualizing health care based on a P H C framework highlights the links between the larger social and environmental context, the structure of health services, and the processes of achieving health. P H C enables a "big picture" view o f the multiple factors - both within and external to the health care system itself - that influence health (O 'Ne i l , 1986). The move towards identifying the Canadian health care system as a P H C system, as the CIHI (2003) report demonstrates, may be viewed as a move towards acknowledging the broader determinants of health that influence the primary care health needs of the population, as well as the secondary and tertiary level needs that are associated with hospital-based care. A key point in understanding the delivery of health services within a framework of P H C lies in differentiating the role of primary care providers per se from P H C , while recognizing the interface between primary care and P H C . The W H O (1978), distinguishing primary care from P H C , identified the provision of accessible, essential primary care as one strategy within an integrated P H C framework (Tarlier et al., 2003). While primary care practitioners chiefly provide health care services directly to individuals and families, the 11 philosophical underpinnings of P H C imply the understanding that health is also influenced by the broader social context and the social determinants of health. Health status reflects more than a linear causal relationship between the provision of health services and health. Thus, a P H C orientation to health services research acknowledges the influence of the broader context of health services delivery while endeavouring to understand and evaluate the more specific relationships that exist between health and the provision of health services. B y opening up a "big-picture" view of health services delivery, a P H C orientation to health services research also opened up a large and complex scope of potential foci for this research endeavour. The possible foci engendered by a P H C orientation suggested that while acknowledging the bigger picture, the boundaries of this research project be clearly demarcated. Thus, while the focus of some P H C research is on the "big picture" itself; that is, on the interface of health and its social determinants at the level of the community (e.g., Hancock, Labonte, & Edwards, 1999), the main focus of the present study was the provision o f primary care services as one strategy within an explicitly defined P H C model of health care delivery. The key relevant features of the broader social and community context of health service delivery wi l l be discussed as these are presented in the literature. P H C provided one o f the theoretical lenses that guided data collection and analysis and the discussion of the study findings. Situating this study within a P H C framework offered a way to focus on how primary care was provided by nurses in one remote First Nations community, while simultaneously accounting for, acknowledging, and better understanding the phenomena of interest within the larger context. 12 The Nursing Role Effectiveness Model A conceptual framework adapted from the Nursing Role Effectiveness Model (Irvine et al., 1998) informed this study. Based on and developed from Donabedien's (1980) work, the Nursing Role Effectiveness Model uses the framework of structure, process and outcome to identify how nurses contribute to clinical outcomes (see Appendix A ) . Using this framework, information relevant to assessing multiple aspects of quality of care is organized in terms of structure, process and outcome. Donabedian described structure as "the relatively stable characteristics of the providers of care, of the tools and resources they have at their disposal, and of the physical and organizational settings in which they work" (p. 81). Process refers to the "set of activities that go on within and between practitioners and patients" (p. 79), and outcome, as used by Donabedian, refers to "a change in a patient's current and future health status that can be attributed to antecedent health care" (p. 83). 6 Although Donabedian's historic work (1966; 1980) was originally directed at evaluating the quality o f medical care (i.e., physician practice), it has since been applied more generally to the evaluation of health care systems (e.g., Aday, Begley, Lairson, & Slater, 1998, as cited in Grembowski, 2001), as well as to evaluations of nursing practice (e.g., Doran, Sidani, Keatings, & Doidge, 2002) and nurse practitioner practice (e.g., Sidani & Irvine, 1999). O f considerable relevance^to this study was a recent research project that used Donabedian's structure-process-outcome framework to examine continuity of care for First Nations clients in remote communities in Ontario (Minore et al., 2002). The framework 61 have deliberately used direct quotes from Donbedian (1980) to define structure, process and outcome, to clarify that I am defining these terms in the sense that Donabedian originally proposed in his framework. I believe that over the course of the two decades or more since Donabedian first presented his framework, there has been a tendency among researchers to continue to use these terms but without clearly defining them. Subsequently, Donabedian's original meanings have become somewhat lost. That is, we presume to understand what is meant when we use the term "structure", "process" or "outcome", but without clearly defining our meaning, there is potential for the meanings to shift, creating the possibility that these terms may hold different meaning for different individuals. 13 of structure-process-outcome was used as an organizing strategy for both data collection and discussion of the study results. The Minore et al. study exemplifies and attests to the continuing usefulness of Donabedian's work to contemporary explorations of quality of health care, and w i l l be discussed further in the literature review section. Donabedian (1980) described his structure-process-outcome model in terms of well-defined concepts and relational statements, but recognized the inherent difficulty in ascribing the status of theory to a general model that encompassed multiple facets of care: I know that many ambiguities remain, as becomes apparent when one tries to classify specific phenomena exclusively under one of the three headings. This is because the three-part division is a somewhat arbitrary abstraction from what is, in reality, a succession of less clearly differentiated, but causally related elements in a chain that probably has many branches. In such a chain, each element is, at least to some extent, a cause o f the element that follows, while it is itself caused by the elements that precede i t . . . . in a chain of this kind it is futile to try to distinguish means from ends. The analyst's purposes and perspectives determine what segment of the chain he wi l l study, and what is a means and what is an end. (Simon, 1961, cited in Donabedian, p. 84) It seems clear that while Donabedian did not view his model as a theoretical model per se, he left the door open to the possibility of theoretical application, or the development of theory, by other researchers who might apply the model to more focused and specific research questions. Nursing researchers Irvine et al. (1998) adapted Donabedian's (1980) framework to facilitate a closer focus on the role played by nurses with regard to patient outcomes, but within a larger understanding o f health system structure. As mentioned, the Nursing Role Effectiveness Mode l has previously been used as the conceptual framework for research evaluating the quality of both nursing (Doran et al., 2002) and nurse practitioner practice (Sidani & Irvine, 1999). Initial empirical tests of the conceptual model have demonstrated that "the model provides a well-defined conceptual framework to guide the evaluation of 14 outcomes in acute care, in-patient settings, and to examine the contribution of nurses within these settings" (Doran et al., p. 38). While noting that further testing is necessary to demonstrate the usefulness of the model in evaluating nurses' independent role function (e.g., in a nurse practitioner role), these researchers also stated "The Nursing Role Effectiveness Model provides a theoretical basis for exploring the relationships among organizational and unit structural variables, process and patient outcomes" (p. 36). More recently, Sidani et al. (2004) discussed a structure-process-outcome model as a framework within which "proposed relationships can be specified" (p. 64), and identified this as a "theory-driven approach to evaluating quality of nursing care" (p. 60). Thus, it is evident that the Nursing Role Effectiveness Model is in the process of evolving as a theoretical model to guide the appraisal of the quality of nursing and advanced nursing practice in a variety of settings. The foregoing passage from Donabedian (1980) also implies that the relationships between structure, process and outcome are potentially multi-causal and not entirely linear, a view that is congruent with a P H C perspective. That is, Donabedian recognized that elements external to rendered health services influence health: "The specificity of many outcomes as a basis for inferences about the quality of care tends to be low. . . because many factors other than medical care influence health in a general sense, and, more specifically, the response to care" (p. 106). Sidani et al. (2004) similarly recognized that multiple factors influence outcomes: "The world of practice is a complex system o f multiple factors, multiple effects, and mutual causation" (p.60, citing Hegyvary, 1993). O f particular relevance to the present study, one of the external factors identified by Donabedian was inequity in the social valuation (and thus, access to health services) placed on "different segments of the population" (p. 16), or in contemporary language, on marginalized populations. Situating the Nursing Role Effectiveness Model of structure-process-outcome within a P H C framework 15 allows such influences to be acknowledged and facilitates a clearer articulation of the broader socio-contextual factors that influence health services structures, as well as the processes and outcomes of health care. The Nursing Role Effectiveness Model offered a tested conceptual framework by which to organize the multiple aspects of health service delivery that were of interest in this study, while simultaneously facilitating a more specific focus on relevant aspects of nursing practice. Integrating the Nursing Role Effectiveness Model within a P H C model offered a conceptual framework that supported a research focus on the specific phenomena of interest while keeping the "big picture" firmly in mind. Continuity of Care Continuity of care, as conceptualized within the framework of this study, was related to fragmentation of care, in that continuity of care was conceptualized as the inverse of fragmented care. That is, health services researchers postulate that improved continuity of care decreases fragmentation o f care (Sparbel & Anderson, 2000a; Haggerty et al., 2003). However, as these researchers have noted, one of the main challenges o f investigating continuity of care as a research phenomenon has been the ambiguity and range of definitions that historically have characterized our understanding of continuity of care (Sparbel & Anderson; Haggerty et al.). Continuity of care is a term that has been used in the health care literature at least as far back as the 1960's, when Donabedian (1966) identified it as a dimension of quality o f care, and observed the need for further conceptual exploration of this (and other related) concepts. Traditionally, continuity of care has been defined in diverse disciplinary-specific ways within the various health disciplines. Haggerty et al. (2003) described the major definitional variations in terms o f four health care domains: primary care (i.e., predominantly 16 arising from the primary care medical literature), nursing, mental health and disease management. Within primary care medicine, continuity has been defined largely as an interpersonal dimension of care: as long-term one-to-one patient-provider relationships (e.g., Burge, Lawson, & Johnston, 2003; Christakis, Wright, Zimmerman, Bassett, & Connell, 2003; Nutting, Goodwin, Flocke, Zyzanski, & Stange, 2003; Saultz, 2003). In contrast, nursing has interpreted continuity of care in terms of the co-ordinated transfer of patients and communication of information between different nurse providers and parts of the health care system (Sparbel & Anderson, 2002a; e.g., Harrison, Toman, & Logan, 1998). In the domain of mental health care, the focus has been on both the coordination of diverse health and social support services, and the stability of collaborative health care provider teams (e.g., Fortney et al., 2003; McCann & Baker, 2003). The disease management domain arose from speciality medical practice in managing complex diseases (e.g., diabetes, cardiovascular disease). The emphasis here is on the continuum of care; that is, the content and coherence of care (Haggerty et al., 2003). Thus, continuity of care has conventionally been conceptualized and defined in disciplinary-specific ways that have challenged efforts to conceptualize a multidisciplinary understanding (Haggerty et al.). However, the concept of continuity of care has recently received renewed attention in both the nursing and health services research literature. Recent work by nursing scholars (Sparbel & Anderson, 2000a, 2000b) and health services researchers (Biem et al., 2003; Donaldson, 2001; Haggerty, Reid, M c G r a i l , & McKendry, 2001; Reid et al., 2002) further clarified and defined a concept of continuity of care that has relevance from an interdisciplinary perspective and across the health care system. 17 Relevant to this study was the recent conceptual work done by Reid et al. (2002) in the context of primary care delivery in Canada. This work identified three types o f continuity: (a) informational continuity, (b) relational continuity, and (c) management continuity, and asserted that continuity o f care is a multi-faceted concept that encompasses all three of the identified types of continuity. Informational continuity encompasses the transfer of information and the accumulated knowledge of a patient. Relational continuity refers to both on-going patient-provider relationships and consistency of personnel. Management continuity depends on coordination and consistency in the approach to treatment management, and flexibility, or an individualized approach to care. Reid et al. offered this summary of continuity of care: Continuity is the result of a combination of adequate access to care for patients, good interpersonal skills, good information flow and uptake between providers and organizations, and good care coordination between providers to maintain consistency. For patients, it is the experience of care as connected and coherent over time. For providers, it is the experience o f having sufficient information and knowledge about a patient to best apply their professional competence and the confidence that their care is recognized and pursued by other providers. (2002, p. iv) A recent study of continuity of care and rural Canadian seniors (Biem et al., 2003) built on the conceptualization of continuity provided by Reid et al. (2002) by identifying seven characteristics of optimal continuity of care. Biem et al. described these seven characteristics as the "seven c's" (p.4): 1. Contact (i.e., patient-provider relationships [relational continuity]); 2. Collaboration (i.e., between providers and patients [relational continuity]); 3. Communication (i.e., between providers [informational continuity]); 4. Coordination (i.e., multidisciplinary teamwork [management continuity]); 5. Contingency (i.e., "timely access to providers for urgent questions, worsening symptoms or complications", p.4 [relational and management continuity]); 6. Convenience (i.e., o f access [relational and management continuity]); and 7. Consistency (i.e., in the approach to treatment management in spite of different providers being involved in care [management continuity]). 18 The seven characteristics of continuity described by Biem et al. (2003) were consistent with the conceptualization o f continuity provided by Reid et al. (2002), and pushed their conceptualization deeper in terms of offering a way in which to conceptualize the facets of continuity in greater detail. The "seven-c's" offered some concrete reference points from which to consider the complex phenomena of continuity of care. For instance, while relational continuity as defined by Reid et al. patently encompassed the ideas of patient-provider contact and collaboration, Biem et al.'s work defined these as specific characteristics that contributed to relational continuity. Biem et al.'s characteristics also suggested areas of overlap between types of continuity; for example, coordination related to both relational and informational dimensions of continuity. Additionally, the characteristics of contingency and convenience suggested that access to care was an important aspect of continuity that might be more clearly addressed in the Reid et al. typology of continuity. A reconceptualization of continuity of care provided by Donaldson (2001) has less relevance to the present study but is worthy of mention from the perspective of two points in particular. First, Donaldson conceptualized and defined continuity of care in terms of agency theory. The Agency Model of Continuity she presented has patent similarities to a structure-process-outcome model. In Donaldson's model, continuity is conceptualized as an intermediate outcome and agency as an outcome: continuity is "an intervening variable that may affect agency" (p. 283). While the notion that a provider-patient relationship may affect agency is congruent with the P H C principle of empowerment, viewing agency strictly within the narrow confines of that relationship, and without acknowledging the effect o f broader social determinants on agency, is incompatible with the P H C conceptual framework that guided this study. Second, Donaldson points out the fallacy of using health outcomes alone as measures of agency, and thus of the process of continuity, because health outcomes may 19 also be influenced by other factors. However, the other factors recognized by Donaldson are restricted to individual patient factors and "redundancy" (p. 284) of providers and treatments. Thus, Donaldson's model of continuity fails to acknowledge the influences on health that arise from the broader context in which the health service system operates. These identified limitations of Donaldson's conceptualization of continuity of care therefore limited its usefulness to this study. Donabedian (1980) also addressed continuity of care, but as one of a triad of attributes of quality o f care. He identified these attributes as accessibility, continuity and coordination. While distinguishing between these three features, Donabedian's description o f these attributes is nevertheless compatible with the conceptualization of continuity presented above: Care may be said to be accessible when it is easy to initiate and to maintain.... Coordination may be regarded as the process by which the elements and relationships of medical care during any one sequence of care are fitted together in an overall design. Continuity means lack of interruption in needed care, and the maintenance of the relatedness between successive sequences of medical care. A fundamental feature of continuity is the preservation o f information about past findings, evaluations, and decisions, and the use o f these in current management, or their orderly and reasonable evolution. Coordination involves the sharing of such information among a number of providers to achieve a coherent scheme o f management, (p. 22-23) Taken together, Donabedian's view of accessibility, continuity and coordination demonstrated many o f the same characteristics evident in the Reid et al. (2002) report and the Biem et al. (2003) study; for example, the communication of information between providers, and a consistent and organized approach to management of care among different providers. Donabedian (1980) not only distinguished between the three attributes of quality of care, but also between these attributes and quality of care per se. That is, the definition o f quality of care is not wholly dependent on accessibility, continuity, or coordination. However, these attributes of quality may be dependent on the definition of quality of care, 20 and in particular, on the existence of some normative definition or concept of rational care, or what might in,contemporary terms be considered evidence-based care, or "best practice". Donabedian (1980) maintained a conceptual separation among the attributes of quality o f care primarily as a means of setting boundaries on the scope of his work; significantly, he made a point of stating that other formulations of these concepts could be equally valid. Thus, the foregoing discussion supported the assumptions on which the present study rested: (a) that continuity of care may be conceptualized in a broad, multi-faceted sense that encompasses the attributes of accessibility, continuity and coordination, and (b) that continuity of care, while related to the overall quality o f care, does not in itself represent the sum o f quality of care. Evaluation of the quality of care requires a normative definition of what constitutes quality care in addition to assessment o f continuity: " I f continuity means noninterruption of needed care, the decision that care is needed involves a judgement of quality" (Donabedian, p. 23). In contemporary terms, quality, or appropriateness of care, is most appropriately evaluated in terms of evidence-based clinical guidelines or "best practice" standards. The inclusion of a component of clinical outcome evaluation in this study grounded the conceptual framework in the reality of clinical practice by allowing for an evaluation of the adequacy of care based on normative standards of practice. Ideally, some relationship must be demonstrated to exist between continuity of care and clinical outcomes (Middleton & Lumby, 1998; Sparbel & Anderson, 2002b), and ultimately, between each of these and changes in global health status. However, I was compelled, just as Donabedian (1980) was, to draw boundaries on the scope of my study. The aim of this study was to explore in a qualitative way, how outpost nurses influenced continuity of care and clinical outcomes, and not to make a judgement on the overall quality of care per se. While there has been some 21 attempt among researchers to develop measures of continuity, the current state of knowledge about continuity of care as a multifaceted and complex concept does not yet support the quantitative measurement of continuity, or thus, the possibility of making strong causal inferences about the relationships between continuity o f care, clinical outcomes and quality o f care (Reid et al., 2002). But as Donabedian claimed, understanding continuity of care is dependent on the quality of care; therefore, it was necessary to include some means of judging quality of care in this study. Considering continuity of care in relation to a structure-process-outcome model (such as The Nursing Role Effectiveness Model) raised the question of whether continuity is more properly conceptualized as part of the process of care, or as an outcome of care. In their conclusions, Reid et al. (2002) referred to continuity as a "result" (p. 15), a "product" (p. 15), and an "achievement" (p. 16), thereby implying that it may be an outcome of the process of care. But drawing on Donabedian's (1980) definition o f outcome, that is, "a change in a patient's current and future health status that can be attributed to antecedent health care" (p. 83), it is difficult to justify how continuity in itself constitutes a health outcome. Continuity, conceptualized in terms of relational, informational and management continuity, seems to fit more appropriately with Donabedian's definition of process: a "set of activities that go on within and between practitioners and patients" (p. 79). The research by Minore et al. (2002), which examined continuity of care in remote First Nations communities, also operationalized continuity of care as a process; specifically, these researchers adopted an operational definition of continuity in the process o f care, which they defined in terms of the patient receiving the required sequence of care within an appropriate time frame. Donabedian (1980) also grappled with questions o f how to conceptualize the interface between process and outcome, as well as with the question of which of these components of 22 assessment represented the best approach to evaluation. He briefly discussed the need for "immediate, process-specific 'mini-outcomes'" (p. 120), in addition to more general, inclusive (and consequently longer-term) measures of both process and health status. The following statement perhaps best summarized Donabedian's conclusions regarding the process-outcome interface: ...process and outcome are fundamentally linked in a single, symmetrical structure that makes of one almost a mirror image of the other, no matter how many attributes are used to test the relationship. Thus, the emphasis shifts to a more thorough understanding of the linkages between process and outcome, and away from the rather misguided argument over which of the two is the superior approach to assessment, (p. xi) Acknowledging that a degree of slippage exists with regard to whether continuity of care is best conceptualized as process or outcome, for the purposes of this study, it was conceptualized as an aspect of the linkage between process and outcomes of care. It was reassuring to note that as Donabedian (1980) stated, his structure-process-outcome framework was proposed "as a guide, not a straitjacket" (p. 89)! Thus, the focus of this study shifted, as Donabedian suggested, to the linkages between the process of care and outcomes of care: explicitly, continuity and specified clinical outcomes. Thus, continuity of care, in combination with specific indicator-condition clinical patient outcomes, was assumed to reflect the appropriateness of care, defined in terms of clinical practice guidelines in use at the selected study site. In turn, appropriate care was assumed to have influenced patient health status. However, consistent with the P H C orientation in which this study was grounded, it is assumed that other factors, such as the social determinants of health, also contributed to and influenced peoples' health status. So while acknowledging the broad array of factors that influenced health at every level in the "big picture" view, the particular 23 theoretical lenses that drove this study directed the focus onto the part o f the picture represented by the structure, process, and outcomes o f health services delivery. Using a broad, multidisciplinary conceptualization of continuity o f care demonstrated how continuity might be conceptualized as the inverse o f fragmented care. Fragmented care was exemplified by the idea of patients "falling through the cracks" of health care delivery. B y the same analogy, continuity of care may be represented by the idea o f ensuring the safe and efficient passage of patients as they navigate the health care system. That is, continuity of care was one strategy by which health care providers kept patients from falling between the cracks. While this evolving conceptualization of continuity of care has significant implications for the way researchers operationalize and measure continuity, much work remains to be done to further clarify and refine the concept, and the links between continuity and related concepts. This study contributes to the growing body of recent empirical work that aims to extend our knowledge of continuity of care. Summary The conceptual framework of P H C that underpinned this study directed attention to a multitude of factors that may be theorized as influencing health services delivery and health status. Theorizing potential relationships between these factors implied a level of complexity that confounded an "outcomes" study in the traditional sense. That is, although the measurement of clinical outcomes was one component o f this study, it was not an objective of the study to assign or infer cause related to outcomes in the usual linear sense. Rather, the objective was to engage in a process of preliminary theorizing, based in the relevant literature, and then to explore how factors identified within the context, structure and process of health services delivery in one remote First Nations community together influenced health 24 outcomes. Thus, the role of nurses involved in the delivery of primary care represented one key factor among many influences on health and health services delivery in the study site. 25 C H A P T E R T H R E E A C R I T I C A L SYNOPSIS O F T H E R E L E V A N T L I T E R A T U R E Introduction The literature included in this review presents a broad overview of the existing knowledge relevant to this research project. Literature relevant to describing and substantiating the context of health services delivery in remote aboriginal communities is presented first, to better situate the model of structure-process-outcome and to provide a context for understanding the multiple and complex factors that influence health services. The outpost nursing literature is then-discussed in terms of organizational and structural issues that influence the delivery of health services in Nursing Station settings, and nursing practice issues. Finally, selected literature pertaining to the evaluation of primary care outcomes is discussed specifically in relation to the assessment of core knowledge and practice competencies needed by nurses employed in a role that includes the provision of primary care. Since relatively little peer-reviewed literature exists relating to several of the identified factors, this review necessarily draws from a wide variety of sources. There is a necessary reliance on relevant grey literature, and many o f the sources included in this overview are government-sponsored reports, research reports that have not (at the time of writing) been published as peer-reviewed papers, and unpublished graduate student theses and dissertations. One of the advantages of including sources o f literature such as royal commission and other government reports is that these often represent a relatively current public perspective on contemporary issues. They are therefore consistent with a P H C orientation that values public participation and thus are potentially a powerful source of relevant knowledge. 26 Context A discussion of the contextual issues that are conceivably relevant to this research project could potentially address a very broad range of the issues that are relevant to the health and provision of health services to aboriginal Canadians living in remote communities. The discussion presented here focuses on two aspects of context that are most important to understanding and substantiating the contextual background and design of this research. One of these aspects relates to understanding the implications of "remoteness" to health services delivery in aboriginal communities. The second aspect of context addresses an understanding of how the relatively poorer health status of aboriginal Canadians, as well as health care in aboriginal communities have been influenced by socio-historical factors that continue to shape the contemporary provision of services. The Implications of Remoteness Defining the concepts, implications and issues relevant to rural and remote health research has become a focus o f attention among health services researchers within the last six years. The Canadian Rural Health Research Society was originally formed in 1999 (at that time, as the Consortium for Rural Health Research) for the specific purpose of advancing rural and remote health research. This was followed by a Canadian Institutes of Health Research [CIHR] initiative in rural health in 2001. The impetus for this growing attention has been both increased recognition of the relatively poorer health status o f rural and remote residents compared to other Canadians (Statistics Canada, 2003), and acknowledgment that health issues in rural and remote communities are different from those in urban centres. The Royal Commission Report on the Future of Health Care in Canada (i.e., the "Romanow Report") released in 2002, underscored the critical need to address these issues: "Unique rural health problems require urgent attention and unique rural conditions need to be taken 27 into account in addressing those problems" (p. 164). However, much o f the early activity was directed towards defining "rurality". The tendency has been towards rolling together the concepts o f rural and remote as though they represent one concept (i.e., "rural/remote"), rather than two related but distinct concepts, with similar and yet distinct implications for health and health services. Contemporary research and policy documents addressing health issues in rural and remote settings reflect the lack of clear conceptual definition of the terms rural and remote (Kul ig et al., 2003). Policy documents have generally failed to clearly distinguish between these two terms (Kul ig et al.). Researchers conducting their work in remote settings have typically attempted to clarify the meaning of remote by adding descriptors such as isolated, semi-isolated, or northern. For example, Vukic and Keddy (2002) described the site o f their research as a "northern remote community" (p. 542), Knopp's (2001) study site was a "remote setting in northern Saskatchewan" (p. 3), and Martin and Gregory (1996) described their setting as "northern" (p. 9), defining northern as being located north o f the 54 t h parallel in Manitoba. Boone, Minore, H i l l , and Hartviksen (2000) conducted a study about non-northern nurses' perceptions o f northern practice, but did not elucidate the meaning of northern. The study by Minore et al. (2002; also, Minore et al., 2005) was conducted in three First Nations communities that were described as both remote and relatively isolated. These terms are not defined explicitly, but their meaning is implied by further description of the geographic location of the communities (i.e., approximately 400 kilometres from a secondary referral centre) and the environmental barriers (e.g., weather) that hinder access to resources outside the community. In my own previous research I used both "northern" and "outpost" to describe similarly isolated aboriginal communities that experience disparities in access and 28 health resources, and that are further characterized by relatively poorer health than other Canadian communities (Tarlier, 2001; Tarlier et al., 2003). While descriptors such as remote, isolated, or northern usefully inform the reader as to the geographic location of the remote community, they fail to explicate the implications of being remote, isolated, or northern. A s K u l i g et al. (2003) observed, the implications of these various terms that are important to understanding health issues in these communities, has generally been assumed or taken for granted rather than made explicit. This failure to define the language in conceptual terms contributes to the overall "lack of analysis regarding the theoretical and practical meaning o f rural" (p. iv), and by extension, of remote. Thus, one of the challenges identified by rural and remote health services researchers is the need to better define the language used to describe rural and remote settings (Heath & Szpilfogel, 2002). This was the purpose of a workshop conducted jointly in 2002 by the Rural Development Institute, Brandon University, and the Department of Community Health and Epidemiology, Dalhousie University, under the auspices of the Canadian Rural Health Research Society (i.e., Workshop on Defining "Rural" and "Rurality" for Health and Health Services Research; Heath & Szpilfogel). One of the conclusions that came out of this workshop was that "meaningful distinctions between 'rural', 'remote', and 'degrees of northernness'" are needed to push forward a conceptual definition (p. 15). A second conclusion stated, "We need to be clear about who and what we are we talking about. Is the conversation about rural [remote] people, or is it about marginalized or disadvantaged populations?" (p. 16). While not discounting the likelihood that rurality and remoteness hold some positive implications for health, from the perspective of the delivery of health services, the literature supports the assertion that residence in remote communities does in fact imply disadvantage 29 and marginalization. I propose that disadvantage and marginalization represent the practical and theoretical piece that K u l i g et al. (2003) saw as lacking in their review o f policy documents. The literature that addresses health and health service delivery in remote settings clearly implies that residents of remote communities experience both poorer health and inequities in access to health services, compared to Canadians l iving in other settings (Romanow, 2002). This is supported by health indicator data compiled by Statistics Canada (2003), based on census data and data from the Canadian Community Health Survey ( C C H S ) . 7 A further and often unstated implication of remoteness is that aboriginal people make up a large segment, i f not the majority, of Canadians who live in remote communities o (Statistics Canada, 2001). Thus, when health services researchers describe a community as remote, the often-unarticulated implication is that the residents of that community are (a) predominantly aboriginal, (b) have poorer health and (c) have access to fewer health resources in their community than do Canadians who live in non-remote communities (Caron, 2005). An interesting paradox is noted in that the CCHS excluded residents of Indian reserves. Thus, these data include few if any of the primarily aboriginal communities that tend to be among the most remote communities in the northern provinces. However, aboriginal people living off reserve make up a significant proportion of the population of non-reserve northern communities. This group of aboriginal people was eligible to be included in the survey and presumably contributed proportionately. While these data substantiate that northern residents have poorer health on several parameters than do residents of Canada's southern regions, it is unclear from the CCHS data how this poorer health relates to aboriginal people. That is, it does not help to clarify whether the poor health of northern residents is more or less associated with being northern, versus being aboriginal, or with community size and remoteness (INAC, 1997), or some combination of these factors. 8 It is difficult to ascertain the exact percentages of aboriginal versus non-aboriginal residents in remote, northern regions, as the available census data are not organized in a way that makes this information obvious or easily accessible. Through a laborious method of searching the on-line demographic tables available on the Statistics Canada website for community profiles, then doing the mathematics myself, I determined that, for example, the population of the Burntwood Health Region, which represents a large part of northern Manitoba and includes remote aboriginal communities as well as the metro area of Thompson, is almost 73% aboriginal. The population of the Northwest Health region of BC (the northwest region of the province) is approximately 25% aboriginal, and the population of the Yukon Territory (excluding Whitehorse) is about 45% aboriginal. In Nunavut, almost 93% of rural / remote residents are aboriginal. 30 The first comprehensive investigation of the implications of rurality and remoteness for nursing practice in Canada was the Nursing Practice in Rural and Remote Canada Study, under the direction of co-principal investigators MacLeod, Kul ig , Stewart, and Pitblado (2004). This large, multi-armed study provided the first in-depth exploration o f nursing practice in rural and remote Canada. One arm of this study relied on analysis o f demographic data available through the Registered Nurses Database [ R N D B ] ; a second arm involved a large national survey (n = 3,933) of nurses employed in a variety of nursing roles and settings in rural and remote locations. The survey data were contextualized and enriched through analysis of qualitative data collected from nurses as part o f a third arm of the study (n = 152). This study represented a foundational work in the field of research related to nursing practice in rural and remote settings and contributions that enrich the knowledge base related to rural and remote nursing are ongoing (e.g., K u l i g , 2005; Stewart, D ' A r c y , Pitblado, Forbes, et al., 2005; Stewart & McLeod, 2005). The fourth arm of the MacLeod et al. (2004) study was the K u l i g et al. (2003) policy analysis report previously referenced. This report represented a preliminary document that provided an insightful view of the implications o f existing health policy documents on rural and remote nursing practice. The inclusion of detailed assessments of these implications from the perspectives of aboriginal health, the educational preparation of nurses, the evolving role of nurse practitioners, and health care delivery made this document of particular relevance to the present study. K u l i g et al.'s discussion of remoteness specifically in relation to aboriginal communities was based largely on documents from Health Canada and Medical Services Branch ( M S B ) (later to become First Nations and Inuit Health Branch [FNIHB]) . These authors noted that "One report implies that remoteness is related to accessibility, access to services, and support" (Health Canada, 1993, cited by K u l i g et al., p. 127). They also drew 31 attention to a F N I H B classification system for designating communities according to degree o f isolation and remoteness.9 Vukic and Keddy (2002) also referred to the F N I H B classification system, which rates aboriginal communities as non-isolated, semi-isolated, isolated or remote isolated, depending on access (i.e., road access versus air access, distance to physician services, and quality of telephone service), and the availability of health services within the community (i.e., a health centre where public health only is provided versus a Nursing Station where nurses provide primary and emergency care in addition to public health services). Notably, by the F N I H B definitions, few communities even in the far-flung reaches of the high arctic would be considered remote, isolated communities, since scheduled flights and good telephone service are now available in virtually all northern communities. For example, Igloolik, located two hours' flying time from Iqaluit, and a further three hour's flight from tertiary level services in Montreal or Ottawa, would not be considered a remote isolated community. However, the F N I H B classification system illustrates how access and resources to services are implied in the way a continuum of remoteness has been conceptualized. 1 0 The Multistakeholder Index of Rurality ( C M A et al., 2003) represented a joint effort o f the Canadian Medical Association ( C M A ) , the Canadian Nurses Association ( C N A ) , the Society of Rural Physicians of Canada, and the Canadian Pharmacists Association. This method for indexing rurality was of interest to the present study primarily because it included parameters of rurality that were also relevant to remoteness, such as distance to a secondary 9 Interestingly, the FNIHB classification system is different from the definitions of geographical zones used by INAC, wherein a "remote zone" is defined as being more than 350 km from the nearest service centre having year-round road access, a "special access zone" is defined as having no year-round road access (i.e., a fly-in community), and a "service centre" is defined as a community having a variety of specified services available, including provincial health and community services (INAC, 2003). This conceptualization begs the question of whether an isolated community is still considered to be isolated to the same degree on the intermittent days that a physician is on site? 32 referral centre, barriers to access, and available resources in the community. However, the index was premised on the assumption that communities were large enough to support a full-time physician practice; therefore, several parameters are specific to physician practice. The rurality index has no way to acknowledge the reality o f health services in remote aboriginal communities that typically lack full-time physician services. Recognizing this shortcoming in respect to remote communities, the C M A et al. document recommended that research is needed to start looking at the differences between not only rural and remote communities, but also between rural, remote and aboriginal communities (p. 24). While the recruitment and retention o f health professionals in rural and remote areas was the focus of both the Ku l ig et al. (2003) report and the C M A et al. (2003) document, each of these documents was driven by recognition of the poor health status of rural and remote residents, and the assumption that improving access to health providers is one strategy to effect better health. The implication is evident that remote communities are associated with a less healthy population that is marginalized in terms of access to health services: The poor health status of rural and remote populations compared with their urban counterparts, as well as the challenges they encounter in accessing health services, have been well researched. Unfortunately, these health inequities are exacerbated by the difficulties that rural and remote communities experience in retaining and recruiting health care providers. The need to address health care workforce issues is therefore a critical component in helping to ameliorate the health status of rural and remote populations. ( C M A et al., p. 1) Remoteness may also imply that remote communities have in some sense dropped through the cracks of broader society. They are remote in the sense of being remote from mainstream Canadian society. This assertion is particularly germane in the case of aboriginal communities. For example, aboriginal communities are not included in the same databases (i.e., the C C H S database); thus, it is easier to lose sight o f their health status in comparison to 33 other Canadians. Neither (in the provinces) are they part of the same health care system, that is, the provincial systems. B y virtue of being distant from the places where provincial health services are located, remote aboriginal communities have been forced to rely on a federal system that has developed, in theory, to meet their health care needs in the absence of available provincial health services. However, the distancing and isolation of remote aboriginal communities from mainstream society has to do more with the historical and socio-political context of aboriginal existence in Canada than with geographical location (Browne & Fiske, 2001). The Historical and Socio-Political Context of Aboriginal Health and Health Services In terms of health status indicators, numerous studies and official statistics consistently demonstrate a wide gap between Indian and non-Indian Canadians.... Concern is often expressed over the inadequate services provided by government agencies and the lack of accessibility of many Indians to the same quality of care enjoyed by the majority of Canadians. (Young, 1984, p. 257) The poor health status of Canada's Aboriginal peoples is a well-known fact and a serious concern not only to Aboriginal peoples but also to all Canadians. The situation is simply unacceptable and must be addressed . . . . [In spite of] a number of initiatives underway in every province and territory, the fact remains that there are deep and continuing disparities between Aboriginal Canadians both in their overall health and in their ability to access health care services. (Romanow, 2002, p. 211) These two passages conveyed the same message: aboriginal Canadians experience poorer health and inequitable access to health services, compared to other Canadians (CIHI, 2004a; Caron, 2005). Significantly, these passages were written almost 20 years apart, suggesting that in spite o f recognition o f the problems, little substantive progress has been made over the last two decades in addressing aboriginal health issues. A s Romanow (2002) noted, some improvements in health status have in fact been achieved (CIHI), primarily attributable to public health strategies such as improved housing (in some communities) and disease prevention. However, the prevalence of chronic diseases such as diabetes (Green, 34 Blanchard, Young, & Griffith, 2003; Young et al., 2000) and heart disease (Shah et al., 2000), infectious disease such as hepatitis A (Jin & Martin, 2003) and tuberculosis (CIHI; I N A C , 2003), as well as the incidence of suicide (Malchy, Enns, Young, & Cox, 1997) and traumatic death and disability (CIHI; I N A C ; Karmali et al., 2005; Young, 2003) continue to be significantly higher in the aboriginal population compared to the non-aboriginal Canadian population. These profound health disparities are further reflected in health status indicators such as infant mortality rates that remain a third higher than the national average (CIHI; I N A C ) , and between 5.5 (women) and 8.1 (men) years' discrepancy in life expectancy between aboriginal and non-aboriginal Canadians ( F N I H B , 2005). Significantly, these statistics are not simply neutral indicators of health and well being, but reflect the impact of social inequities and the social determinants of health on the health of aboriginal people. A n understanding and acknowledgment of the historical and socio-political context of aboriginal health is relevant to this research for three main reasons. First, understanding the phenomena o f interest within its broader context supports a P H C orientation to the research and is consistent with the conceptual underpinnings of this study. Young (2003) pointed out that contextual factors such as geographic isolation and socio-economic status "are central to the aboriginal experience in many regions and removing them takes away the most powerful explanatory variable" (p. 121). Second, understanding the broader historical, socio-political and economic context of aboriginal health is key to understanding how the organization and structure of aboriginal health services have developed at the macro and meso-levels, and how and why these services differ from health services to non-aboriginal Canadians (O 'Nei l , 1986). Third, previous studies suggest that the historical and socio-political context of aboriginal health influences patient-provider interactions (e.g., Browne, 2005; Browne & Fiske, 2001; O ' N e i l , 1989; Vukic & Keddy, 2002); thus, the process o f care, and potentially, 35 health outcomes (Browne & Fiske). A contextual understanding of the broader issues that influence aboriginal health and health services added depth to the present study by offering direction for data collection and facilitating a more insightful analysis of the data. A few classic papers dating back to the early 1980's have formed the foundation of a growing body o f literature that has effectively established that the poorer health of aboriginal people is linked to social inequities (O 'Nei l , 1986, 1989; Young, 1984; York , 1989), and that these inequities are the product o f the historical social, political, and economic relationships between aboriginal people and (predominantly) European settlers. These are relationships that have been characterized as colonial in nature (Hackett, 2005). For example, in a synopsis of the key historic developments and government policies that have shaped aboriginal health services, Young stated, "the dominant philosophy can be described as benevolent paternalism" (p. 260). Young related this philosophy to the federal government policy of the time (i.e., dating from Confederation until the post Wor ld War II era), which was a policy of displacement and assimilation of aboriginal people ( R C A P , 1996). Driven by this policy, health services to aboriginal people at the time were in reality little more than "benign neglect" (Young, p. 262). Even following the institution of Medical Services Branch" in 1962 (Young, 1984), as a branch of the federal Department of National Health and Welfare 1 2 dedicated to providing health services to aboriginal people, aboriginal health services have continued to be characterized by colonial attitudes and paternalism (Gregory, 1992; O ' N e i l , 1986) and the belief that aboriginal people were wards of the state, incapable of controlling their own 11 Renamed First Nations and Inuit Health Branch (FNIHB) in 2000. 1 2 Later Health Canada. 36 health services. The locus of authority and responsibility, and thus, control over health services remained firmly in government hands (Lavoie, 2004; O 'Nei l ) . The need to repatriate authority and control for health services was recognized in the late 1970's as a first step in rebuilding a system of health care that could better address the profound health issues that had become increasingly evident in the aboriginal population. The initial impetus for the eventual transfer of authority for health services back to aboriginal people came primarily from a new federal Indian Health Policy, implemented in 1979, and the Berger Royal Commission Report of 1980 (Vukic & Keddy, 2002; Young, 1984), which recommended consultation and participation of aboriginal groups in health care decision-making. Since then, health transfer has been a slowly-evolving ideal that proceeds despite some scepticism and the critique that it may represent an "attempt to off-load federal responsibilities for health care that may disadvantage communities as they inherit underfunded, medicalized systems of health care" (Browne & Fiske, 2001, p. 129). However, the gradual and responsible transfer of authority for health back to aboriginal groups is seen as the definitive strategy for eventually effecting improvements in aboriginal health status (Lemchuk-Favel & Jock, 2004). A s of December 2004, health transfer agreements had permitted almost half (48%) of First Nations communities to assume greater responsibility and control over their health care resources (FNIHB, 2004). A number of contemporary government reports have recognized the link between poorer aboriginal health and the social inequities that are the legacy o f past government policies that have been racist and discriminatory (Romanow, 2002; R C A P , 1996; Seaton et al., 1991). The R C A P report has been the most definitive of these, clearly l inking the present with the past: "Current social problems are in large part a legacy of historical policies of displacement and assimilation, and their resolution lies in recognizing the authority of 37 Aboriginal people to chart their own future" (vol. 3, Introduction). However, it is questionable to what extent other official federal and provincial government reports have acknowledged the link between poor health status and the broader socio-political context of aboriginal health, which is the cumulative result of over a century o f oppressive government policy and which is now embedded into the fabric of Canadian society (Browne & Smye, 2002). For example, while Romanow observed that "young Aboriginals are more often exposed to problems such as alcohol abuse and drug addiction than other Canadians of the same age" (p. 218), he failed to clearly attribute the problem of widespread substance abuse to the socio-historical context, stating instead that due to such abuse ''combined with [emphasis added] pervasive poverty, persistent racism, and a legacy of colonialism, Aboriginal peoples have been caught in a cycle that has been perpetuated across generations" (p. 218). This is a subtle difference in semantics that is nevertheless significant, in that it is indicative of the tendency to recognize the different pieces of the puzzle without accurately acknowledging how the pieces are linked together. Framing the issues in this way obscures how social problems such as substance abuse are in fact deeply rooted in the socio-historical context of poverty, racism and colonialism (Hackett, 2005). Browne and Fiske (2001; also Browne & Smye, 2002) noted a similar issue in regard to health research, and in particular, epidemiological research that has inadvertently reinforced and perpetuated negative stereotypes of aboriginal people as passive, dependent and irresponsible, thereby "reinforcing unequal power relationships and justifying ongoing paternalism and dependency in health care" (p. 129). For example, attributing diabetes risk to factors such as genetic susceptibility, obesity or inactive lifestyle, or trauma risk to driving and alcohol use, negates the role o f social determinants such as poverty, unemployment, or lack of access to resources (Caron, 2005), as well as the socio-historical context that may 38 potentially influence the development of risk factors such as alcohol use, obesity or inactivity. Moreover, the effects o f the social determinants on health, (e.g., poverty), have often been treated as confounding effects to be "controlled for", rather than acknowledged as potentially critical explanatory variables (MacMil lan et al., 1996; Young, 2003). Decontextualizing research findings in this way not only reinforces the negative images of aboriginal people that perpetuate the cycle of dependency, but also contributes to an attitude of victim-blaming (Bartlett, 2005; O ' N e i l , 1989), wherein individuals are seen as lacking the wi l l - rather than the resources, to improve themselves. There are several contemporary, comprehensive histories of the evolution of aboriginal health services from the perspective of nursing (e.g., Browne, 2003, 2005; Martin, 1997; Vukic , 1997), although notably, these tend to exist as the unpublished portion of 13 master's and doctoral theses. For example, Martin presented a history o f aboriginal health care from pre-contact (i.e., prior to 1670) to the present. This work included a description of common traditional health practices and illustrated how contact with colonizing influences changed both health practices and the health of aboriginal people. Mart in drew attention to racism from a historic perspective, and to the insidious development o f racializing discourses based in economic and political colonialism, and described how these discourses continue to influence both aboriginal health services and the practice of outpost nurses who provide health services in remote settings. O ' N e i l (1989) studied physician and nurse-patient interactions in a remote Inuit setting and found that these interactions, which were mediated through an interpreter, were constrained by "political and ideological barriers inherent in a colonial medical system" (p. 1 3 Graduate theses and dissertations are now widely accessible through the National Library of Canada and National Archives of Canada electronic collection and thesis retrieval website: 39 325). He referred to the persistence of such colonial relationships as "internal colonialism" (O 'Nei l , 1986, p. 119), referring to the deliberate or incidental subjection of an indigenous population by a more powerful mainstream population. The constrained interactions described by O 'Ne i l contributed to a loss of confidence in local health providers on the part of the local Inuit population, and subsequent confrontation and conflict between the Inuit community and local providers, as well as the larger health system structure (i.e., at that time, M S B ) . O 'Nei l ' s work offers a graphic illustration of how the socio-political history of colonialism continues to influence and shape contemporary health services delivery in aboriginal communities. More recently, from the perspective of nursing practice in remote aboriginal communities, Vukic and Keddy (2002) found that the ability of nurses to develop trust with aboriginal clients was constrained by the "structural, administrative, historical, cultural and political contextual realities that have shaped northern community nursing" (p. 542). Similarly, Browne and Fiske (2001) found that "affirming encounters" (p. 139), or provider-patient interactions that were viewed as positive by First Nations women, were characterized by respect and trust. The lack of trust within provider-patient relationships has implications for health services delivery, nursing practice and patient health outcomes at both the aggregated, or community level (Vukic & Keddy), and at the individual level (Tarlier, 2004). Contemporary nursing scholars have turned a critical lens on the structures and processes that typify interactions between nurses and aboriginal patients. Recent work, both published and unpublished, demonstrates how critical social perspectives have informed and added depth to our understanding of how the historical and socio-political context of the organization, structure and processes o f health services to aboriginal people continue to be influenced by internal colonialism and racializing discourses. For example, feminist and 40 critical social theory informed a multiple case study of the nature of primary health care nurse practitioner work (Martin-Misener, 2000). 1 4 Perspectives from critical medical anthropology and cultural safety informed Browne and Fiske's (2001) study o f First Nations women's interpretations of provider-patient interactions. Vukic and Keddy's (2002) study examined the nature of outpost nursing work using an institutional ethnography that identified "disjuncture in the discourse of northern nursing" (p. 542), while another recently completed ethnography (Browne, 2003, 2005) relied on post-colonial, critical and feminist perspectives to analyze interactions between nurses and First Nations women patients in an acute care setting. Previous research grounded in critical perspectives suggests that maintaining a broad contextual understanding of the structure and process of care is vital to an accurate interpretation that is meaningful from the perspective of the aboriginal people whose health is central to the aims of the present project. While this research did not take a specifically critical theoretical stance, it was nevertheless important to acknowledge that aboriginal health services take place within a historical socio-political context that has shaped, and continues to shape not only the traditional and emerging organization and structure o f health services at the macro and meso-levels, but also the processes of health care at the meso and micro-level; that is, at the level of individual providers and patients. Outpost Nursing Literature Gregory (1988) noted the lack of literature, and particularly, of research-based literature related to outpost nursing in Canada. More recent writers have made the same observation (Chaytor 1994; McLeod , 1999; Tarlier et al., 2003), suggesting that outpost 1 4 While Martin-Misener's study did not focus specifically in the area of health services to remote aboriginal communities, most of the PHC nurse practitioners (n = 5) that took part in this study did in fact work in a remote aboriginal community, or in a rural or urban health center where aboriginal patients formed a significant part of the patient population. 41 nursing may be an under-recognized field of nursing practice. Much of the literature that does exist addresses organizational and structural issues as well as practice issues, suggesting that it is difficult to separate outpost nursing practice from the context in which it occurs. While outpost nursing is unique to the Canadian context, both Australia and the U . S . have experienced a parallel development of similar nursing roles in rural and remote areas. For example, the Frontier Nursing Service in the Appalachian region of the U.S. , and Remote Area Nursing in the Australian outback. These nursing roles each evolved to meet the health needs of under-served rural and remote populations, and they continue to face common issues and challenges in the present day. Shared themes include working with marginalized populations, lack of resources and supports, and identifying the role as an advanced nursing role (Bushy, 2002). Australia and Canada struggle with similar issues with regard to the historical socio-political context of aboriginal health, socio-economic inequities and the need to offer culturally relevant health services (Nel & Pashen, 2003; Wilson, 2001). Bushy noted a lack of literature about rural and remote nursing at the international level, as well as a need to identify "core nursing content" to support advanced rural / remote nursing practice (p. 109). Despite these similarities, the nursing role in each of the three countries has also been shaped by unique considerations, such as the national health services system, health care funding, political and economic ideologies, and the particular physical characteristics of each country (i.e., weather, geography). For example, early acceptance of the N P role in the U.S . led to the development of a recognized N P certificate program (now a graduate level N P degree program) as early as 1970, to prepare nurses for work with the Frontier Nursing Service (Frontier Nursing Service, 2002). Although an outpost nursing type of role exists in the U.S . and Australia, these roles are carried out within distinct social and health services 42 contexts that limit the usefulness of the American and Australian literature to the purposes of this study. Thus, this literature review focuses on outpost nursing in the Canadian context. A common theme through most of the outpost nursing literature is that historically, there has been a notable lack of support and resources available to outpost nurses, and that this has influenced both the nursing role and practice (Gregory, 1992; MacLeod, 1999; Tarlier et al., 2003; Vuk ic & Keddy, 2002). For example, how outpost nurses learned to adapt to the context of practice emerged as a sub-theme in my previous research with outpost nurses (Tarlier et al.). While this section of the literature review has been divided into two main sections that address (a) organizational and structural issues, and (b) practice issues, it is with the recognition that in reality, no clear demarcation separates these two areas. Organizational and Structural Issues In previous studies, organizational and structural issues arising from the context and structure of nurses' work and living environments in remote aboriginal communities emerged as important influences on nurses' practice. These issues are discussed here in terms of four key themes that are drawn from the existing relevant literature: (a) isolation, (b) understanding the context of aboriginal health, (c) nursing turnover, and (d) worklife issues. The literature reviewed here offers an important background to the present work. Isolation One of the earlier published studies to look at outpost nursing specifically was Hodgson's (1982) anthropological research with outpost nurses in northern Manitoba. Hodgson identified how organizational and structural aspects o f working and living in remote aboriginal communities contributed to role ambiguity and conflicting role expectations for the nurses in her study. Nursing practice expectations (i.e., advanced and expanded role and level o f autonomy), social and living conditions (i.e., communal housing attached to the 43 Nursing Station), and living and working within an unfamiliar culture were all viewed by Hodgson as contributing to nurses being distanced from, and therefore isolated from the aboriginal community. These issues were experienced as a source of stress that in turn affected nurses' ability to form satisfactory interpersonal relationships within the aboriginal community, and thus also affected both their work lives and nursing practice. Hodgson discussed the effect of such isolation on nursing practice indirectly in terms of its negative influence on communication and building relationships with the aboriginal community. Gregory's (1988, 1992) work has become a classic reference in the Canadian outpost nursing literature because it presents a clear, concise description and analysis of how the organization and structure of outpost nursing practice and health services delivery in remote aboriginal communities have been shaped by the broader socio-political historical context. Drawing on Hodgson's (1982) work, Gregory similarly discusses the key issues in terms of nursing role expectations, and l iving and working in isolation and within a different culture. Nurses in remote aboriginal communities tend to be isolated from not only the personal, professional and social supports they have left behind, but also from the aboriginal community itself. Most non-aboriginal nurses are unprepared for the degree of "culture shock" (Gregory, 1992, p. 188) they experience on taking up an outpost nursing position in a community that is unexpectedly foreign in terms of culture, language, socio-economic status, and often, the extent of social problems such as substance abuse and violence. More recently, nursing scholars have conceptualized the notion of nurses' isolation from the community as "being other" in the community (Tarlier, 2001; Vuk ic , 1997; Vukic & Keddy, 2002). Interestingly, Martin's (1997) study compared quality of worklife issues as experienced by aboriginal versus non-aboriginal nurses. Differences between aboriginal and non-aboriginal nurses who participated in Martin's study were based largely on the 44 aboriginal nurses' being able to integrate into the community better. That is, they were less "other" than the non-aboriginal nurses because they understood the local language and culture, and were more readily accepted by patients and the community. Both aboriginal and non-aboriginal nurses perceived this as positively influencing patient care, in that aboriginal nurses were seen as being more effective. Partly as a result of this, the aboriginal nurses also tended to be more satisfied with their worklife. Nurses often find themselves in highly charged political situations, wherein aboriginal community leaders perceive nurses as the representatives of an antagonistic federal bureaucracy: "Nurses unaware of political realities may face a significant power struggle in their interactions with community leaders" (Gregory, 1992, p. 190). Being in some sense held accountable for the actions of the larger bureaucracy is another way in which nurses experience being "other" and are distanced from the community. The politicized nature of aboriginal relationships with the larger health services delivery organization may also create distance between nurses and their supervisors, as nurses are caught in the middle of a health services "tug o' war", trying to meet the sometimes conflicting demands of both their employing organization, and the patients to whom they owe a duty of care. The socio-historical development of health services to aboriginal people has resulted in services being administered externally and at arm's-length from the communities (O 'Ne i l , 1986). Paradoxically, as politicians and bureaucrats at higher levels of government have become acutely sensitive to aboriginal leaders' concerns, nurses have become increasingly vulnerable; often taking the brunt of a conflictive situation that may have developed historically between their employer and the local community (Gregory, 1992). For example, nurses are vulnerable to "being B C R ' e d " , which refers to the ability of local Chief and Council to pass a Band Council Resolution requiring the immediate removal 45 of a particular nurse from the community. Because Band administrators, such as chiefs and mayors "have direct access to powerful individuals in the departments of Indian Affairs and Health and Welfare" (Gregory, p. 190), a B C R is rarely questioned at the local or regional supervisory level. To illustrate, a nurse following F N I H B ' s policy regarding the number of Tylenol 3's that may be legally dispensed from the nursing station pharmacy may face being B C R ' e d i f a dissatisfied patient lodges a complaint with the Band administration. Thus, the phenomena o f B C R reflects the larger socio-political context that has led to political strife between communities and federal agencies at the macro and meso levels, but which also influences nursing practice and patient care at the micro-level, as nurses sometimes find themselves walking a fine line between balancing employer expectations (and standards of practice) with community expectations, where failure may mean loss of position and expulsion from the community. 1 5 Vukic (1997; Vukic & Keddy, 2002) found that nurses' status as other in the community effected their ability to build trusting relationships in practice. Trust was viewed as fundamental to being able to work effectively with the community (Vukic & Keddy, p. 542). Vukic and Keddy linked difficulties in building trust to the broader socio-political context: "the difficulty in part stems from the constraining, structural, administrative, historical, cultural and political realities that have shaped northern community nursing" (p. 542). The finding that experienced outpost nurses developed skil l in building trusting relationships and viewed it as essential to their ability to influence patient health outcomes was similarly reflected in the findings of my own previous research with outpost nurses 1 5 The power to pass a BCR also has a positive side, in that it gives aboriginal communities a powerful option for dealing with nursing station administration or nurses that have proved unresponsive to community concerns. The purpose in discussing BCR in this paper is to present the risk of being BCR'ed as a factor that may potentially influence nurses' practice and interactions with patients. 46 (Tarlier et al., 2003). This finding was one o f four themes to emerge from the data of an interpretive study describing the role and practice of experienced outpost nurses. This study was also a master's research project (Tarlier, 2001) and therefore somewhat limited in scope. Nine experienced outpost nurses participated by sharing practice narratives, which were interpreted in accordance with Benner's model of interpretive phenomenology (Benner, 1994). The other three themes that emerged from the interpretation of findings in this previous work were: (a) nurses evolved into the outpost nursing role by learning to adapt to the context and by developing an understanding o f primary health care; (b) experienced outpost nurses became comfortable with the autonomy and responsibility of outpost practice, which is related to the advanced and expanded expectations of the primary care nursing role (i.e., as defined by the F N I H B nursing scope of practice, reviewed in the following section that addresses nursing practice issues); and (c) experienced outpost nurses perceived competency in primary care as fundamental to outpost nursing practice. Notably, the present study built on and extended this previous research. Understanding the context of aboriginal health Martin (1997) briefly addressed the consequences of outpost nurses failing to examine their practice within an understanding of its historical and socio-political context. Identified consequences included having a limited perspective on aboriginal health care issues, failing to address the source of problems that nurses perceived to effect their practice, and failing to understand the broader picture of health that implied a need for community development approaches. Martin's findings in relation to both the ethnocentrism of non-aboriginal nurses and the failure of non-aboriginal nurses to examine their practice from a broad historical and socio-political perspective are consistent with a recent doctoral study conducted by Browne (2003, 2005). 47 Browne (2003, 2005) used ethnographic methods informed by standpoint theory (Smith, 1987, 1992, as cited in Browne, 2003) and post-colonial and feminist perspectives to explore the socio-political and historical context of clinical encounters between First Nations women and nurses in an acute care setting. Browne found that " in the absence o f competing frames of reference, embedded assumptions about Aboriginal peoples, culture and 'difference' influence the relational aspects of nurses' work with First Nations women" (2003, p. i i) . Racializing discourses, reflecting embedded societal discourses and assumptions, created and perpetuated a process of social distancing between nurses and First Nations patients that in this predominantly non-aboriginal work setting contributed to othering practices directed at First Nations women. While Browne's study did not examine nursing practice or outcomes of practice per se, her observations suggested that embedded racism and insidious discriminatory practices exerted a negative influence on communications and relationships between the acute care nurses and the First Nations women patients who participated in her study. Nursing turnover Several of the papers already presented in this review have referred to or have implied nurse turnover, or transience in staffing as an important issue that effects outpost nursing practice (Gregory, 1992; Martin & Gregory, 1996; Minore et al., 2002, 2005; Vuk ic & Keddy, 2002). Nurse turnover has long been a critical health services issue in remote aboriginal communities (Morewood-Northrop, 1994) and is related to many o f the organizational and structural issues already delineated; for example, isolation, culture shock, powerlessness, unpleasant living conditions, l iving in fear, lack of resources, lack of support from and trust in the nursing supervisory structure, and inadequate preparation to meet the expectations of an expanded scope of practice. Martin and Gregory discussed nurse turnover 48 in terms of increased workload arid ultimately, nurse fatigue. Excessive workloads and fatigue contribute to the "burnout" and rapid turnover of outpost nurses that is typical of many communities (Kul ig et al., 2003; Vukic & Keddy), and also to the difficulty in recruiting and retaining adequately qualified nursing personnel. The Sioux Lookout region in northern Ontario recently commissioned a study to explore non-outpost nurses' perceptions of outpost nursing practice (Boone et al., 2000), as a recruitment and retention strategy. While this study sampled nurses (n = 237) who had not necessarily experienced outpost nursing work first hand, study findings supported all of the organizational and structural issues already identified in the literature as being important to nurses' decision to pursue or not pursue outpost nursing. Minore et al. (2002, 2005) identified nurse turnover and workforce instability as a potential factor contributing to lack of continuity of care and poor patient health outcomes in remote First Nations communities. Although organizational and structural issues may explain a large part of the high rate of nurse turnover in remote communities, the literature also supports the possibility that nursing practice issues equally influence high turnover rates (Morewood-Northrup, 1994). Worklife issues Martin and Gregory (1996) presented preliminary findings from Martin's (1997) master's thesis research, which was an ethnographic study of quality of worklife issues of outpost nurses in northern Manitoba. Worklife issues were defined in terms o f four categories: (a) the interplay o f home and work life, (b) the nature of nursing work, (c) the nature of the work environment, and (d) the larger health care system. The theoretical framework (O'Brien-Pallas & Baumann, 1992, as cited by Martin & Gregory) that guided Martin's study postulated that aspects of nursing worklife influenced both nurses and their practice with patients. Having been conducted as a master's thesis, the scope of the study was 49 small, being based on a total of 10 days' fieldwork in four different remote aboriginal communities, and interviews with 11 outpost nurses (n = 5 aboriginal and 6 non-aboriginal nurses) Consistent with Hodgson's (1982) research, the findings o f Martin's (1997) study suggested that outpost nursing "was inundated with contradictions and conflicts" (p. 241). Quality of worklife issues perceived by both aboriginal and non-aboriginal nurses to negatively influence patient care were nurse fatigue (i.e., in response to excessive hours worked, on-call hours worked, and lack of sleep secondary to being on-call at night), lack o f resources (e.g., understaffmg, mental health resources), and conflicts with physicians (i.e., physicians at a distance who provided telephone referral / consultation services but who did not visit the community and thus had no particular on-going relationship to the nurses or the patients). The quality of worklife issues described by Martin are not unique to nursing practice; these issues may also affect the quality of worklife of other health professionals (e.g., physicians) who practice in rural, remote or under-serviced areas. A key difference is that in remote outpost settings, nurses take on responsibilities that are more usually associated with physician practice, for which nurses are sometimes inadequately prepared (Minore et al., 2002), and that are not generally considered usual nursing practice in other settings. Minore et al. (2002) found that nurses' relationships with distant physicians exerted a negative influence on inter-professional communication and access to health services. A s Gregory (1992) had previously observed, the politicized environment was also perceived to exert a negative influence over nurses' practice. Martin (1997) characterized the political influence over various aspects of nurse's worklife as powerlessness: 50 Politics was a key factor influencing nurses' relationships with clients, NICs [nurses in charge], ZNOs [zone nursing officers] and particularly the Band Chief and Council. Many nurses felt powerless, given the Band Chief and Council's perceived influence over their work. Several participants identified that political structures and processes impeded their roles as nurses, (p. 231) Other factors embedded within the larger organization and structure of the health services delivery system were identified as nurses' lack of support from and trust in ZNOs (i.e., the immediate supervisors of nurses in the field), isolation, and living conditions. Moreover, Martin's (1997) study revealed outpost nurses were sometimes "living in fear" (p. 229) in regard to their personal safety both in the workplace and when out in the community. Nursing Stations riddled with bullet holes and elaborate (if often ineffective) security alarm systems attest to the fact that in some communities, the Nursing Station is seen as a tangible symbol of oppressive colonial authority. Nurses have at times been caught in the crossfire (occasionally quite literally) that is the result of the socio-political legacy of aboriginal relations with federal governmental authority (Hackett, 2005). Nurses are also placed at risk as the front-line providers of health services to patients with the type of mental health and substance abuse problems that are largely secondary to the socio-economic inequities that have shaped contemporary aboriginal health issues. In some communities, nurses still work on-call alone at night and without the benefit of support, such as a reliable local policing presence.16 This, along with the social distancing and the isolation of nurses in some communities, makes nurses a vulnerable target within the community. Notably, this is the exception rather than the rule; my personal clinical experience suggests that, as in other parts of the world, the communities where nurses are most at risk are often the communities most plagued by socio-economic inequities, that is, the most impoverished communities. 1 6 FNIHB-operated Nursing Stations in Manitoba now have security guards on site during the night; however, this is not the case in all provinces or territories, where nurses commonly continue to work alone when on-call. 51 Vukic (1997; Vuk ic & Keddy, 2002) conducted an institutional ethnography to explore the social organization of outpost nurses' work. Vukic ' s study was informed by Smith's standpoint theory (Smith, 1987, as cited in Vukic & Keddy). In the review of outpost nursing literature that supported her study, Vukic (1996; Vukic & Keddy) identified five themes that were relevant to the social organization of outpost nursing work: (a) the expanded role of outpost nurses, (b) the cultural context of practice, (c) the context of northern health care, (d) the changing role of outpost nurses (i.e., an increasing emphasis on P H C ) and (e) quality of worklife issues. Vukic isolated two further variables from the available body of outpost nursing research literature that pertained to the quality of worklife: professional isolation and a lack of adequate educational preparation of outpost nurses. The findings of Vuk ic ' s (1997; Vukic & Keddy, 2002) research included identifying factors arising from the organization and structure of outpost nursing work that influenced nurse's practice. Vukic discussed her findings in terms of documentation practices and the nurse as other in the community (Vukic & Keddy). Documentation practices imposed on nurses by the organization within which they worked were viewed as shaping nursing practice. "The documents organize, regulate, and manage what work should be done, thus enforcing compliance in doing and recording what was done" (p. 545). Examples of types of documentation included the nurses' daily patient logs, the month-end report, and the unit work plan. Although the nurses in Vukic ' s study expended considerable effort towards meeting the requirements of the documentation expected of them, she found that the documents failed to "reflect the complexity of the decision-making process" (p. 545) that nurses engaged in as part o f their daily practice. Moreover, while documentation tended to increase the visibility of task-oriented nursing work, the lack of documentation regarding other less task-like aspects of nursing perpetuated the invisibility and therefore the 52 insignificance of such other aspects of work that nurses' considered important to their practice. For example, nurses' work in building trusting relationships with patients and within the community went undocumented and thus remained largely invisible and unvalued. Martin (1997) similarly identified several factors related to both nurses and patients that negatively influenced nurses' practice. Patient factors included demands on nurses that were sometimes perceived as unreasonable and unrealistic, over-dependence on the health system, failure to take responsibility for health at the community level, and valuing acute care services at the expense of community health programs. Nurse factors included inadequate educational preparation, inexperience in nursing generally and in the outpost nursing role, and attitudes and values. For example, Martin found the non-aboriginal nurses who participated in her study exhibited ethnocentric attitudes and values: "They expressed and demonstrated a superior value for their own cultural beliefs, attitudes and norms" (p. 258). In my previous study of outpost nursing (Tarlier et al., 2003), nurses' evolution into the outpost nursing role was related to the contextual, organizational and structural factors that influenced outpost practice. Two sub-themes suggested that: (a) nurses adapted to the context of practice (i.e., adapted to the culture, isolation, and lifestyle), and (b) nurses shifted to a primary health care orientation to practice as they gained experience in the role and developed a "big picture" understanding o f the role played by the social determinants o f health. That is, while experienced outpost nurses regarded primary care competencies as essential to practice, at the same time, they understood that primary care by itself represented little more than a "band-aid solution". Comprehensive community-based health prevention and promotion, as well as community development strategies were required to address the root causes of poor health in the communities in which nurses practiced. 53 Nursing Practice Issues The outpost nursing literature suggests that outpost nursing is an advanced community-based nursing role that requires advanced nursing knowledge and skills as well as specific knowledge and skills in primary care. This combination of advanced and expanded nursing competencies is most often described in the literature as a nurse practitioner (NP) role. Gregory (1992) saw there was a "substantial need for nurses who possess nurse practitioner skills" (p. 196) to provide health services in remote aboriginal communities. MacLeod, Browne, and Leipert (1998) suggested that the ability to practice in an advanced community health nursing role in addition to offering primary care services was a potential benefit of employing advanced practice NPs in rural and remote settings. In view of research findings suggesting that outpost nurses offer primary care services but sometimes at the expense of community health care (Martin, 1997; Minore et al. 2002; Vukic & Keddy, 2002), employing advanced practice NPs in outpost settings may represent a significant potential benefit. In concluding their comprehensive analysis of policy documents pertaining to nursing practice in rural and remote settings in Canada, K u l i g et al. (2003) recommended the implementation of national initiatives to support and promote the recognition, regulation and integration o f advanced N P roles in health services to rural and remote communities. NPs able to work to a "full scope of practice" (p. 66) were viewed as having the requisite knowledge and skills to address health service delivery issues in rural and remote areas. A common finding in the outpost nursing literature is that outpost nurses have in fact historically been identified as nurse practitioners. In interviews with graduates of the Dalhousie Outpost and Community Health Nursing diploma program, Chaytor (1994) found that this group of outpost nurses believed themselves to be functioning in a N P role. Martin-54 Misener, Vukic , and M a y 1 7 (1999) viewed the outpost nursing role as similar to that o f a primary health care N P but differentiated the two roles on three points: (a) outpost nurses were responsible for initial management of emergency situations in settings where there were no supportive back-up services (i.e., physicians), (b) the outpost nursing role encompassed responsibilities in community health and community development, and (c) outpost nurses practiced within specifically aboriginal cultures and thus immersion in aboriginal culture was implied. However, despite differentiating the N P and outpost nursing roles, the tendency to identify the outpost nurse as a N P has persisted. Martin-Misener (2000) stated, . . . the federal and territorial governments have resisted calling nurses who practice in remote nursing stations or health centers primary health care nurse practitioners.... However, the nursing community at large perceives the role of nurses who practice in remote nursing stations to be a nurse practitioner role, as do the nurses themselves, (p. 15) For example, in a study of clinical outcomes in a remote aboriginal community, Knopp (2001) identified the outpost nurses in her sample as NPs , despite the N P title not being recognized at that time by the local provincial nursing regulatory authority. The C N A Position Statement on the Nurse Practitioner (2003) also recognizes N P practice in nursing outposts: "NPs practice in a variety of community, acute care and long-term care settings. These include... nursing outposts..." (p. 1). In my previous research with outpost nurses, study findings suggested that the experienced outpost nurses who participated in my study shared domains and competencies of practice with American NPs (Tarlier et al., 2003). I adapted the framework of domains and competencies of N P practice developed by Brykcznski (1989) to reflect additional competencies specific to outpost nursing practice that had emerged from my research data. 1 7 All were long-time faculty with the Dalhousie Outpost and Community Health Nursing program, Dalhousie University. Ruth May originally established the program at Dalhousie University in 1967 (Chaytor, 1994) and continued to teach in the program following her official retirement in the mid 1990's. 55 The resulting framework of outpost nursing competencies reflected the core knowledge and skills required to support the complex, multi-dimensional work of outpost nurses in both primary care and community health (Tarlier, 2001, p. 104-5; Tarlier et al., 2003). Brykcznski 's (1989) original framework of domains and competencies of N P practice formed the basis of the National Organization of Nurse Practitioner Faculties (NONPF) Nurse Practitioner Primary Care Competencies, the most influential set of N P competencies in the U .S . (Price et al., 1992). These "core competencies are a gold standard" (NONPF, 2002, p. 5) for nurse practitioner education, practice, and credentialing, both in the U.S . and internationally. For example, the N O N P F N P competencies were used to guide curriculum development o f the Family Nurse Practitioner Master's of Science in Nursing graduate degree program at the University of British Columbia ( U B C ) ( U B C School of Nursing, May, 2003). Thus, the finding that experienced outpost nurses share domains and competencies of practice with NPs , as explicated by the N O N P F N P competencies, supports the assertion found in the literature that the outpost nursing role is similar to the N P role. Outpost nurses employed by F N I H B are not required to meet the N O N P F N P competencies. Instead, F N I H B has developed their own clinical competencies as well as clinical practice guidelines for nursing practice in remote aboriginal communities. F N I H B recently revised and updated the guidelines (Health Canada, 2000c, 2001a) and also added a competency self-assessment tool for nurses (Health Canada, 2001c, 200Id). The F N I H B clinical guidelines and the competency self-assessment tool are based on the original M S B scope of practice document (Health Canada, 1994) that "establishe[d] the basic parameters for nursing practice" (p. 1) in Nursing Stations. The scope of practice document explicated specific "duties", responsibilities, skills and tasks expected for basic Nursing Station practice, and provided recognition that many of these responsibilities and skills were beyond the 56 accepted scope of nursing practice in other settings. Thus, the document supported in a "legal" sense the practice of M S B (and later FNIHB)-employed nurses beyond what provincial nursing regulatory bodies supported. For example, the advanced and expanded scope of practice of nurses working in Nursing Station settings included conducting a history and physical examination of a patient " in order to establish a working diagnosis" (p. 8), "initiating... medical intervention... [and] drug therapy... according to predetermined protocols" (p. 9), and independently carrying out procedures such as suturing, venipuncture and diagnostic testing (e.g., blood specimens, radiographic studies). The "predetermined protocols" referred to are the Clinical Practice Guidelines. Nurses' prescriptive and dispensing authority was further guided by the Nurses' Drug Classification System, a drug formulary specific to nurses' practice in federally operated Nursing Stations. 1 8 The scope of practice document also explicated specific public / community health nursing responsibilities, such as communicable disease control, maternal-child and reproductive health. Notably, the competencies and guidelines developed by F N I H B continue to represent the accepted standard of practice (informally in practice, i f not also formally acknowledged) in outpost settings in the Yukon, the Northwest Territories, Nunavut, and transferred health authorities as well as those that continue to operate under direct F N I H B authority. The F N I H B competencies are not identified as or compared to N P competencies; as Martin-Misener (2000) observed, the federal and (until recently) territorial governments have resisted identifying outpost nurses as NPs. However, with the development o f a national N P competency framework ( C N A , 2005a) and implementation in November 2005 of the 18 The Nurses' Drug Classification System has been integrated into the latest version of the FNIHB Clinical Practice Guidelines (Health Canada, 2000c). The FNIHB drug formulary is now Pharmacy Standards of Practice for First Nations and Inuit Health Branch Health Facilities (Health Canada, 2001b). 57 Canadian Nurse Practitioner Examination [CNPE] ( C N A , 2005b), a mechanism now exists for assessing and recognizing outpost nurses who are able to demonstrate N P competencies through the examination process. The use of competencies and clinical practice guidelines is increasingly promoted in nursing and medical practice generally, in the expectation that their use contributes to better continuity of care (Biem et al., 2003) and patient outcomes. However, problems with implementing guidelines have been identified. Biem et al. cautioned that guidelines may be "o f varying quality" (p. 5), and also noted that guidelines are not sufficient when health providers lack other resources, as is often the case in rural and remote settings. Similarly, Vukic and Keddy (2002) remarked that These guidelines can be significantly problematic as the assumption, when referring to these guidelines, is that the nurse has identified the appropriate 'medical' diagnosis when treating patients. These guidelines can be the major source for prescribed practice i f nurses are not adequately prepared, (p. 546) Thus, Vukic and Keddy implied that an adequate base of core knowledge is required to safely implement practice guidelines. Despite the availability of clinical practice guidelines, the outpost nurses in Vuk ic ' s (1997) study found that the level of educational preparation most often provided to outpost nurses was inadequate (i.e., short courses of three to six months' length). The lack of adequate educational preparation to support safe outpost nursing practice was a major theme identified by Vukic and Keddy (2002) in the outpost nursing literature. Based on Vukic ' s research, these authors concluded, "nurses want tools, knowledge, skills, and organizational support to reflect the broad base of knowledge needed to carry out activities in the areas identified as outpost nursing work" (p. 547). Romanow (2002) also noted the need to adequately prepare health providers to better meet the health needs of 58 aboriginal communities. But significantly, the need for nurses to have additional educational preparation for the work they were expected to do in remote communities was recognized as long ago as 1964, more than 40years ago. The Dalhousie Outpost and Community Health Nursing Program that operated from 1967 - 1997, for the specific purpose of preparing nurses for northern and remote practice in aboriginal communities across Canada, was established in response to the Hal l Royal Commission on Health Sei~vices (Hall , 1964, cited in Martin-Misener et al., 1999). Despite early recognition that nurses providing primary care services in remote communities needed additional educational preparation, and the establishment of a handful of programs, such as the Dalhousie program, to provide such education, a 2001 paper by Silverman, Goodine, Ladouceur, and Quinn suggested that nurses working in remote aboriginal communities perceived a need for "continuing education" related to basic primary care clinical competencies, such as physical assessment and perinatal health. The Silverman et al. paper described a survey of FNIHB-employed nurses (n = 369) wherein nurses were asked to identify their continuing education needs. Employer's expectations of nurses' practice competencies were not discussed (e.g., as delineated by the F N I H B Competency Assessment Tools, Health Canada 2001c, 200Id). Survey respondents also identified emergency, acute care, gynecological and obstetrical skills as "cri t ical" (p. 43) areas for continuing education. Interestingly, while respondents identified "victims of violence, non-compliant clients, substance abuse, fetal alcohol syndrome" as areas where continuing education was needed, there was no indication that nurses or the researchers related issues such as these to the socio-political context of aboriginal health, nor did either group identify knowledge related to aboriginal health or cultural issues, community health or community development skills as areas for continuing education. 59 What constitutes adequate educational preparation for outpost nurses? While the type of educational preparation adequate to prepare nurses for a N P role has long been a point of controversy in Canada (Tarlier, 2001), there is little evidence in the literature of discussion regarding adequate education to prepare nurses for outpost nursing practice. Most research involving outpost nurses has confirmed that the majority of outpost nurses are educated at the diploma or baccalaureate level, and may or may not have additional preparation in the form of a short (three to six month) course in primary care skills (Knopp, 2001; Martin, 1997; Silverman et al., 2001; Stewart & MacLeod, 2005; Tarlier et al., 2003; Vukic & Keddy, 2002). This has been in contrast with the trend in the U.S . towards graduate level education for NPs. Moreover, while N P education in Canada remains fragmented across the various jurisdictions that currently recognize and regulate N P practice, at the national level the trend is towards identifying NPs as advanced practice nurses requiring graduate level education ( C N A , 2003). From a historical and socio-political perspective, development of the N P role in Canada may be viewed as being an evolutionary process. Notably, this process in Canada follows the pattern of N P role development over the past 30 years in the U . S . (Price et al., 1992; Ray & Hardin, 1995), and is also similar to contemporary events in the U . K . (Carnwell & Daly, 2003). For the foreseeable future, NPs in Canada wi l l likely continue to be recognized with a variety o f combinations of education and experience through which they have developed the requisite N P core knowledge and competencies ( C N A , 2003). Thus, in considering outpost nursing competencies, it may be more productive to look at core knowledge and competencies required for the role rather than at educational preparation per se. This is consistent with Minore et al.'s (2002) call for health services research to evaluate the core knowledge required by health providers in remote aboriginal communities. 60 There is general agreement among Canadian nurses who see themselves as practicing at an advanced practice level in remote settings that some preparation beyond basic nursing education (i.e., at the diploma or baccalaureate level) is required to support advanced nursing practice (Kulig, 2005; Schreiber et al., 2003). A s more evaluative research of nursing practice is carried out, there is also increasing recognition that nurses with a higher level of educational preparation may influence more positive health outcomes (Aiken, Clarke, Cheung, Sloane, & Silber, 2003). Scholars in the area of remote and outpost nursing practice have identified that the preparation of outpost nurses is inadequate (Minore et al., 2002, 2005; Vukic & Keddy, 2002). For example, Stewart and MacLeod (2005) found that only 40% of nurses who identified themselves as practicing in N P positions in rural and remote settings had advanced practice nursing preparation. There is a need for evaluative studies of health services in remote aboriginal communities to substantiate the need for additional educational preparation for nurses and to suggest what core knowledge is required by nurses and other health providers (Minore et al., 2002; Tarlier et al., 2003). Literature Relevant to the Evaluation of Primary Care A guiding principle o f health services delivery is that health strategies should be focused on health outcomes (Romanow, 2002). This principle is equally applicable whether health services are delivered in remote aboriginal communities or to mainstream urban populations. However, there is an alarming deficit of clinically based health outcomes research in remote aboriginal communities. Only two studies were identified that addressed the delivery of primary care in remote aboriginal communities from a broad clinical perspective. One of these studies was the case study of continuity of care conducted by Minore et al. (2002, 2005). The other study compared primary care practice patterns o f NPs and physicians in a remote aboriginal community (Knopp, 2001). Since there is so little 61 literature related to primary care services delivered by outpost nurses, and none that addresses patient clinical health outcomes per se, a brief overview of the body of research pertaining to the outcomes of primary care N P practice is offered, on the premise that outpost nursing has been identified as a nurse practitioner type role. A s well , a summary o f the literature pertaining to the Burlington Randomized Controlled Trial ( B R C T ) of nurse practitioners (Spitzer et al., 1974) is provided, as both the original study and subsequent studies that further tested the methods and tools of the B R C T are of particular relevance to this research. The Burlington Randomized Trial: Appraising Quality of Primary Care The Burlington Randomized Trial o f the Nurse Practitioner (Spitzer et al., 1974), one of the earliest large clinical trials of N P practice, was a Canadian study that remains a landmark in the contemporary N P literature. The B R C T randomized 1,598 families (n = 1,598) to physician (conventional; n = 1,058) or N P (experimental; n = 540) practice. The physicians and NPs worked in a collaborative practice. Families were chosen as the unit of randomization "because many clinical problems in primary care involve an entire family" (p. 252). Patients were followed for one year. Pre and post trial data were collected on patient health status, clinician activities and practice activities. Measured outcomes included mortality, and clinical effectiveness and safety, evaluated using criteria developed to reference selected specific indicator conditions. The B R C T concluded that NPs were capable of providing "first-contact primary clinical care as safely and effectively, with as much satisfaction to patients, as a family physician" (p. 255). The indicator conditions and chart data abstraction tools developed for use in the B R C T formed the basis of the data abstraction process that was used in the present study. Indicator conditions refer to health states, diseases or a constellation of presenting symptoms, 62 and are selected on the basis that (a) management or treatment of the condition may influence the outcome, and (b) the condition occurs frequently enough in primary care settings to provide adequate data for analysis (Sibley et al., 1975). The development of indicator conditions was an attempt to adapt the tracer disease concept (Kessner, Kalk , & Singer, 1973, as cited in Sibley et al.) to quality of care evaluation in primary care practice. Ten indicator conditions were identified for the original B R C T study; an additional 11 conditions were identified in subsequent similar studies by the same group of researchers (Sibley, 1976). Explicit criteria for the management of each condition were developed based on the available empirical evidence and through a process of consultation with a Peer Advisory Group. This group was composed of three family physicians in clinical practice rather than academia, to ensure that criteria reflected community standards of adequacy in primary care rather than academic standards, in keeping with the stated purpose of the B R C T . Physicians in the Peer Advisory Group were highly regarded by both their clinical and academic peers (Sibley et al., 1975). Criteria for each indicator condition included drug management, referral decisions and monitoring and follow-up of patients. Evaluation of the referral process was included "as a supplementary measure of quality" of care (p. 46). Chart data were abstracted on the basis of episodes of care, which Sibley et al. (1975) defined as "al l encounters for the management of an indicator condition" (p. 48). For example, an episode of otitis media might include several patient-provider encounters between initial diagnosis of the condition and its eventual resolution. There was an underlying assumption that a single practitioner was primarily responsible for the management of an episode of care. Both the first and last encounters of an episode had to fall within the time frame of the defined study period for an episode to be eligible for inclusion in the study. Criteria defining an episode of care specific to each indicator condition were 63 developed (e.g., to be eligible for inclusion in the sample of the indicator condition "Care of the Newborn", a child's first visit for well-child care at the clinic had to occur within the study period). Initial pilot testing of the data abstraction tools developed for the B R C T demonstrated high inter-rater agreement on scoring, ranging between 88-94% (Sibley et al., 1975). The researchers also claimed good internal consistency between the abstracted data scores and scores obtained through the three different approaches used (i.e., indicator condition management, drug management and opinion o f consultants). The results were also consistent with the outcome measures of the B R C T , that is, mortality, and physical, social and emotional well-being, as evaluated by various measurement strategies described by Sackett et al. (1974). The original group of researchers subsequently used the indicator conditions and tools developed for the B R C T in further studies of quality of care in primary care practices. The ten indicator conditions developed initially were used to evaluate the quality of care in an additional 15 practices, including practices in rural settings. The 11 new indicator conditions were used to evaluate another 13 primary care practices (Sibley, 1976). The method of quality of care appraisal developed for the B R C T was further used in two primary care studies conducted in B . C . , which are presented here briefly. Robertson (1984) used seven o f the original B R C T indicator conditions to evaluate the quality of care at an urban Vancouver primary care clinic that similarly to the B R C T study sites, employed NP/physician teams. Notably, this master's thesis research did not compare the practice of NPs and physicians but appraised overall adequacy of care by all providers. The indicator conditions selected for use in the Robertson study were based on a list of conditions most frequently seen at the study site (i.e., otitis media, hypertension, 64 prenatal care, depression, urinary tract infections, childhood immunizations, and care o f the newborn in the first year). The clinic's medical director reviewed criteria for each of the selected indicator conditions to ensure they reflected current accepted practice standards (Sheps & Robertson, 1984). Robertson conducted a pilot study of 32 charts to assess reliability of the data abstraction process and found an overall agreement o f 81%. A second reliability study demonstrated overall agreement of 85%. A retrospective chart review over the one-year period prior to the study was carried out. The unit of analysis was defined as an episode o f care (n = 103). Results were compared with the results of the B R C T and no significant differences were found in adequacy of care. Identified limitations of the method included (a) indicator conditions allowed only a "sample of care given" (Sheps & Robertson, p. 885), thus limiting the generalizability of results to other health conditions, (b) the criteria focused on technical aspects of care and neglected psychosocial aspects of care, and (c) exclusive reliance on chart data, which may be an incomplete source. Robertson also observed that i f the relationship between quality of care and outcomes were better established, it would be possible to use the B R C T criteria to assess quality of care in the absence o f data regarding outcomes. Overall however, Sheps and Robertson found the B R C T indicator conditions and data abstraction tools to be a useful method to assess the adequacy o f care in a primary care setting, and stated the criteria were "practical to use, clinically relevant and unambiguous" (p. 886). The B R C T indicator conditions criteria and data abstraction tools were also used for an evaluation of prenatal care provided to low-risk women by nurse midwives (n = 44) compared to family physicians (n = 88) (Buhler, Gl ick, & Sheps, 1988). Using a peer advisory process, the criteria and data abstraction tools were updated and revised to reflect the current standards of prenatal care. Interrater reliability in this study was 95%, intrarater 65 reliability was 96%, and overall agreement in chart scoring was 95%. Overall, the B R C T method of using indicator conditions and chart data abstraction tools has proved practical and adaptable to assessing the quality of care in a variety of primary care settings, and has demonstrated acceptable validity and reliability. Appraising Nurse Practitioner Practice A vast amount of research has accumulated over the 30 years since the B R C T to further substantiate the safety, efficacy and effectiveness o f nurse practitioners in primary care. In 1992, Dr. Loretta Ford, a co-founder o f the first N P program in the U.S. , observed, "No professional role has been more thoroughly discussed, described, studied and reported than that of the nurse practitioner" (Ford, 1992, p. 289). A s Price et al. (1992) noted, "the literature is now fairly replete" (p. 11) in terms of describing the N P role from a variety of perspectives, including competency and the processes o f care delivery by NPs, cost effectiveness, and patient satisfaction. The wealth of research addressing the N P role has been critically synthesized and summarized in three key projects, which are briefly presented here. Feldman, Ventura, and Crosby (1987) provided a systematic review of the early N P literature. From an initial reading of nearly 1,000 documents, 350 were found to be relevant to evaluating N P effectiveness. Effectiveness was defined as N P performance and / or outcomes of practice (p. 303). Studies were evaluated on the basis o f scientific merit; that is, on study design, validity, reliability, generalizability and statistical analysis. Studies were excluded from the review " i f NPs could not be isolated from other health professionals in terms of outcomes evaluated" (p. 303). Through a systematic process of critical analysis by a group of " N P expert" nurse scholars, it was determined that 56 studies met pre-established criteria in terms of being average or above average in regard to (a) relevance, (b) clarity and 66 (c) lack of methodological flaws. The results of the review were presented in tabular format and included several large clinical trials of N P outcomes, all of which had demonstrated equivalent or better outcomes of N P managed care compared to physician managed care. Feldman et al. also determined that several large quasi-experimental studies met the stringent criteria o f their review process, as well as a handful of descriptive studies. While there were some differences between study findings, an overall high degree o f consistency was demonstrated supporting the effectiveness, safety and acceptability of N P practice. Notably, virtually all o f the studies included in this review compared N P practice to physician practice. Several years following the Feldman et al. (1987) review, Brown and Grimes (1995) conducted a meta-analysis evaluating patient outcomes of NPs and nurse midwives. The basis of comparison was again physician practice. Thirty-eight studies o f N P practice were included in the meta-analysis. Practice was evaluated on the basis of 33 specified outcomes. Brown and Grimes found that while outcomes of N P and physician practice were equivalent on most variables in controlled studies, in studies that were randomized to provider, N P -managed patients demonstrated better patient compliance with treatment recommendations. Moreover, " in studies that controlled for patient risk in ways other than randomization, patient satisfaction and resolution o f pathological conditions were greater for N P patients" (p. 332). O f relevance to the present research, Brown and Grimes noted that a limitation in the data of the studies they reviewed "precluded answering questions of why and under what conditions these outcomes apply" (p. 332). Thus, there is an implied need for research that considers outcomes of practice within a contextualized understanding that encompasses structure and process. 67 More recently, Horrocks, Anderson, and Salisbury (2002) undertook a contemporary systematic review of randomized controlled trials and prospective observational studies of N P practice compared to physician practice. This review included many of the studies included in the earlier reviews as well as several contemporary clinical trials (e.g., Kinnersley et al., 2000; Mundinger et al., 2000; Shum et al., 2000). Studies were excluded i f data on patient outcomes were not provided. Outcomes included patient satisfaction, health status, process of care and health costs. Based on criteria that were developed by the Cochrane Effective Practice and Organisation of Care Review Group, a mere 11 randomized trials and 23 observational studies were included in the Horrocks et al. review. 1 9 N o significant differences were found between physicians and NPs in terms of health outcomes; thus, conclusions drawn from this review were consistent with previous reviews: "Nurse practitioners seemed to provide a quality of care that is at least as good, and in some ways better, than doctors" (p. 821). The literature addressing the evaluation o f N P practice offers three important insights for the proposed research. First, as noted previously, physician practice has been the "gold standard" against which N P practice has historically been judged, both in the B R C T and in subsequent research. At the time the B R C T was conducted, physician practice in primary care would have offered the only standard, albeit a normative one, against which to evaluate N P practice. However, the advent of evidence-based medicine in 1992 (Evidence-Based Medicine Working Group, 1992) has led to the development of evidence-based standards of practice, thus providing evidence-based anchors against which to evaluate the quality of practice. Second, a point not addressed in the literature reviewed here, the narrow focus on 1 9 Notably, the BRCT withstood the test of time and was included in the small group of 11 randomized control trials that met the Cochrane criteria for inclusion in the Horrocks et al. (2002) review. 68 the part of N P practice that overlaps with what has been traditionally considered physician practice, has been at the expense of better understanding the part of N P practice that remains grounded in nursing and that NPs purportedly enact as advanced practice nursing. Nursing is now also moving towards developing an evidence-base for practice, implying the possibility that advanced nursing practice may also eventually be evaluated against evidence-based anchors. Third, as Brown and Grimes (1995) observed, research is needed to develop a better understanding of how and why N P practice influences patient outcomes. Appraising Outpost Nursing Practice In contrast to the plethora of research that has established NPs as safe and effective primary care providers, only two studies were located that specifically addressed patient health outcomes in remote aboriginal communities, where most primary care is provided by outpost nurses. The first study, which was specific to nurse-managed primary care in a remote aboriginal community, was a master's thesis by Knopp (2001). Knopp conducted a comparative retrospective chart review (n = 985) over a six-month period in one remote aboriginal community in Saskatchewan. The variable of interest was the management of ambulatory infectious disease in children aged one to five years (e.g., upper respiratory tract infections, skin infections and otitis media were the three most common diagnoses). The care of patients managed by NPs (n = 574) was compared to that of patients managed by primary care physicians (n = 411). Ten NPs and three primary care physicians practiced in the health center during the six-month time frame of the chart audit. Practice patterns were compared on six parameters: (a) population of children seen by health care provider, (b) diagnosis of children seen by provider, (c) diagnostic tests ordered, (d) prescribed treatment (primarily antibiotics prescribed), (e) follow-up suggested, and (f) adherence to practice standards. 69 Knopp (2001) identified the outpost nurses in her study as NPs , based on the nurses' having "special advanced training in primary care" (p. 26). The literature review provided by Knopp focused on situating the N P as an advanced practice nurse ( A P N ) prepared at the graduate level and having significant knowledge and skills beyond basic nursing education. However, of the outpost nurses in Knopp's study, nine held diplomas in nursing and one held a baccalaureate degree in nursing. The meaning of "special advanced training in primary care" was not made explicit. Thus, it is unclear how or whether the outpost nurses in Knopp's study met her definition o f N P as an advanced practice nurse, or by what criteria it was determined that these nurses were NPs , other than that their role included primary care. Despite this lack of clarity, it appeared clear that the educational preparation of the nurses in Knopp's study was similar to that of most outpost nurses. Chart data were abstracted using a data abstraction tool developed and pre-tested by Knopp (2001) and reviewed by "experts" for face validity (personal communication, J. Knopp, March 2, 2004). Overall, few significant differences were found between the practice patterns of the NPs and physicians in the study (Knopp). The main significant difference noted by the researcher was that while rates of antibiotic prescription were equivalent between N P and physician providers, the class of antibiotic differed by provider, with NPs tending to prescribe more penicillin based drugs, and physicians prescribing more cephalosporins and sulphonamides. There is no evidence that data related to non-pharmacological interventions were collected. NPs also documented requested follow-up with patients more frequently than did physicians. The usefulness of Knopp's (2001) study is limited by several important considerations. For example, the study relied exclusively on patient chart data. Knopp noted problems such as inconsistency and incompleteness that are associated with this strategy, and 70 which have been noted by other researchers as well (Buhler et al., 1988; Gilbert, Lowenstein, Koz io l -McLa in , Barta, & Steiner, 1996; Robertson & Sheps, 1984). However, inconsistent, incomplete or unclear chart information has important implications for continuity of patient care. In particular, informational continuity may be compromised in settings where patients are seen by different health providers or where there is a high turnover rate of health care providers. Knopp remarked on the difficulty of acquiring data from other sources (i.e. databases) in a remote health center setting. She also noted that N P s ' chart documentation tended to provide more complete data than physician's chart documentation. Knopp (2001) observed that structural factors might have influenced N P ' s and physician's practice. For example, she stated, "When describing and comparing the treatment prescribed by health care providers in Pelican Narrows, it is important to first understand the structure of the cl inic" (p. 55). However, data regarding key structural factors was not presented in a systematic manner as part of the data collection and analysis. Confidence in study findings would be strengthened i f possible confounding and mediating structural factors had been considered at the outset of the study. Evaluating practitioner adherence to practice standards (using C M A practice guidelines) was found to be problematic due to the lack of guidelines available for several of the clinical diagnoses, and the finding that the physicians in the study did not use standard guidelines. Thus, while practice patterns were found overall to be comparable between the NPs and physicians, it remains unknown whether the practice of either group was of an acceptable standard. It is also unclear how care may have been shared between providers, such as might be the case i f a N P consulted with a physician. Overall, Knopp's (2001) study was significant in that, despite identified conceptual and methodological concerns, it appeared to be the only existing evaluative study of primary care delivery by outpost nurses. 71 The second study that addressed patient health outcomes in remote aboriginal communities was the study by Minore et al. (2002, 2005), discussed earlier in this dissertation from the perspective of its relevance to conceptualizing continuity of care. While this was not a study of nurse-managed primary care per se, because outpost nurses were "the principle primary care providers" (p. 13) in the remote communities where the study took place, it was ipso facto a study of nurse-managed care. The Minore et al. (2002) study was commissioned by F N I H B (Ontario Region) and the Shibogama First Nations Council in response to concerns about continuity of care that were raised by a physician who had worked as a locum in the Shibogama First Nations communities. The multi-site case study, conducted in three Shibogama First Nations communities, used both quantitative and qualitative strategies to explore the adequacy, timeliness and cost-related factors of health services in the communities. One key research question addressed the relationship between lack of continuity of care and client health outcomes. Donabedian's structure-process-outcome framework provided a conceptual framework for the research. The study was a retrospective chart review that extracted chart data on three selected health conditions (i.e., diabetes, oncology and mental health) over a five-year period. These conditions were selected on the basis that "within the domain of each of the selected pathologies, adverse clinical outcomes among patients underscore the importance of continuity" (p. 3). Chart data were collected using data extraction tools developed specifically for this study (Minore et al., 2002). The tools were based on the F N I H B Clinical Practice Guidelines (Health Canada, 2000c), which provided the standard of practice for nursing. These tools .. .were designed to give a holistic assessment of the health system's response to clients, rather than measure the performance of individual care providers, which is conventionally the outcome of chart audits. The tools also had to capture information 72 along several dimensions: time sequence and actions taken by provider category in assessing, diagnosing, developing/implementing a treatment plan and evaluation. (Minore et al., p. 7) Data were also collected through in-depth interviews with 30 individuals and interpreted in accordance with qualitative data analysis strategies. Specific data on health outcomes were not presented in the Minore et al. (2002) study report. However, the discussion o f findings suggested that poor outcomes were related to discontinuity of care. Major factors contributing to the lack of continuity were inadequate preparation and education of health providers (referring to non-nurse providers such as physicians and mental health workers as well as nurses), the failure of primary care providers to do "holistic assessments" of patients (p. 21) (i.e., process of care), lack of follow-up care, and failure to communicate patient information between providers (i.e., due to structural factors such as chart organization and appointment-booking systems). Continuity of care was also viewed as being related to the larger system of health services organization and what might be characterized as micro-meso-macro system interplay. The study by Minore et al. (2002) offered two key implications for the present research. One, it demonstrated the existence of problems in each component of the structure-process-outcome model and its value to research in remote aboriginal communities, allowing evaluation of outcomes as a reflection of a system of care and not merely as the result of a specific regime of care or an individual provider or patient. Thus, it verified the conceptual validity of exploring continuity o f care as a link between the process and outcomes of health care. A n d second, it demonstrated the utility of using multiple research strategies, or combining qualitative and quantitative methods, to achieve an in-depth understanding of "how things work" in the unique context of aboriginal health services in remote settings. 73 Summary This substantive critical synopsis of the literature has attempted to thread together extensive bodies of literature from four separate, albeit related, fields of knowledge: (a) rural and remote health research, (b) the historical and socio-political context of aboriginal health and health services in Canada, (c) outpost nursing, and (d) the evaluation of primary care practice. It was not each distinct body o f knowledge in itself that was most relevant to the present research, but the discrete ways in which each body of knowledge interfaced with the others. That is, there was a particular area of overlap between each of these four bodies of knowledge that I endeavored to extract, critically examine and present in a way that engendered clearer understanding of the unique, complex and multi-dimensional issue of health and health services in remote aboriginal communities. The conceptual framework of the Nursing Role Effectiveness Model , adapted to fit the specific context of this research, offered a way to organize this information that clarified the inter-relationships suggested by the literature (see Appendix B) . One startling observation about the literature was that it was largely based on graduate student research. Another observation was that so few scholars are actively involved in researching health services in remote aboriginal communities, despite broad recognition that aboriginal people and residents of northern and remote regions experience the poorest health status of all Canadians. Data related to health outcomes in remote communities are largely unavailable or difficult to access. Very little research exists linking the structure and processes of health services to health outcomes in remote communities. In particular, there is little research describing how the process of nursing care may influence health and health outcomes. While previous research suggests that organizational and structural issues affect nurses' practice, 74 there is little information about how these issues influence either the process of nursing care or health outcomes. There is limited knowledge regarding the core knowledge and competencies that nurses need to work in remote aboriginal communities, about how nurses acquire needed competencies, or what education and experiential backgrounds might best support nursing practice in the Nursing Station setting. There is even less knowledge about how nurses enact necessary knowledge and competencies, or about how the three dimensions of continuity of care identified by Reid et al. (2002) are demonstrated in the process of nursing care. The present research aimed to address these gaps in current knowledge by providing insights and increasing our understanding of the complex inter-relationships that exist between the context and structure of health services delivery, the process of nurses' work and continuity of care, and health outcomes in remote aboriginal communities. 75 CHAPTER FOUR RESEARCH DESIGN AND IMPLEMENTATION Introduction The phenomenon of continuity of care and how it is influenced by multiple contextual, structural and process factors represents a complex research problem. The unique nature o f the organization of health services in remote communities adds another distinct layer of contextual complexity. There are few evaluative studies of nursing practice in remote communities and little relevant literature informing us of the inter-relationships between the context, organization and structure of health services delivery, the process of nurses' work and health outcomes. Even data relevant to health services research is challenging to access; few ( i f any) remote health facilities have yet implemented an effective computer database system for maintaining patient records. Patient data from these communities is also notably missing from the larger national and regional level administrative databases (e.g., C C H S ) . Thus, patient data related to health outcomes remains relatively inaccessible. A l l of these issues contributed to the blurred boundaries that were evident between the research phenomena and its context; they created a research environment that was both challenging and problematic. Conventional health outcomes research designs lacked the scope and flexibility required to shift focus between the broad gaze necessary to integrate understandings of the broader context, or the macro-level of health services delivery, the meso and micro-levels of the processes of care delivery (i.e., the processes of nursing practice) and the micro-level of individual patient health outcomes (Sidani et al., 2004). The limitations of conventional research to answering research questions about complex phenomena in health care are becoming more widely recognized: "The search for adequate tools to monitor and evaluate the quality of nursing care is not new, but it has been hampered 76 by narrow conceptual models and inadequate research tools" (Sidani et al., p. 60). Given the complex, multi-factorial and contextualized nature of outpost nursing practice, these issues were highly relevant to the present study. The objectives that directed the present research reflected this degree of complexity and demanded a method o f inquiry that facilitated an understanding of the phenomena of interest, as it exists within a complex system. A n ethnographic method offered the necessary scope and flexibility to achieve the objectives of this study. Method of Inquiry: Ethnography While different forms of ethnographies exist (e.g., institutional ethnographies, critical ethnographies, medical ethnographies, and ethnographies o f work or educational settings [Atkinson, Coffey, Delamont, Lofland, & Lofland, 2001]), they are each underpinned by common guiding concepts. These conceptual underpinnings include "culture, a holistic perspective, contextualization, emic perspective and multiple realities, etic perspective, non-judgemental orientation, inter-and intra-cultural diversity, structure and function, symbol and ritual, micro and macro, and operationalism" (Fetterman, 1998, p. 475). Thus, ethnography is conceptually consistent with the guiding concepts o f this research, presented in Chapter Two. Exploring health and nursing practice within the broader contexts of a remote First Nations community, and aboriginal health and health services in Canada, necessarily implied not only the concept of culture, but also the intersection of multiple co-existing cultures. For example, intersecting cultures included the self-identified culture of the First Nations community, the "mainstream" or dominant Canadian culture of most o f the non-First Nations participants, the culture of l iving in an isolated northern community (Tarlier et al., 2003), the organizational culture within the Nursing Station, as well as the organizational culture of F N I H B . The concept of cultures in the context of the study site also implied the enculturated 77 social conditions of inequity, poverty, and dependence that are the result of the colonial roots and socio-political history of aboriginal peoples in Canada. A broad conceptualization of culture as complex, dynamic, relational, and shaped by political, social, and historical influences (Browne & Varcoe, in press; Reimer Kirkham et al., 2002) underpinned the understanding of culture in this study and supported the assumption o f multiple co-existing cultures. Kirmayer, Simpson, and Cargo (2003) defined a similarly broad view of culture: While older anthropological writing conceived o f cultures as closed, homogenous and sometimes static systems, contemporary ethnographers view cultures as local worlds that are constantly in flux. There is great variation in knowledge, practice and attitudes among individuals within a cultural group, with significant conflict, resistance and contestation of dominant values. Local worlds are embedded in larger global systems that bring diverse peoples together through migration, mass media and other forms of contact and exchange. A s a result, most individuals have access to and participate in multiple cultures. Individuals use this multicultural background to navigate, communicate and provide rhetorical supplies and discourses within which to locate and construct socially and psychologically viable selves, (p. s i9) It is important to acknowledge that while some contemporary nursing scholars are applying a critical theoretical lens and advocating a broader conceptualization of culture (e.g., A l l en , 1999; Browne & Varcoe, in press; Doane & Varcoe, 2005; Reimer Kirkham et al., 2002), within nursing practice generally, culture continues to be defined primarily in narrow culturist terms. Such narrow views of culture have led to the prevalence within nursing of understanding culture as static, and chiefly associated with specific traits that are readily identifiable with a particular culture (e.g., specific customs or beliefs) (Reimer Kirkham et al.). Narrow understandings of culture typically serve to reinforce aspects of difference between the dominant culture and another culture, thereby also reinforcing the type of pervasive negative stereotypes about non-dominant cultures that exist in wider social discourses (Browne & Varcoe). Recognizing the tensions between a broad versus a narrow 78 conceptualization of culture was critical to analyzing the data in this study related to participants' discussions of culture. To support the identified objectives of this study, a research method was conceptualized wherein qualitative and quantitative data were equally privileged, each perspective informing the other. While ethnography has come to be considered primarily a qualitative research method (Morse & Field, 1995), ethnographers have historically combined qualitative and quantitative methods of data collection and analysis to foster a complete and comprehensive understanding of the research phenomena (Fetterman, 1998). Combining or integrating quantitative and qualitative strategies under the umbrella of ethnographic methodology has gained contemporary recognition also: In terms of methodological innovativeness, the most interesting examples of ethnographic inquiry that include qualitative and quantitative approaches, as well as varied levels of analysis, are those in which the ethnographer plays an active i f not leading role in research design and is thereby able to maintain a primary interest in cultural analysis (Chambers, 2000, p. 859). Therefore, ethnography was appropriate to the mixed methods design of this study. The present ethnographic research was further framed as a case study to clarify the boundaries that were placed around the phenomena of interest. I have previously made the point that an understanding of the broader context of health services delivery in remote aboriginal communities is critical to conducting meaningful research in this setting. However, neither the context itself nor the overall organization o f health services in aboriginal communities was the primary focus of this research. Having acknowledged these as being inextricably linked to the delivery of health services in aboriginal communities in general, the focus of this research shifted to the delivery of health services in one particular First Nations community. In identifying a case, Stake (2000) observed, 79 It is common to recognize that certain features are within the system, within the boundaries of the case, and other features outside. Some are significant as context... boundedness and behavior patterns are useful concepts for specifying the case (p. 436). Thus, the boundaries of this research were set around the delivery of health services in one study site, while acknowledging the influence of the larger context. The Study Site This research was conducted in the Nursing Station of a remote First Nations community in northern Canada. The study site was purposefully selected on the basis of four specific criteria: (a) primary care was provided chiefly by nurses / nurse practitioners working in an expanded and advanced nursing role and within a primary health care framework, (b) the community population was greater than 1,000, (c) the Nursing Station employed a minimum of four nurses, and (d) the community was isolated both geographically and in terms of access to health care providers other than nurses (i.e., no doctor was resident in the community). These selection criteria are typical of most remote aboriginal communities where outpost nurses practice; therefore, the study site may be considered typical of Nursing Stations in most mid-sized or larger remote aboriginal communities in northern Canada. Additionally, requiring that the study site Nursing Station be operated under F N I H B or a territorial government authority ensured that nurses were legally covered to work in an expanded and advanced role, and that they were more likely to have access to resources to support expanded and advanced role practice (i.e., a scope of practice document, clinical guidelines, and explicit expected practice competencies). Many remote aboriginal communities in Canada have fewer than 800 residents. A community o f this size would employ only two to three nurses, and the number of health-80 related clinic visits for primary care would be correspondingly small. To meet the objectives of this study, a larger community and a larger Nursing Station offered greater scope for exploring how aspects such as staffing, nursing workload, and communication and referral patterns among providers influenced continuity of care. Additionally, a larger community population size was necessary to provide an adequate number o f patient charts to allow for a statistically meaningful analysis of chart data. Gaining Entry Access to several potential sites that met the pre-established selection criteria was initially negotiated with the Regional Nursing management of F N I H B in Manitoba. Access to a specific site was then negotiated and finalized with the Nurse in Charge (NIC) o f the Nursing Station in that site. F N I H B committed to providing support in kind by supplying me with accommodation in the study site while fieldwork was carried out. 2 0 During a preliminary visit to the site prior to beginning data collection, the N I C provided me with a liaison to the Band Chief and Council, the recognized authority within the local community (CIHR, 2005). The liaison was an experienced Community Health Representative (CHR) , a health worker who was a member of the local community and worked in the Nursing Station, but who was employed by the Band. The C H R arranged for me to present my research plan to the Band Councillors. After considering my research plan the Band Councillors provided me with a letter approving the study and authorizing me to carry out research within the community. During the three-day preliminary visit, I presented my study to the nurses and staff working in the Nursing Station, spent time talking with nurses and staff, responded to 2 0 Accommodation in remote aboriginal communities is often at a premium and very difficult to obtain; thus this represented a significant offer of support. 81 questions about the study, and familiarized myself with the site. This was also an opportunity for the staff and nurses to become familiar with me, and to become accustomed to seeing me "hanging around". I also conducted a pilot test of the data abstraction instruments during this time. This proved to be time well spent, as when I returned a few weeks later to begin data collection, I had already begun to establish a working relationship with the individuals who were key informants and who were instrumental in helping me to gain and maintain access within the study setting. Units of Analysis The main unit of analysis was the Nursing Station. Individual patients and nurses represented embedded units o f analysis within the larger "case". While data from several sources (i.e., observations, interviews, review of patient charts) were collected from individuals, and a preliminary analysis o f the chart review data occurred at the individual level, the data was ultimately interpreted at the Nursing Station level. To illustrate, regarding survey data collected at the individual level, Y i n (2003) stated, "These data may be highly quantitative, focusing on the attitudes and behaviour o f individual clients, and the data w i l l be used along with archival information to interpret the success and operations at the given center" (p. 52-53). Thus, in this research, although data related to individual patient outcomes was collected and initially analysed at the individual level, the real utility of these data was in how individual findings contributed to a better understanding of continuity of care at the organizational level (see Appendix C) . Overview of Data Collection Strategies Data collection was directed and guided by the research objectives and the adaptation of the Nursing Role Effectiveness Model conceptual framework that guided this study. The framework specified structure, process and outcome factors identified in the literature that 82 were thought to influence the practice of outpost nurses (see Appendix B) . Specific questions designed to target the collection o f data addressing each identified factor were explicated in the data collection protocol (see Appendix D). These questions served as "reminders regarding the information that needs to be collected and why" (Yin , 2003, p. 74). Ethnographic data collection strategies were used, with the goal of generating the level of "thick description" that supported an in-depth analysis (Morse & Field, 1995, p. 156). These strategies included (a) conducting interviews with key informants (e.g., nurses, Nursing Station staff and managers); (b) direct observations (e.g., of patterns o f interactions between nurses, patients, other health professional and paraprofessional staff, and relationships within and external to the Nursing Station); (c) focused conversations with key informants, patients and community members that helped to clarify or provide insight into field observations, (d) writing field notes; and (e) collecting documents and physical artefacts (e.g., work schedules, maps, photographs and communication tools such as patient chart forms). Quantitative and narrative data (in the form of contextual notes) were also collected from the review of patient charts, using the chart data abstraction tools. Each strategy is discussed in greater detail below. Data collection took place during two visits to the study site, approximately one month apart. In total, I spent approximately six weeks in the study site. I conducted observations and interviews, collected artefacts, and abstracted chart data throughout these six weeks. The Nursing Station operated regular hours between 8:30 a.m. and 5 p.m., Monday through Friday. Observations took place primarily during regular clinic hours. I was allowed access to the Nursing Station after-hours, and I often worked in the Nursing Station during the evening hours and over the weekends, abstracting data from patient charts. During these times I often had the opportunity to observe nurses' interactions with staff and patients 83 when nurses were called back to the clinic after regular hours. Thus, data collection was continuous during the times I was in the study site. Most of the nurses who agreed to be interviewed expressed the preference to be interviewed during their off-work hours; therefore interviews were most often conducted when the clinic was closed (i.e., during evenings and week-ends). Some nurses and all clinic staff preferred to be interviewed during work hours. The Data Set The data set was comprised of data derived from patient chart review, interviews, observations, artefacts (including photographs), and a personal reflective journal I kept while in the field. In total, this represented a substantial data set. I endeavoured to collect as much of the data as practical directly on a laptop computer, to simplify the task of cataloguing and organizing. Patient Chart Review Quantitative data related to the identified clinical health indicator conditions, health outcomes and continuity of care were abstracted in accordance with the data abstraction tools developed for this study. Two data abstraction tools were developed, one for each of the two indicator conditions: prenatal care and care of the infant through the first year. Chart data abstraction instruments Data abstraction tools were based on health indicator conditions and criteria originally developed for the B R C T (Spitzer et al., 1974), updated and modified to reflect current clinical practice expectations o f nurses providing primary care in F N I H B Nursing Stations (Health Canada, 2000c, 2001a). Data were abstracted from review of patient charts for each indicator condition in accordance with the relevant Data Abstraction Instrument Protocol (see Appendix E) . 84 The decision to base chart data collection on the two conditions of prenatal care and infant care through the first year was based on the expectation (confirmed during pilot testing) that these two groups of patients would provide an adequately large sample for the purpose of this study. Each indicator condition was managed primarily by nurses in the Nursing Station setting, yet potentially offered referral to a physician or other health care provider as an appropriate intervention, thereby lending itself to exploring informational and managerial aspects of continuity of care. To increase the utility of the data abstraction tools to exploring the phenomena of continuity of care in remote aboriginal communities, the B R C T tools (Sibley, 1976) were further adapted to a format that was similar to the format developed by Minore et al. (2002). Minore et al. developed tools for abstracting chart data on three health conditions: diabetes, mental health and oncology. These tools were developed to capture chart data along the dimension of time, which is key to understanding continuity. The tools were based on the F N I H B clinical guidelines for nurses; thus, they were developed specifically for use in F N I H B Nursing Stations in remote aboriginal communities. Minore et al. pilot-tested and substantially revised the tools prior to implementation. B y superimposing the B R C T data abstraction tools for the selected two indicator conditions onto a longitudinal framework that was conceptually similar to the one developed by Minore et al. (2002), I refined the B R C T tools so as to render these more specific to both application in a remote aboriginal health centre setting, and to the exploration of continuity o f care. I defined episodes of care more broadly than they were originally defined for the purposes of the B R C T , to better capture the dimension of time. This broader definition, when combined with a tool that was adapted to capture the dimension of time, facilitated capturing 85 data related to care provided between or outside of the more discrete episodes o f care that were defined by the B R C T . Each of the two data abstraction tools was also used to collect basic demographic data on each patient (i.e., age, gender). Date of birth was not collected, as this information could constitute a risk to anonymity in a small community. Patient charts were assigned a unique identifier number. Basic data included the first and last date of the identified episode of care. Data were collected for each patient visit during the episode of care, which allowed for the appraisal of care over a longitudinal continuum that facilitated an appraisal of the continuity of care. Chart information that was collected included: • Type of health provider seen (e.g., nurse, physician, resident, specialist) and number of different providers; • Type of visit (i.e., assessment, follow-up, referral / consultation, follow-up of referral); • Completeness of assessment; • Completeness of management and indicated interventions for routine care; • Completeness of patient monitoring and follow-up, including whether referrals and consultations were appropriately carried out; • Clinical outcomes appropriate to each indicator condition and reflected as either satisfactory or unsatisfactory outcomes (e.g., prenatal health outcomes were infant birthweight, gestational age at delivery, complications of pregnancy, labour or delivery, and attendance for routine postpartum care between six and eight weeks' postpartum; infant health outcomes were infants at age one year who had up-to-date immunization status, adequate weight gain and evidence that the hemoglobin had (a) been checked between six and 1 2 months of age, and (b) the hemoglobin value had been within normal limits); • Other clinical data specific to the indicator condition (e.g., gravida /para status, diabetic status for prenatal patients). The adapted and modified instruments were pilot tested in a small sample (n = 2 0 ) of patient charts that were sampled from a time outside of the time period defined for the purpose o f chart data collection. These charts were identical in format and type of content to the charts of patients later included in the sample. 86 The data abstraction tools were also updated to reflect contemporary methods of data collection using Excel© computer software. Data from each patient chart were put directly into a separate spreadsheet. Prenatal data spreadsheets were collected in a " P N Workbook", and infant care spreadsheets were collected in a " N B Workbook". Key indicator values were pre-set to roll data over automatically from each spreadsheet to a "Summary Sheet" o f statistics in each of the two workbooks. Basic descriptive and some comparative statistics could then be easily run within the Summary Sheet of the Excel© spreadsheet. Reliability and validity of the chart review instruments The original criteria for each indicator condition developed for the B R C T were developed in the early 1970's, at a time predating the formalized existence o f the evidence-based clinical practice guidelines that represent the standards of "best practice" today. Similar to the "expert" consensus approach upon which the development of evidence-based practice guidelines continue to rely, the original criteria were developed and pre-tested through a process that relied on a Peer Advisory Committee, composed of three family practice physicians who were acknowledged to have experience and expertise in primary care practice (Sibley et al., 1975). A comparison of the criteria with mortality, morbidity and patient outcomes data supported internal consistency and validity (Sackett et al., 1974; Sibley et al.). Subsequent studies appraising the utility of the method further supported the validity of using these criteria to evaluate patient outcomes (Sibley, 1976). Early studies using the indicator condition criteria and data abstraction tools also demonstrated a high level of interrater reliability, ranging from 88% to 94% (Sibley et al., 1975). Later researchers using the same instruments found similarly high levels of reliability, as discussed previously (i.e., in Chapter Three). Minore et al. (2002) did not offer statistical support for the reliability and validity of the data abstraction tools developed by their team; 87 however, after initial pilot-testing revealed weaknesses in the tools as originally developed, the tools were substantially revised and re-tested. As discussed above, the indicator condition criteria and data extraction tools adapted for this study were updated to reflect the F N I H B clinical guidelines, which represented the current standard of practice expected of nurses employed by F N I H B . These guidelines were also compared with other current and widely accepted evidence-based clinical guidelines (e.g., C M A guidelines) to ensure that the F N I H B guidelines reflected current standards of practice. During the pilot test, the criteria were reviewed by "expert" local practitioners familiar with practice expectations of nurses in the study site and were found to reflect current practice expectations. Chart Review Data A n important insight that resulted from pilot testing the sampling protocol and abstraction instruments was that patient charts could be sampled on the basis of maternal-infant dyads. This change in sampling strategy resulted in acquiring chart data that were uniquely appropriate to investigating the phenomena o f continuity of care, as it allowed continuity to be traced from prenatal care of the mother, through the perinatal period, and to the completion of the first year of infancy. This resulted in much richer data related to continuity of care than originally anticipated. In total, 128 patient charts were manually reviewed, representing approximately 22% of all births that occurred during the pre-determined eligibility period. For the first round of data collection, the time period was selected so as to ensure that all infants would have reached their first birthday prior to the start of data collection. This ensured that the infant chart review would capture data from the entire first year of infant health. For the first round of data collection, the charts of 50 mother-infant dyads were randomly selected using the 88 random number generator feature available in Excel© software, and the "Birth Book": a handwritten record of all infants born to women resident in the community. There were several instances when the chart of either a mother or an infant could not be located in the Nursing Station, and in these cases, another name was selected at random from the Birth Book. Early in the chart review process, two o f the infant charts could not be located after the mother's chart had already been reviewed and the data extracted. I decided to keep the data extracted from these two maternal charts, but in the future I ensured that both charts of a maternal-infant dyad were available prior to reviewing either one. If one chart of a dyad was missing, another dyad was selected at random. Thus, the first round of data collection included the charts of 50 mothers and 48 infants. In the four instances of twin births, only the chart of the infant that had been randomly selected was included in the chart review. The charts of five women were reviewed twice, for two different pregnancies during the defined time period. Being able to follow women through subsequent pregnancies in this way added a rich dimension to understanding the data related to continuity of care. For the second round of data collection a further 30 charts (i.e., 15 maternal and 15 infant charts) were selected using the same process, although the criterion for the eligible time period was moved forward by one year to offer the opportunity to review charts during a more recent time period. Thus, for the second round of chart reviews, seven infants had not yet reached their first birthday and therefore, data related to infant health outcomes at age one year were not available for those seven infants. Both qualitative and quantitative data were extracted from patient charts. Qualitative data were abstracted in narrative form as "contextual notes". These notes were a brief synopsis of each chart (i.e., there were 128 separate notes) based on review of the chart as a 89 whole. The meaning of a contextual note was often reflected upon in an accompanying interpretive memo. The contextual notes and interpretive memos provided context and an analytical strategy to help "explain" the quantitative data; or often, an explanation for missing or unclear chart information. The quantitative data described what was found on chart review; by adding a contextual understanding, the qualitative contextual notes created the possibility for also understanding how and why. Interview Data Data were derived from two types of interviews: formal interviews that were arranged beforehand with the participant, and informal interviews (i.e., focused conversations) that occurred " in the moment" or during the course of engaging in other fieldwork. Most formal interviews ranged between 30 and 45 minutes in length. These interviews combined semi-structured and unstructured questions, and loosely followed a pre-set list o f "trigger questions" (see Appendix F). Formal interviews were taped and later transcribed, with one exception; one interviewee declined my request to tape the interview. During this interview, I jotted down the participant's responses to my questions and wrote up these jottings as a comprehensive field note immediately following the interview. Two o f the interviews were done with two nurses who preferred to be interviewed together. Thus, 15 participants were represented in the 13 formal interviews. Formal interview data consisted of 12 interviews with direct health care providers (i.e., nurses, physicians, C H R s ) and three interviews with administrative level staff. The informal interviews represented substantive or key discussions with 16 individuals. In a few instances these individuals also participated in a formal interview. The informal interviews occurred spontaneously, at times when it was not practical or desirable to tape the conversation, or when requesting permission to tape (or to leave to get the tape 90 recorder) might have disrupted the flow o f the discussion or caused the participant to be less open in his or her communication. Informal interview data were written up as field notes as soon as possible following my interaction with the participant. Several participants appeared to view the taping of interviews with trepidation. Even after having read and signed informed consents agreeing to participate in a taped interview, they were careful to verify with me beforehand that no names would appear to connect them with the interview. A few participants appeared uncomfortable and in a hurry to complete the interview, but after the interview had "officially" ended and the tape recorder had been turned off, they continued to talk, sometimes for up to another 30 minutes. These participants appeared more relaxed and spoke more openly after the taping ended. In these instances, I jotted notes as soon as possible following the interview. These notes were later added to the transcribed interview as either pre- or post-interview observations. I considered certain of the Nursing Station staff to be key informants, being both members of the local community and also familiar with the operation of the Nursing Station. The purpose o f interviewing selected key informants was to illuminate and provide insight into aspects of the data collected through observation and chart review. M y initial research plan had called for interviewing patients and other community members (such as health council members or community health workers) who were identified as key informants largely on the basis o f their ability and willingness to share an "exemplar" experience of continuity or discontinuity of care, either as a patient themselves, or from a position of familiarity with patients' experiences. This strategy was consistent with the notion of sampling critical cases, or those cases that the researcher judges wi l l best illuminate or offer insight into particular aspects of the phenomena of research interest (Maxwell , 1998). It was also consistent with Fetterman's (1998) description of "judgemental sampling" (p. 480), 91 wherein the researcher selects "the most appropriate members of the subculture or unit, based on the research question" (p. 480). I was able to interview a few key informants who were community members and health workers. However, despite several requests made through the C H R s to negotiate access to patients or other community members who were wil l ing to be interviewed, I was unable to gain access to informants from this group. I was unable to determine an exact reason for this, but my sense was that my designated "liaison" was uncomfortable approaching individuals on my behalf, and I speculated that this was possibly because it was not considered part of the regular C H R job duties, thus my requests competed with the C H R ' s other expected work responsibilities. Observational Data Observational data were written up as field notes. Jottings were entered in a small notebook as events were observed and were written up more fully as field notes later the same day. Field notes were entered directly into a Word© template on a laptop computer. The template was designed in six sections: (a) date and time, (b) where the observation took place and who was present, (c) a description of the observation, (d) interpretive insights, (e) follow-up ideas, and (f) special notes. The latter three sections were not consistently completed, but provided a space for capturing ongoing ideas, insights and musings about the observations, and for connecting different observations, or sometimes linking interview and observational data through a process of contrast and comparison. There were two categories of observations in addition to the informal interviews that were written up as field notes, described above. The other two categories were patient-provider interactions (i.e., "encounters") and observations related to the structure and function of the Nursing Station and the community (i.e., "environment"). There were 16 observed patient-provider interactions and 38 observations related to context and structure. 92 I routinely joined the nurses in "morning report", which occurred at 8:30 a.m. every morning. This was the time when the nurses, the N I C , and any visiting health providers (e.g., physicians, residents, regional personnel such as educators or managers) met to review the patients that had called the nurse on-call after-hours the night before, as well as any other patients or situations that nurses wanted to discuss with their colleagues. It was also a general communication forum. One morning a week, morning report was extended to provide the nurses with time for "in-service", an educational session. Morning report and in-service time provided an opportunity to observe interactions among the nurses, and between nurses and other health providers, including communication patterns, referrals and consultations between nurses and physicians. Thus, morning report was a source of rich observational data. Demographic Data Demographic data pertaining to the study site community were accessed on-line at the Statistics Canada Community Profiles website. The demographic data are based on 2001 census data. Demographic statistics are related to population, family, housing, education, income, employment and religious characteristics of the community. These statistics provided important background information that helped to describe the broader community context in terms o f socio-economic status and the social determinants o f health. The demographic data also substantiated my impressions and insights about the community derived from observations and interviews collected during fieldwork. Demographic data were incorporated into the overall data analysis to provide context and help explain the analysis and findings. Artefacts Artefacts included photographs and miscellaneous maps, schematics and documents collected at the study site. These were categorized as being related either to the context of the 93 community and Nursing Station or to the process of delivering health care. Artefacts were not only a method of recording data, they also provided a graphic representation o f the data and were useful for helping to communicate a sense of the data to an audience. The artefacts were also useful as an aid to my recollection of the community and events after I left the field. There were 46 digital photographs (taken by myself) illustrating the community, Nursing Station, and nurses' residences. While the photographs included pictures of some individual nurses and staff members, no photographs were taken of patients or o f nurses during a patient encounter. The photographs evoke a sense of the remoteness of the study site, and of some of the everyday challenges community members and health care providers encountered in living and working in this remote setting. Locating a large-scale map of the community proved more difficult than I anticipated. In the end, I hand-drew a map showing the main features of the community. Similarly, when no one could locate a schematic of the Nursing Station, I hand drew a schematic o f the layout of the facility, graphically showing the structure and space in which nurses' work was organized. Other artefacts included a copy of an aerial photograph of the site, receipts and lists of prices of some common items at the local Northern Store, a newspaper article related to the effects of gang violence in remote communities, schedules o f flights in and out o f the site, and a book of local stories that offered a glimpse into the recent history o f the community. Artefacts related to the process of delivering health care included documents such as patient chart forms (e.g., the Antenatal Record, Wel l Chi ld Record), copies o f the relevant guidelines nurses were expected to follow (e.g., Gestational Diabetes Screening guidelines), and copied sections of pertinent resource manuals available to nurses (e.g., F N I H B 94 Community Health Resource Manual). I also obtained charts of the organizational structure of each of the two organizations that were responsible for health services in the community: F N I H B and the local Health Directorate, and examples of the types of teaching and information resource materials that were available to nurses. Personal Journal and Notebook While in the field I kept two handwritten documents related to data collection: a personal journal and a notebook. The personal journal was a repository of my impressions and perceptions of the community, the Nursing Station and the people. The journal provided a venue where I could reflect on the process of data collection, or elaborate on insights, ideas and inspirations or "hunches" that came to me as I began to mentally work through the meaning in the data I was collecting. The process of writing reflectively also provided an opportunity to explore the effect that I, as the researcher, exerted on the research process, as well as the biases and "blind spots" that I brought to the process. The journal is a piece of the "audit trail" whereby I may retrace my thinking during the process of data collection and recall how I came to particular insights or beliefs about the data. The notebook was a simple memory aid, a tool for capturing all the details of data collection " in the moment", or as events occurred. Like the journal, the notebook is also a piece of the audit trail. Overview of Participants The three main groups of participants in this study were health care professionals (i.e., nurses, physicians and administrators), Nursing Station para-professional and support staff (i.e., C H R s , clerical, security and maintenance staff), and patients. Roles sometimes overlapped; for instance, some nurses also worked in administrative positions, and as in most Nursing Stations, maintenance and housekeeping staff were also responsible for taking x-rays. Due to the general nature o f some observations made within the public areas of the 95 Nursing Station and around the community, many individuals contributed anonymously to the overall data collection but did not directly participate in the study. For example, interactions were observed between clerical staff and patients in the waiting room. The individuals who are described here as participants each participated in the study in a formal sense, either through formal and informal interviews, or formal observations. The largest group of participants, and the participants who were the main foci of interest in this study, were nurses. Fifteen nurses participated in the study (see Table 1). This group included nurses employed in dedicated public health and home care roles who were employed by the Band, as well as the nurses employed by F N I H B in the Nursing Station. Nurses' ages ranged from the early 20's to the mid-50's. Two nurse participants were male. Thirteen nurses were of European or Asian ancestry, and two nurses identified as First Nations. O f the seven nurses who were in permanent staff positions, three had been in their permanent position for less than three months, and three had been in a permanent position with F N I H B for more than one year. Two nurses in permanent positions were employed in administrative nursing roles. One of these two nurses identified as First Nations. A l l aboriginal participants in this study further identified as specifically First Nations people. Four nurses employed in a relief nurse position in the study site had previously held a permanent position with F N I H B in a different community for between eight months and three years. Three of the relief nurses were relatively new to Nursing Station work, having worked intermittent temporary assignments for a year or less. Most of the nurses who held (or had held in the past) a permanent position with F N I H B had taken the Primary Care Skills Course, and diploma-prepared nurses in a permanent position had also taken the Community Health Nurse Upgrade course offered by F N I H B (these two courses wi l l be described in greater detail in Chapter Five). A t the time of the study, one nurse was actively engaged in 96 obtaining a master's level nurse practitioner degree, and one diploma-prepared nurse intended to start a B S N degree within the year. Table 1 Demographic Characteristics of N = 15 Position Education Female = 13 Male = 2 Permanent staff = 7 Diploma prepared = 6 Age range: early 20's to mid-50's Relief staff = 7 B N / B S N prepared = 6 European or Asian ancestry = 13 Orientee = 1 Primary Care Skills Course = 6 First Nations — 2 Community Health Nurse Upgrade Course = 3 Unknown =3 O f the other health care professionals who participated in this study, two were physicians (one general practitioner and one specialist). Both had worked in the study site for several years and were familiar with the community, the operations of the Nursing Station and nursing practice expectations. One physician was male and the other was female. Physicians were not in the community on a permanent basis but visited on either a weekly or an intermittent basis. Nursing Station para-professional and support staff that participated in the study included two CHRs , a health facility administrator, and staff such as maintenance, housekeeping, security and clerical workers. These participants were Band employees, had been in their positions for several years, identified as First Nations and were members of the local community. Patients' participation in the study was limited to observations of their encounters with health care professionals (i.e., nurses) and in five cases, with para-professional health staff, and informal conversations that arose during such observations. Twenty women participated in observations; six observed encounters involved prenatal care and therefore involved only the woman as the patient, and 14 encounters involved women and their 97 children. O f the children who participated in the study, all but one were accompanied by their mother (one child was accompanied by a grandmother). The ages of the children ranged from newborn to three years of age, with only three children over the age of one year. Protecting the Anonymity of Participants While recognizing the need to convey to the reader whose perspective is represented or illustrated by specific data presented in this dissertation, the relatively small number o f individuals involved as participants, along with the impracticality of maintaining complete anonymity of the study site, mandated that some descriptive details be omitted or disguised to protect the anonymity of individuals. Therefore, wherever feasible, and particularly when the specific identity o f an individual participant may influence how the reader understands the data, nurses have been identified as nurses, physicians as physicians, administrators as administrators, and so forth. I have been as specific as possible given the obligation to protect participant anonymity. However, where more specific identification of an individual participant may compromise anonymity, I have deliberately used a vague descriptor to identify the participant. For example, in some instances I have used "health professional" rather than identifying whether the participant was a nurse, physician or administrator. While both male and female health care professionals participated in this study, I have elected (in most examples) to refer to all health care professionals as female to further protect the anonymity of participants. I have also endeavoured to identify whether a health professional was First Nations or not, in examples where knowing this information about a participant's perspective may be relevant to better understanding the data. Data Analysis Analysis is one of the most engaging features of ethnography. It begins at the moment a fieldworker selects a problem to study and ends with the last word in the report or ethnography. Ethnography involves many levels of analysis. Some are simple and 98 informal; others require some statistical sophistication. Ethnographic analysis is iterative, building on ideas throughout the study. (Fetterman, 1998, p. 494) The process of data analysis began in the field, and was conducted concurrently with data collection. Initial analysis involved repeatedly reviewing the data, reworking and rewriting from jottings in the field, to field notes that were continually expanded through the addition of interpretive memos. This strategy allowed me to identify key ideas in the data and to follow up on these while still engaged in fieldwork. Fetterman (1998) stated that "analyzing data in the field enables the ethnographer to know precisely which methods to use next, as well as when and how to use them" (p. 494). Thus, analysis of data in the field represented a critical part of the iterative process that characterizes ethnographic data analysis (Fetterman). Quantitative data related to patient demographics and health outcomes that were extracted from patient charts were analyzed using both Excel© and SPSS 10® software. Statistical data were primarily categorical and ordinal; therefore statistical analysis was limited to descriptive and comparative statistics. These statistics were sufficient to meet the intended purposes of including quantitative data in this study, which were to describe patient health outcomes on the selected two indicator conditions, and to complement and enrich the qualitative data related to continuity of care and nursing practice. Qualitative data related to the chart review process (i.e., the contextual notes data) were analyzed using content analysis. Several key dimensions that characterized prenatal and infant care were identified through the process of content analysis. These key dimensions reflected the dimensions of continuity of care described by Reid et al. (2002). Content analysis of the contextual notes data also revealed dimensions of adequacy of care, health outcomes, issues embedded within the broader community context that influenced health 99 (e.g., poverty, illiteracy), and the context and structure of health services (e.g., communications between health providers or agencies). Interview and observational data were also initially analyzed using content analysis. Analysis of the qualitative data set proceeded in two stages. Notably, I differentiated between the initial step of content analysis of observational, interview and chart review data, and the subsequent thematic analysis of the data overall. The first stage was based on a process of open coding and memoing, wherein data was coded line-by-line without reference to a priori theoretical assumptions. The researcher's goal at this stage "should be to generate as many codes as possible, at least initially, without considering possible relevance either to established concepts in one's discipline or to a primary theoretical focus" (Emerson, Fretz, & Shaw, 1995, p. 152). Thus, the process of open coding helped to avoid imposing my own biases into the process of analysis. Through the process of open coding, patterns and core themes began to emerge from the data. During the second stage of analysis I began to integrate the data from various sources by l inking together the core themes and integrating the findings that had emerged through content analysis of each of the different parts of the data set, as common over-arching themes emerged. I examined the data using the three dimensions of continuity o f care (Reid et al., 2002) as a guide to analysis (see Appendix G). The selection of major themes was guided by the purpose and objectives of the research explicated in Chapter One, by the conceptual underpinnings of this research, presented in Chapter Two, and by the relevant literature that was critiqued in Chapter Three. These analytic strategies enabled a focused thematic interpretation of the data set as a whole and allowed me to make sense o f the large amount of data I had. Thus, there were two levels of analysis. Each dimension of the data set was first analyzed independently using content analysis, and was subsequently interpreted in 100 conjunction with the findings that emerged from each other dimension of the data set. A t a second, and higher level of analysis, qualitative data from the overall research process were analyzed thematically in conjunction with the findings o f the quantitative data analysis. Different levels of analysis were not viewed as discrete linear analytical events, but as an iterative and ongoing development of ideas (see Appendix C). Validity Fetterman (1998) claimed that triangulation is "at the heart of ethnographic validity" (p. 495). Stake (2000) similarly describes triangulation as key to instilling confidence in the results of qualitative research: "Triangulation has been generally considered a process of using multiple perceptions to clarify meaning, verifying the repeatability of an observation or interpretation" (p. 443). B y encompassing multiple sources of data and diverse data collection strategies, this research incorporated triangulation strategies that contributed to confidence in the validity of study findings. Trustworthiness or credibility in the data in qualitative research is fostered through two main strategies. One o f these is the collection of, "r ich" data, or "data that are detailed and complete enough that they provide a full and revealing picture of what is going on" (Maxwell , 1998, p. 94). Through "thick" description, rich data provide a sufficiently comprehensive view of the research phenomena to allow the emergence of contrary data, or data that challenges the researcher's assumptions, biases and theories (Maxwell) . Rich data therefore "tests" (p. 94) the credibility of the research by forcing the researcher to examine his or her biases and pre-conceived ideas about the phenomena o f interest in the light of what is actually found in the data. The ethnographic data collection strategies that were used in this study resulted in the accumulation of thick description and rich data. 101 The second main strategy for increasing the trustworthiness of qualitative research is through auditability o f the research process. The purpose of an "audit trail [is] to clearly document the researcher's decisions, choices, and insights" (Morse & Field, 1995, p. 144), that is, to make the research process transparent. The process of recording and memoing fieldnotes offered a strategy to maintain a clear and transparent audit trail. Keeping a reflective journal o f my thoughts and insights as the researcher, and of my perceptions of my own influence on the research process, was another strategy for maintaining an auditable record. The work o f writing reflectively as an adjunct to writing field notes offered a space in which to explore my own pre-existing biases and assumptions as these emerged during the research process, and was an important piece o f the "audit trail". In this way, biases, assumptions or "blind spots" were made explicit, which Fetterman (1998) suggested was a way of "mitigating the negative effects of bias" (p. 473). Additionally, reflexively engaging with the data during the process of writing field notes was a strategy that allowed me to situate myself in relation to the data, so it became transparent not only how my perspectives shaped the data, but also how as the researcher, my perspectives were shaped through my engagement with the research process (Emerson et al., 1995). Thus, reflexivity was key to the credibility o f negotiated and mediated meanings that I aimed to construct and have represented as research findings (Emerson et al.). While this research was conducted in a study site that was selected in part because it was considered to be typical of Nursing Stations in remote aboriginal communities generally, it was not the intention o f this study to be able to generalize research findings per se to other settings. However, through developing a better understanding o f how nurses in primary care 102 influenced patient health outcomes and continuity of care in the study site, I expect this study wi l l contribute to our knowledge of how health services are delivered in similar settings. Ethical Considerations This research was guided by the Tri-Council Policy Statement: Ethical Conduct for Research Involving Humans, and in particular by Section 6, Research Involving Aboriginal Peoples (CIHR, Natural Sciences and Engineering Research Council o f Canada, Social Sciences and Humanities Research Council o f Canada, 1998). 2' Prior to commencing this study, ethical approval was obtained from the Behavioural Research Ethics Board, Office of Research Services, at the University of British Columbia (see Appendix H). Approval to conduct the research in the study site was obtained from the appropriate stakeholders at the regional and facility levels, as well as from the Band council on behalf of the community. Confidentiality and anonymity of all individual participants has been maintained. Data collected from patient charts was collected using coded identifiers that were known and accessible only to myself as the researcher. Identifiers that might link individuals to chart information were maintained until the completion of data analysis, at which time this information was destroyed. Anonymity of the study site and the community has been maintained to safeguard the anonymity of individuals employed or associated with the study site. In general, the following specific measures were employed to ensure that ethical standards and integrity were upheld: (a) written informed consent was obtained from each individual directly involved in direct observation and / or interviews (see Appendix I); (b) participants understood that their participation in the study was voluntary and they 2 1 As of October 2005, the CIHR Guidelines for Health Research Involving Aboriginal Peoples have been made available as a draft for consultation. The finalization and adoption of these guidelines within the Tri-Council Policy Statement is expected to occur by Fall 2006 (CIHR, 2005). 103 understood that they could choose to decline to participate or withdraw from the study at any point; (c) no information was collected to be used as data that could be used to identify individual participants; and (d) any tape recordings or records that might link individual identities to the study w i l l be stored in a secured location during and following the study and wi l l be destroyed at the conclusion of the research. Summary A n ethnographic method supported the use of a mixed methods strategy, or a variety of data collection strategies and perspectives, which were essential to providing a multifaceted and contextualized picture of health outcomes and continuity of care in the study site. Data were collected through observations in the clinical setting, and formal and informal interviews with health professionals, patients and para-professional staff in the Nursing Station. Relevant documents, photographs and artefacts were also collected as data. Quantitative data related to clinical health outcomes were abstracted through retrospective chart review. Data abstraction was based on two health indicator conditions: prenatal care and care of infants through the first 12 months. 104 C H A P T E R FIVE T H E C O N T E X T AND S T R U C T U R E O F NURSES' W O R K : W O R K I N G A T T H E MARGINS Introduction to the Research Findings The research findings are presented in four chapters that are organized in terms of (a) the context and structure of nurses' work (b) health outcomes and continuity of care, (c) the process of nursing care, and (d) the intersection of context and structure, process and outcomes. Maintaining a primary focus on continuity of patient care, the interpretation o f the findings presented in each chapter builds on the data and interpretative insights presented in the previous chapter(s), fostering increasingly abstract levels of interpretation. I begin this chapter with a short introduction to the study site community, based on my field observations and Statistics Canada demographic data, followed by an interpretation of the data related to the context of the broader community. The second part of this chapter describes specific aspects of the context and structure of health services organization and nurses' work in the study site. The interpretations presented in Chapters Five through Eight are grounded in data collected through the data collection strategies as described in Chapter Four. Stepping into the Community: First Impressions The community of Snowy Lake is located in northern Manitoba, situated between one and two hours' flying time north of Winnipeg. 2 2 Snowy Lake is accessible year-round by air, and by an ice-road for several weeks during the winter. Two commercial air companies provide at least daily scheduled air service between Snowy Lake and the closest larger service centre. A handful of neighbouring First Nations communities are accessible by water during the summer or by snowmobile during the winter months, allowing families and friends to visit between these communities. Flying into Snowy Lake on a clear day reveals the extent to which the community is surrounded and isolated by water. The flat landscape below is a myriad of green, densely wooded patches of land surrounded by blue lakes and rivers; at times it is impossible to Snowy Lake is a pseudonym. Specific details about the study site have been omitted or disguised to protect anonymity of the community and of the individuals who participated in the study. 105 discern between larger islands and the "mainland". N o roads or buildings or rising smoke, nor any other signs of human habitation interrupt the pristine views that pass below, until finally the sun glints silvery off something in the distance. A s the plane approaches, the glint grows into a string of buildings lining a point of land. The plane banks and turns, and across the water a second small patch of buildings appears. A few tiny boats trace white wakes across the water as we circle the airstrip to land, and a plume of yellow dust races a vehicle along the one road. This is Snowy Lake, and Snowy Lake could be any one of several northern Canadian communities. Like many northern First Nations communities, Snowy Lake is a divided community. The airstrip, Royal Canadian Mounted Police ( R C M P ) , Northern Store and a handful of residences are located on one side of a body of water, while the reserve (where the Nursing Station is located) is on the other. Thus the commute between the airstrip and the Nursing Station involves crossing either by boat (in the summer) or by snowmobile or vehicle (after the ice has formed). The crossing takes just a few minutes, a bridge could easily span the distance, but a short drive through each of the two adjacent communities reveals that more than water separates them. The majority of residents are First Nations people who live on the reserved Indian lands. The reserve land of the Snowy Lake First Nation is an area of less than 50 square kilometres. While most nurses and teachers live in housing provided on the reserve, a handful of non-aboriginal residents live off reserve. Non-aboriginal residents are generally temporary, and have come to the community specifically to provide services such as nursing, teaching, policing or managing the local Northern Store. Less than 1% of the population of Snowy Lake identified as being non-aboriginal in the 2001 census (Statistics Canada, 2001). Much of this 1 % lives together, segregated on one side of the water (the side where most of the services and resources are located), while the majority of the population lives on the far side, the "reserve side". The weather and the march o f seasons dictate much of life in Snowy Lake. The community is located north o f 53° latitude (for comparison, Prince George and Edmonton are between 53-54° N) , in the taiga, or northern forests. The forest surrounds Snowy Lake, creeping up close and silent around the houses. The climate is characterized by short, cold, humid summers and long, cold, snowy winters. Freeze-up occurs about November each year, and break-up is usually in late spring. For five or six months of the year snow covers the landscape and the daily temperature rarely inches past freezing. Winter temperatures regularly drop to minus 40° C , while average summer temperatures seldom reach 20° C , and other than in July and August, frost may occur on any night of the year (Environment Canada, 2005). Despite daily scheduled flights, transportation in and out of the community is often tenuous. Flights are sometimes "weathered out", meaning that weather conditions prevent flights from landing. High winds, storms, blizzards, fog and extreme low temperatures may preclude flying on any given day. The ice-road too is dependent on weather conditions. Some winters the ice isn't thick enough to support vehicular traffic across the lakes, or i f freeze-up is late or spring thaw arrives early, the ice-road may be open for just a few weeks. Freeze-up and break-up, as in other northern communities, present particular challenges to transportation within the community, as for a week or two each fall and again in the spring, the ice is too thick to permit boat traffic, yet not thick enough to allow people, snowmobiles or vehicles to safely cross. Freeze-up and break-up have the tendency to leave southerners indignant at nature's power to effectively strand an entire community for a week or two each 106 fall and spring: " in this day and age!" But the locals take it in stride, uncomplaining; it is just part of the pace of life in Snowy Lake. The First Nations population o f Snowy Lake numbers between 2-3,000 people. At the time of the 2001 census, the median age of residents was about 18 years old, and over 40% o f the population were younger than 15 years (Statistics Canada, 2001). Only 3.5% of the population were aged 65 years and older. Traditionally, the people of Snowy Lake found their living in hunting and fishing, and in the post-contact era, in the fur trade. Many families continue to rely on hunting and fishing to supplement store-bought foods, and a few individuals continue to trap. Fewer than 10% of the population have fulltime, year-round employment and the unemployment rate hovers around 30%, about five times higher than the overall Manitoba unemployment rate o f 6% (Statistics Canada). In 2001 the median average income of all Snowy Lake residents aged 15 years and older was less than $9500, compared to more than $12,400 for Manitoba residents overall (Statistics Canada). Close to 40% of income earned in Snowy Lake was government transfer payments (i.e., paid out largely as social assistance dispersed by local Band administration), versus 23% for the rest of Manitoba (Statistics Canada). In contrast, the average earnings of all persons who had earned employment income was about $18,600, compared to $19,200 for Manitoba overall. The Band is the largest employer in the community and administers nearly all jobs on the reserve. For example, other than the Nursing Station nurses, who are federal employees, all o f the Nursing Station staff (including the public health and home care nurses) are Band employees, as are the social services staff and the teachers. The Band administration forms the local governing body. Band members elect the Chief and several council members. The Band operates the schools, a local Band constabulary, most social services, and most other community services. A n y programs within the community require the permission o f the Band administration. Band chiefs and councils wield legal powers beyond that of off-reserve municipal authorities; for instance, Band chiefs and councils have the authority to order searches (i.e., for contraband drugs or alcohol), enforce curfews, or deport undesirable non-Band members from the community. Snowy Lake has both an elementary school and a high school up to grade 12. Less than 2% of Snowy Lake residents have completed high school. About 6% of men over age 25 have completed high school, compared to only 4% of women (Statistics Canada, 2001). However, while no men aged 25 and older held a university degree at a bachelor's level or higher, 2.7% of women had obtained that degree (Statistics Canada). Ninety-seven percent o f Snowy Lake residents who identified as First Nations spoke a First Nations language at home, and 95% identified that language as their first language (Statistics Canada). At the time of the 2001 census, there were fewer than 500 houses in Snowy Lake. A l l but a few houses on the reserve were owned by the Band and were provided to families at no charge, although families were expected to pay for heating and hydro costs. A small number o f newer houses are available to residents as rentals. Houses are similar to houses on reserves across Canada: usually small (2 bedrooms, 1 bathroom) bungalows built either on a slab or off the ground. The 2001 census identified that more than 50% of the dwellings in Snowy Lake were more than 14 years old, over 40% were in need of minor repairs, and over 30% were in need of major repairs (Statistics Canada, 2001). Health Canada (2003) defines "adequate housing" as "housing units that do not require any minor or major renovations or replacement" (p. 65). In 1999, only 56.9% of First Nations on-reserve houses were considered adequate (FNIHB) , suggesting that fewer adequate housing units exist in Snowy Lake than in most other reserve communities. Moreover, the 2001 census identified 107 approximately 25 more families in Snowy Lake than there were houses, suggesting that more than one family often shared a house. Significantly, the average family included 5 members (Statistics Canada). Overcrowded housing is recognized as a problem in First Nations reserve communities generally (FNIHB). There is an arena in Snowy Lake but few other recreational facilities, although the school gyms are sometimes opened for activities and events. Notably, the arena was closed and awaiting renovations during the study period. There is no theatre, no pool, no mall and no Internet cafe. T im Hortons '®, the trademark of rural Canada, is notably absent. A local television station scrolls community events, and a local radio broadcasts in the local language. Local residents make do with dial-up Internet access via ancient telephone lines that are subject to debilitating static. There are both a Catholic and a Protestant church on the reserve; the majority of residents are Protestant (Statistics Canada, 2001). Bingo is an important part of community life, as in other First Nations communities. Snowy Lake is a "dry" community, meaning that alcohol is not permitted on the reserve, although off reserve residents who are not Band members are permitted to bring in alcohol for personal use. Thus residents of the off-reserve community may enjoy alcoholic beverages at home and at social events while on-reserve homes and events are "dry". The few roads are unpaved and alternate between thick yellow dust, slick mud gumbo or ice, depending on the weather. There are no sidewalks or shoulders to walk on, and the uneven footing makes walking along the roads risky, yet walking is the common mode of transportation for many people. Vehicles can be driven into Snowy Lake over the winter road, but the ungraded roads, mud and potholes, and the lack of resources to maintain vehicles ensures that vehicles have short and generally unhappy lives. Abandoned vehicles litter yards and road edges, left to rust where they died. This is Snowy Lake, and Snowy Lake could be any one of many communities in northern Canada. It's a Different World Here At first glance Snowy Lake might appear to many newcomers new to the "the north" as a somewhat rustic northern community, intriguingly quaint and isolated and in some ways attractive by virtue of being a new and different experience. On arrival, nurses and other health care providers are taken directly to the Nursing Station residence, where they are settled into living quarters that are comparable in most ways to their l iving situation at home: suites are private (or semi-private), have central heating, running water, hot showers, laundry facilities, stoves, microwaves and satellite television. Nurses experience living conditions that are substantially better than those experienced by most community members, and at least 108 initially, nurses tend to be isolated from the harsh reality of the housing conditions and poverty that prevail in the community. Despite being insulated to some degree by l iving and work conditions that took place within the "protected bubble" of the Nursing Station environment, most nurses eventually encountered the reality of the surrounding environment. This is an environment drastically different to the environment to which most nurses were accustomed. It is an environment hallmarked by poverty, inadequate housing, a lack of resources that are taken for granted in most other Canadian communities, and profound health and social issues. While most of the nurses who participated in this study had received a three weeks' employment orientation that included an introduction to the isolation and the different environment they could expect to find when they arrived in the community, most felt nevertheless unprepared for the reality: "nothing prepares you for the reality of it". Many social issues became apparent just by taking a drive around the community. However, since opportunities for nurses to get out into the community were limited, they sometimes spent a considerable time in northern communities before they really became aware of the prevailing social conditions. One nurse, who had been working as a relief nurse in the region for the past year, had been introduced to the realities of life in the community only shortly before I interviewed her. Her reaction revealed her struggle to come to grips with an unfamiliar context and the inequity it implied: "it just was amazing to me that in this day and age, this is like a third world country here, there's no running water, I mean I cannot believe it. I think it's disgusting, atrocious". Sub-standard housing conditions, the lack of running water, transportation and the high cost of food and supplies were among the broad contextual issues that were raised repeatedly by all participants. One First Nations participant (a C H R ) summed up these issues in an interview: 109 We're faced with 3rd world conditions here, we are surrounded by water but still we have no running water, more than 50% are without running water. The majority o f the people get their water from these pipe stands that you see on reserves and during the winter time it's very inconvenient because most of those water lines are frozen. Most people don't have a vehicle. [Families on welfare] only get income once a month and that's their welfare, and their welfare has to stretch for the whole month. Sometimes the winters get really harsh and they have economic problems, so that they can't keep up financially. From the nurses' perspective, access to the local Northern Store was problematic, given the hours they were expected to work in the Nursing Station and be "on-call" after-hours. This was an obstacle that some nurses had not been prepared for prior to arriving in the community. Most nurses brought some food supplies into the community with them, especially i f they were staying for a shorter time, but i f they needed to replenish their supplies locally, they experienced first-hand the challenges faced by local community members both in accessing the store and the high costs o f purchasing items in the community. One nurse described local costs as "exorbitant". M y field observations substantiated the high cost of buying basic needs items in the community; for example, a trip to the community Northern Store revealed a case of 12 tins of infant formula cost $58 (compared to $26 for the same case in an urban centre), $35 for a box of disposable diapers (less than $20 in an urban centre), and $25 for a box o f laundry soap (about $8 in an urban centre). Transportation to the store (either in a small open boat with no life jackets on board, or across the ice) was also costly and inconvenient, as well as potentially treacherous. Thus the simple, taken-for-granted chore of purchasing needed items at the local store entailed an inequitable degree of hardship that would be considered unacceptable in most Canadian communities. In seemingly minor, every-day tasks of l iving both community residents and nurses experienced marginalization. 110 While non-aboriginal health care professionals developed an appreciation for the challenges o f everyday life in the community, it was the First Nations health care professionals and para-professionals whose own social location and lived experiences as community members allowed them to express most eloquently the impact these challenges had on the day-to-day life of community members. Non-aboriginal professionals may have lacked the same depth of appreciation in part because they seldom went into homes in the community. O f the non-aboriginal health care professionals who participated in this study, only two or three could recall having ever made visits into homes on a reserve. This finding was in contrast to established or traditional nursing practice in F N I H B Nursing Stations, where home visiting has long been an expectation of the nursing role. Home visiting has traditionally exposed nurses to social and economic conditions in the local community. Thus, while non-aboriginal nurses understood from their own perspective that social conditions presented challenges, they remained largely unaware o f the full extent of the implications these same challenges posed to community members who lived in less privileged circumstances. A s one First Nations nurse described in an interview: The life is very difficult here. Running water and heat are big factors, in the morning especially because they have to make a fire to warm up the house in the morning; sometimes when you go into the home early in the morning, they're in bed just so they can stay warm. In some of these homes their water is frozen, so then they have to bui ld a fire and melt the water so they can wash up. There's two or three families l iving in one three-bedroom house. There's grandparents and their grown up kids and then their kids, there's little babies running around and those are the great grand children in that home. People don't eat properly only because they're on a budget, on welfare. There's no running water so therefore hygiene is a factor. It's very difficult, l ike, I don't want to be mean or anything but that's just how people live, like some of them don't even change their clothes for a week, for example, only because they don't have any [water], how would they wash clothes? Y o u need lots of water to wash your clothes, they don't have dryers at home, they don't have towels to wash with, things like that. Their priority is to buy food. I l l For the most part both First Nations and non-aboriginal health professionals recognized the essential social inequity that local l iving conditions represented: "it's still not up to the same standard as the rest of the country", and "the l iving standards are not the same as in the city". Health care providers in general also articulated an appreciation of the relationship between inequitable social conditions and marginalized health. In response to my question asking nurses what they viewed as the major health issues in the community, most indicated an appreciation of the relationship between health and the social determinants of health that were rooted in the context of the community. For example, the lack of running water, of adequate housing, recreational facilities, and healthy nutritional choices at the local store, were all identified as contributing to poor health or unhealthy lifestyle "choices". However, not all nurses demonstrated this appreciation, nor did all nurses comprehend the influence of the social determinants of health to the same degree or on a consistent basis. For non-aboriginal health care professionals, there was a certain paradoxical quality to l iving and working in the different world of the First Nations community. The paradox was based in the circumstance that non-aboriginal health care professionals lived in two worlds in the community: the familiar world of the Nursing Station, and the unfamiliar world of the surrounding community. These professionals came to grips on a daily basis with the realities imposed by the broader community context, yet they lived and socialized for the most part in a world that was in some ways not very different from their lives in mainstream Canadian society. The paradox existed because these non-aboriginal health care professionals were not really a part of the community or of the other world they associated with the community; the local community was not really their world. This is part of the paradox of outpost nursing that Hodgson described in 1982, and while her work first drew attention to the contextual issues that created the paradox more than 20 years ago, little has changed since 112 that time. For example, a seemingly insignificant observation that struck me, as a researcher, as singularly representative of the paradoxical world that non-aboriginal heath care professionals inhabited was finding workmen planting a lawn 2 3 in front of the nurses' residence. Landscaping the nurses' residence in a community that was typified by many local residents and health care providers as "third world" illustrated a quintessentially post-colonial dynamic that in turn illustrated the perpetuation of inequity and marginalization within the community. The paradox that nurses experienced in living and working in the different world of the First Nations community allowed nurses to maintain a distance between themselves and the gritty reality they encountered in the clinic and out in the community. This distance between nurses and the community created space for nurses, physicians and other visiting health professionals to remain disconnected from the broader context and disengaged from patients' lives, and ultimately contributed to incongruence between reality and some professionals' perceptions of health and health services. Thus fragmentation and discontinuity were rooted in the historical and socio-political fabric of aboriginal health. Nurses were employed in a community-based nursing role, yet lived and worked within the "protective bubble" o f the Nursing Station and nurses' residence that isolated and insulated them from the surrounding community. This paradox is key to understanding the interpretation of findings presented here. The Local Context of Health Services: Structure and Infrastructure Responsibilities for health services in the study site community were shared between F N I H B and the local Band government. F N I H B maintained responsibility for the Nursing 2 3 Most homes in the community did not have a lawn, although in contrast, several homes in the off-reserve area did have lawns, gardens and in some cases, were fenced. 113 Station and direct authority over the services that were provided from the Nursing Station, which included nursing, medical and dental services. During the data collection period three of the seven full-time nursing positions were filled by permanent FNIHB employees, and the other four positions were staffed by "relief nurses for varying lengths of time, from a few days to two months. Relief nurses were usually employed by FNIHB on a casual basis, although occasionally were contracted by FNIHB through a nursing agency. Medical and dental service providers were also contracted by FNIHB (both a physician and a dentist were in the community for several days each week). However, most of the para-professional Nursing Station staff were employed by the Band under the health transfer agreement with FNIHB. Band-employed auxiliary staff included clerical, housekeeping and maintenance staff, as well as the Community Health Representatives (CHRs), who worked more directly with the nurses as health care providers and medical interpreters. The Nursing Station was directly supervised by the Nurse-in-Charge (NIC), who was a FNIHB employee, and administered by an administrator who was Band-employed. The NIC reported through the Zone Nursing Officer (ZNO) to upper-level FNIHB management, while the administrator reported to Band and Council through the Band-employed Health Director. Thus there were two lines of authority within the Nursing Station, and the two groups of employees experienced different conditions of work. For example, Band-employed workers sometimes received extra time off with pay (e.g., for local holidays) that FNIHB employees weren't entitled to. Both employees and managers were at times unclear as to who was responsible for particular aspects of work or for particular employees, for example, one participant who worked in an administrative role stated: "they have their own supervisor, I tell the supervisor that this is what they should be doing and if it doesn't come to that, I 'll have to speak with them... I have to intervene sometimes". Differences in working 114 conditions and reporting lines fed into a perception of difference or inequity between the two groups of employees, and sometimes to tensions. This kind of tension and fragmentation of health services started within the Nursing Station itself and created a ripple effect that spread through the different agencies that ostensibly worked in collaboration with the Nursing Station to provide health services in the community. Several aspects of health and social services delivered within the community had been transferred to the Band. Transferred services included the Home Care and Mental Health programs, as well as a variety of programs such as the Native Alcohol and Drug Abuse Prevention Program (NADAP) and social and family services. About a year prior to data collection for this study, the Public Health services had also been transferred to the Band. This transfer enabled the Band to employ a Public Health Nurse, who was responsible for the "public health" programs (i.e., immunizations, infectious disease control) previously provided by nurses working in the Nursing Station. The Public Health nurse and public health services were located in the same building as the Nursing Station. However, as a Band employee, the Public Health nurse was not under the supervision of the Nurse-in-Charge, but reported to the Health Director. While some of the auxiliary staff who worked in the Public Health unit were under the supervision of the Public Health nurse, others were under the supervision of the administrator, and some (e.g., the CHRs) also at times and in certain situations worked under the supervision of the NIC or the Nursing Station nurses. Thus, lines of communication, reporting and responsibility within the Public Health unit and between the Public Health unit and the Nursing Station tended to be ambiguous and further contributed to fragmentation within health services in the study site community. The Band-administered health and social service programs were health resources within the community. For example, nutritional counseling, diabetic teaching and home care 115 were some of the available programs, yet access to and utilization of these resources was a source of contention and tension between the different groups of health service professionals. Health care providers in the public health area (e.g., public health and home care nurses) perceived that nurses and physicians in the Nursing Station failed to take advantage of the available Band-administered resources: "I don't think that the Nursing Station utilizes the resources that we have down here... it's unfortunate that Nursing Station staff [have been slow] to clue into what valuable resources we have and make referrals appropriately". Field observations supported the perceptions of public health workers and suggested that nurses in the Nursing Station were in fact sometimes unaware of either the range of available resources, or of their own role in the referral process. In contrast, nurses working in the Nursing Station perceived that the Band-administered resources were either not readily available or that referrals were not followed through on, or not followed through on in a timely fashion. Data from review of patient charts suggested that follow-through on referrals might have been inconsistent, as the following excerpt, recorded as a contextual note during a chart review, illustrated: 19 year-old multip. Glucose tolerance test at 12 weeks revealed diabetes. Referred to [nutritionist] for diabetic counseling at that time, referral from R N on chart. This was followed up appropriately on 3 subsequent visits, and each time it was discovered that the patient had not yet received diabetic teaching. The intervention in each instance was to re-refer the patient. It appears that the patient reached at least 26 weeks with "no knowledge of diabetes", and there is no evidence that diabetic teaching ever occurred. Delivered an infant greater than 4000 grams. Thus the utilization and availability of resources was a factor that influenced continuity of patient care. Resource use was compromised by fragmentation of health services that began with organizational and structural ambiguities that contributed to perceptions of difference and inequity between Band-employed health care providers and FNIHB-employed nurses. Perceptions, misperceptions or assumptions that resources were 116 accessed inappropriately, were devalued, or that resources were unavailable when needed, created tension between Nursing Station and Public Health nurses. The lack of clear and unambiguous lines of communication and responsibility to support clarification of these issues was a source of inter-agency discord. Failure to clarify and resolve these issues contributed to conflict and further fragmentation between health services within the community. These tensions and conflicts were often not apparent on the surface, and in fact, both interpersonal and inter-agency relationships among professionals within the community were visibly collegial and congenial. Tensions and conflict that contributed to fragmentation of services were more often enacted covertly and at the individual level. The discourse of non-aboriginal participants suggested a subtle devaluing of Band-administered health resources, a variation on othering practices (Johnson et al., 2004) that may inadvertently propagate inequitable power dynamics between groups and in consequence, lead to "marginalization, decreased opportunities and exclusion" (p. 254). For example, a field note recorded the following observation during nurses' morning report: There was then some discussion about community resources (which I asked about) but the ANIC [Acting Nurse in Charge] more or less shrugged her shoulders and said the Band-run community health services had a Mental Health Worker, " i f you could get hold of her". These services seem to be considered a write-off [by the nurses]; my impression is that there is not a lot of interaction / collaboration between US and THEM. Thus, othering practices were reflective of the "other world" of the local community that non-aboriginal health care professionals perceived as being different to their own world. The concept of othering both reflected and contributed to the finding that nurses worked at the margins within the context of health services in the community. 117 It's a Demanding Job: Contextual and Structural Issues that Influenced Nurses' Practice Although several issues embedded within the broad context and local structure of health services delivery influenced nurses' practice and their ability to influence continuity of patient care and health outcomes, two issues emerged as paramount. These two issues in particular influenced both nurses' practice and their effectiveness in influencing continuity of care. The first of these influences had to do with the high rate of turnover among nursing staff in the Nursing Station and related issues regarding nursing workload. The second influence was related to nurses' preparation to work within the community context and in the expanded and advanced nursing role that was expected of them. Compared to nurses working in mainstream health services in other Canadian communities, nurses working in the Nursing Station experienced inequitable working and social conditions as well as limited access to resources. Nurses worked at the margins of mainstream nursing practice, expected to function outside the traditional scope of nursing practice, yet often were without work-role preparation comparable to that of nurses prepared for out-of-scope practice in other settings, such as nurse practitioners. Thus, inequity and marginalization were reflected within nurses' preparation for practice, and exerted a strong influence on nurses' ability to engage with practice, posing particular issues and challenges for nurses. Moreover, nurses' preparedness for the role and the context of their practice directly influenced how effectively nurses managed workload issues, thus the interrelationship between the two factors also exerted a synergistic influence on continuity of care. 118 Being Too Busy: High Nursing Turnover and Working Short Staffed An underlying assumption of this research was that there would be a high turnover rate of nursing staff at the study site and that the Nursing Station would be short of nursing staff at least on an intermittent basis. This assumption was based on the literature (Kulig et al., 2003; Martin & Gregory, 1996; Minore et al., 2002; Vukic & Keddy, 2002), on my own clinical experience in similar sites, and on anecdotal evidence from nurses with similar experience. The findings of this study supported this assumption. There can be no question that the study site experienced a high rate of nursing staff turnover. Using the monthly staffing sheets I counted the number of different nurses who had worked in the Nursing Station in the two-year study period covered by the chart review. The total numbered 73 nurses. There had been a stable core of two to three nurses through most of that period, but the remaining nursing positions had been filled by a constant stream of nurses who came to work in the community on a temporary basis, for anywhere from a few days to several months. To illustrate the rate of turnover, there was a turnover of six nurses in the four weeks between my first and second visits to the study site to conduct data collection. From my perspective as a researcher, this was problematic in that I had to "orientate" six new nurses to my research, which involved finding time when the nurses were not busy to explain the study to them, to achieve "buy-in" and to obtain informed consents. Minore et al. (2002) described a similar experience during their research in remote First Nations communities in northern Ontario. Later, when a participant stated "I can understand why [patients] get frustrated when they come into the Nursing Station and it's another nurse they have to tell the same story to", I realized that my own sense of frustration with the high rate of nursing turnover reflected the frustration patients felt when continually faced with seeing unfamiliar health care 119 professionals. Patients in the study site potentially were faced with seeing a new and unfamiliar nurse at each health care visit. Thus, the high rate of nursing staff turnover had a direct influence on relational, or interpersonal continuity of care (Reid et al., 2002) in the study site. A l l participants were highly aware of the high turnover of nursing staff. One nurse stated, "changeover is huge, um, - huge. I don't really know how else to say that but, you know, there aren't a lot of us that have been around for any length of time". A physician told me that three years was a long time for a provider to remain in a community, while a nurse agreed that four years' conferred "old-timer" status. A nurse who had worked intermittently in northern communities for 19 years estimated the longest time she had spent in one community had been six or eight months, illustrating the relative transience of nurses working in the north. The Nursing Station was officially a nine-nurse station until the year prior to data collection for this study, when staffing was decreased by one position, leaving eight Nursing Station positions. At about the same time, the Band-transferred public health nursing position was created. It was commonly believed in the Nursing Station that the public health position had been created at the expense of the Nursing Station position. However, findings suggested that confusion existed among health professionals and para-professional staff in the study site as to the number of nurses that constituted "full" staffing for the community. This confusion was an important aspect of tension and conflict between nurses in the Nursing Station and public health. Regardless of the official quota of nurses, the Nursing Station was seldom fully staffed: "usually we're running at five or six nurses". Health professionals and administrators perceived a close association between the high turnover of nursing staff and the chronically 120 short staffing levels in the Nursing Station. Working short staffed implied that nurses' workload was increased, since patient demand for health care remained constant. Not only did each nurse have to see more patients during regular clinic hours (which might necessitate overtime work), but each nurse also carried a greater responsibility for taking after-hours call for emergency services. Several nurses estimated they worked between 50 and 60 hours a week. Being on-call during the night meant that nurses might be awakened several times during the night, or might be up all night attending to a patient, yet nurses were expected to work the following day despite having worked during the night. During the data collection period I observed that the nurse on-call literally never had an uninterrupted nights' sleep; "call" was very busy in the study site. The high turnover rate of nurses and being short-staffed contributed to nurses' perception that they were "too" busy. The increased workload that resulted from being short staffed was perceived as one reason for the high turnover rate among nurses, thus short staffing and "being too busy" contributed to the high turnover rate, and the high turnover rate in turn, exacerbated the staffing situation and created an increased nursing workload. Being too busy meant that nurses had to work harder to "catch up" with the work of providing patient care. As data from the chart review, observations and interviews showed, when nurses were too busy to keep up with patient care needs, patient care suffered and patients were at risk of "slipping through the cracks", or of experiencing fragmented care, particularly the dimensions of informational and management continuity of patient care. Being too busy and working to catch up were in part the process manifestations of the structural and organizational issues of being short-staffed and the high rate of staff turnover. These larger organizational issues shaped the context of the work environment and were therefore central 121 to the theme of nurses' being too busy. However, they were not the only issues that contributed to nurses' busyness. In common with health care providers in Nursing Stations generally, participants in this study referred to illness-oriented patient care as "treatment". Treatment-oriented services (e.g., Health Canada, 200Id) were differentiated from preventive health services and health promotion activities, which were referred to collectively as "public health" (e.g., Health Canada, 2001c). Treatment included primary care for common episodic health problems (e.g., "colds", headaches) and chronic health conditions (e.g., hypertension, diabetes) as well as for urgent and emergent problems (e.g., trauma, chest pain). Public health programs included maternal-newborn follow-up, well child assessments, immunization programs, school health programs, infectious disease reporting and follow-up, chronic health programs (e.g., diabetes education) and (to some extent) follow-up of chronic conditions, and prenatal education classes. Historically, as the only health care providers in the community, nurses employed in Nursing Station settings have been responsible for both the treatment and public health aspects of health services in remote aboriginal communities. "Community Health Nurse" continues to be the formal position title used by FNIHB and the territorial employers of Nursing Station nurses, reflecting the community health basis of the role. In the study site, the creation of a Band-transferred public health nursing position at about the same time as Nursing Station staffing had been decreased was the source of some confusion among nurses in the Nursing Station, as questions of their continuing responsibility for certain aspects of the public health role were unclear to nurses. However, role confusion related to the public health responsibilities of nurses working in the Nursing Station had also existed prior to the implementation of a Band-employed public health nurse. 122 When nurses were challenged to keep up with the nursing workload, it was the preventive and public health programs that suffered, as nurses' energies were necessarily taken up meeting the more pressing needs of patients that required urgent and emergent services. One nurse stated, "You get overwhelmed and the focus is more on the emergent stuff. It shouldn't be but what do you do when they're knocking on your door". For example, nurses had discontinued making maternal-newborn home visits several years prior to the study time, because in the words of one nurse, "there was no staff to go out and visit". The more urgent necessity of providing treatment services had become the priority that shaped nurses' work in the study site. Some new nurses coming into this setting, who were unfamiliar with the comprehensive, dualistic nature of Nursing Station work perceived the prioritizing of treatment services as diminishing the public health aspects of the role. This resulted in a devaluing of the public health aspects of the role among nurses employed in the Nursing Station and contributed to nurses' role confusion. One nurse participant who had worked in both public health and in the Nursing Station described this devaluing: "some of the [Nursing Station] nurses think that it's just fun and games kind of thing in the public health side, like there's no real work that has to be done". The data that suggested the existence of role confusion is explored further in Chapter Seven. Being too busy and focusing on urgent and emergent patient care at the expense of preventive health programs placed nurses in a perpetual loop of "falling behind" and needing to "catch up". Participants spoke of falling behind and needing to catch up on programs, on organization, on administrative work, and especially on immunizations: "It's very busy down here and we're very far behind so there's a lot of catch-up to do". Some Nursing Station nurses viewed the implementation of a public health nursing position as a panacea to the problem of catching-up, in that public health was to "take over" the immunization programs. 123 However, Nursing Station nurses were unclear what their continuing role was in providing immunizations in the new order; some nurses working in the Nursing Station perceived that immunization was no longer part of their responsibility. According to one informant, the public health nurse would also become responsible for the school health and the infectious disease programs (e.g., STD follow-up). Perhaps reflective of the work-related tensions that existed between some of the Nursing Station and public health nurses, it was unclear how the expectation could be realized that one public health nurse would be capable of taking over the time-consuming public health programs (all of which had fallen behind and therefore required catching up), that a complement of nine Nursing Station nurses had found "overwhelming". Both professional and para-professional participants identified additional reasons that they perceived helped to explain the high turnover - short staffing cycle. Explanations included a perceived lack of support within the larger organization: "typically [FNIHB] hasn't been an organization that's treated people very well", as well as a perceived lack of support within the community. A physician described this by saying that some people in the community "didn't understand" that the high demand for non-emergent services during the night was one reason there were often not enough nurses in the community. During an interview, a nurse participant also described lack of support within the community in terms of behaviour toward nurses that was perceived as abusive: I think every nurse has [experienced] situations of abuse whether it be, you know, physical, verbal or just emotional or whatever but I think it's increased up here. I think there's a lot of expectation from the community that we're here to, you know, to be at their beck and call and I think when you're here around the clock and sometimes you're forced to work the long hours and there's not a lot of support, you just kind of burn yourself out, so I think that's why there's such a high turnover up here. 124 With this statement, this participant linked non-support within the larger organization with a lack of support from the community, and suggested that nurses might be caught in the middle of the sort of political situation that at times may exist in communities between Band and federal government agencies. Nurses in this study not only perceived a lack of support from the local community as well as from their employer, but at times experienced stress secondary to the local political situation within the community (Gregory, 1992; O'Neil, 1986). Nurses' work in the Nursing Station was characterized by a heavy workload that included being on-call at night and often-disturbed sleep, perceived lack of support, stress secondary to working and living within an unfamiliar contextual and socio-political environment, and working short-staffed, which shifted the focus of care to urgent and emergent health services at the expense of preventive care. These were all factors that reflected the demands and challenges that confronted nurses in their work and contributed to the high rate of nursing staff turnover, which in turn exacerbated nurses' workload. Nurses' preparation to meet the challenges of being primary health care providers in such a demanding environment was another key factor that influenced both how "busy" nurses were and how nurses coped with the challenges of their work, which in turn influenced nurses' ability to provide continuity of patient care. Being Prepared: Nurses' Readiness for the Nursing Station Role Although nurse participants had begun their Nursing Station careers with a variety of different levels and types of previous experience and education, none of these participants believed her previous experience or education had been sufficient preparation for the Nursing Station role. Few perceived any part of their formal nursing education to have contributed specifically to their preparedness to work as a primary care provider, suggesting the primary 125 care role required knowledge and skills beyond what nurses learned in basic nursing education programs. Participants also described feeling unprepared and surprised by the context of isolation and poverty they found when they first arrived in a northern reserve community, and unprepared for the challenges of working within the cross-cultural context. Not only did nurses find themselves in a position they felt inadequately prepared for both clinically and culturally, at the same time they were expected to work in a more independent role and with less support than they had experienced in past nursing positions. Several nurse participants had previous experience as emergency room (ER) nurses and identified ER experience as being "definitely" needed in the Nursing Station. While ER experience undoubtedly provided nurses with many critical skills that were fundamental to Nursing Station work (Tarlier et al., 2003), other nurse and administrator participants identified that ER experience alone did not provide nurses with the full spectrum of competencies required for the broad-based, generalist role that Nursing Station work entailed (Vukic & Keddy, 2002). Nurses also needed clinical and procedural competencies related to the diagnosis and management of common episodic conditions (e.g., chest infections, colds, otitis media) and the long-term management and follow-up of common chronic and "multi-system" conditions such as diabetes and hypertension. Nurses needed to be competent providing care for patients of all ages, from newborns to elders, and providing routine prenatal and postpartum care, including preventive services such as obtaining and handling specimens for "pap" smears and giving immunizations. There was a perceived high-priority need for nurses to be prepared to cope with emergency situations that occurred relatively infrequently, yet paradoxically, as one nurse pointed out, the "bread and butter" day-to-day practice of nurses involved providing basic primary care, patient teaching, preventive services and health promotion for patients of all ages. Importantly, while children and 126 childbearing women were two patient groups that comprised a significant part of nurses' day-to-day practice in the study site, no nurse participant in this study had come to the Nursing Station role with previous experience or expertise working with either of these two patient populations. Nurse participants described how they had started out "green" and began to develop primary care and PHC competencies once they began work in the Nursing Station. The first strategy that was available to help nurses develop these competencies was the three-week orientation that FNIHB provided. The first two weeks of the orientation were conducted in a classroom setting in Winnipeg, where new recruits focused on learning about the structure and organization of FNIHB, the routine paperwork, assessment and diagnostic skills (i.e., history taking and physical examination consistent with a medical model of practice), specific skills such as suturing, and the health programs that were routinely offered in the Nursing Stations, for example, chronic health programs, prenatal care and diabetic teaching. This was followed by a week to a week and one half of hands-on practice in a Nursing Station under the tutelage and supervision of a nurse educator. After this, new recruits were on their own, sometimes "buddied" with a resource nurse, or with a second nurse on-call as "back up" for after-hours emergency calls. Orientation included an introduction to the public health aspects of the Nursing Station role, that is, to preventive health and health promotion. Nurses were oriented to mandatory public health programs that included the immunization programs, well child clinic and women's health. In contrast to the ambiguity expressed by some of the more experienced nurse participants in this study regarding the role of the Nursing Station nurses in providing immunizations, well child clinics and other "public health" programming, nurses who had received the orientation were clear that these responsibilities had been part of their 127 orientation to the role and that these had been clearly presented as being part of the responsibility and expectation of nurses working in the Nursing Station. While those nurses who had been through the orientation viewed it as having been a valuable experience, they also perceived that it fell short of adequately preparing them for either clinical practice in the expanded and advanced role or the reality of living and working within the context of aboriginal health in a remote community. One participant with several years' previous acute care nursing experience viewed the orientation as having not prepared her at all for the reality of practice: "I didn't really know what was expected of me until I got up here and saw, oh my god, this is what I've gotta do - it freaked me out". Other participants identified specific aspects of the orientation that they viewed as not conducive to learning: "It was really difficult learning things at the beginning of those two weeks and then expecting to apply them your first week out. You forgot what you learned at the beginning of those two weeks: it was information overload". The second formal strategy that supported nurses' development of primary health care competencies were the Primary Care (PC) Skills course and the Community Health Nurse (CHN) upgrading course that FNIHB has traditionally offered to employees. All of the nurses in this study who had (or had held in the past) a permanent fulltime position with FNIHB had been sponsored to a PC Skills course. The PC Skills course was a three-month course that included classroom study and a clinical practicum under the supervision of a physician preceptor. Nurses generally were required to work for some length of time in a Nursing Station prior to being sponsored to take the PC Skills course. However, one nurse believed that it would have been useful to take the course before starting to work in the Nursing Station, a statement that was consistent with other participants' views related to the inadequate preparation provided by the orientation. Nurses who had taken the PC Skills 128 course perceived that it had been key to helping them develop or enhance their primary care competencies: "the thing I rely on the most is having really strong assessment skills, and that was drilled into me from Primary Care Skills". Another nurse stated, "[PC Skills] pertains exactly to what we're doing, everything that we studied there, we do here. It was an excellent course". Notably, this participant's statement overlooked the fact that the focus of the PC Skills course was primary care, which was just one aspect of the broader primary health care role expected of nurses. The other important aspect of the PHC role was the community health nursing aspect. Nurses prepared with a baccalaureate degree were assumed to have developed community health nursing (CHN) competencies during their formal educational program. Conversely, while degree-prepared nurses may have had more exposure to public health nursing during their basic nursing education than diploma-prepared nurses were likely to have had, this did not guarantee that they had had sufficient exposure to develop community health competencies. For example, although degree-prepared nurses in this study perceived that exposure to community health nursing during basic nursing education provided a helpful basis for the further development of competencies in community health, they did not perceive themselves as being fully competent in the role. An administrator described degree-prepared but inexperienced nurses who had previously come to work in the community as public health nurses, saying "they up and left because they couldn't cope". Nurses without a degree had traditionally been offered the FNIHB C H N upgrading course, which provided formal education related to the "public" and community health expectations of the Nursing Station role, as well as to community development. The C H N course also included content on "aboriginal cultural" and working with families. Of the participants in this study who worked fulltime for FNIHB, or who had worked in a fulltime 129 position in the past, all had either taken the C H N upgrade course or were degree-prepared nurses. Being prepared for Nursing Station work included understanding the socio-political context of aboriginal health and being familiar with the broader community context, including issues related to the social conditions that prevailed in the community, such as poverty and the lack of running water. Interestingly, each of the First Nations health care professionals and administrators who participated in this study clearly identified the need for non-aboriginal health care professionals to be better prepared to work within the context, culture and enculturated social conditions of a remote aboriginal community. These First Nations participants believed that non-aboriginal nurses required more education about northern aboriginal communities, in particular, about how contextual issues affected health and about the differing value systems. One First Nations nurse described how the lack of contextual understanding influenced non-aboriginal nurses' understanding of a different value system: "People don't eat properly only because they're on a budget, on welfare, and there's no running water so therefore hygiene is a factor; there's just different values. So they [i.e., non-aboriginal nurses] have to try and understand the native culture". This statement demonstrated the propensity among both First Nations and non-First Nations nurses to conflate aboriginal culture with enculturated social conditions. However, in making the link between the need to understand contextual differences in order to support thinking critically about the links between the social determinants of health, differing values, and health conditions in the community, this participant implied that non-aboriginal nurses required knowledge related to the broader socio-historical context of aboriginal health (Anderson et al.; 2003, Browne, 2005; Papps ^Ramsden, 1996; Smye & Browne 2002), and not just more information about aboriginal culture per se. 130 Each of the nurses who participated in this study believed she had been unprepared and relatively unsupported when first starting out in the Nursing Station role. Nurses who had subsequently committed to fulltime employment with FNIHB had been supported to develop primary care and PHC competencies through being sponsored to attend the PC Skills course and in a few cases, the C H N upgrade course. Notably, the majority of nurses who worked in the Nursing Station during the study period were employed by FNIHB on a relief, or casual basis, and one nurse was employed by a nursing agency that contracted short-term relief nurses to FNIHB, therefore the number of participants in this study who had attended either the PC Skills course or the C H N upgrade course was relatively small. One participant described an exemplar experience as a new employee. This participant had elected to do a final clinical practicum in her baccalaureate nursing program in a FNIHB Nursing Station. Following graduation, she committed to working fulltime with FNIHB. She was provided with an extended orientation that included spending three months in a small hospital to develop acute care skills. On her return to the Nursing Station she was "buddied" with a supportive, more experienced nurse until taking the PC Skills course. This participant described the process of orientation and developing primary care and primary health care competencies in positive terms, in contrast to some of the participants who had been less well supported during the process, and she remained highly positive about the Nursing Station role. Nurses who fell between the cracks as far as being supported to develop competencies in primary care and PHC included those nurses who did not survive the interval between orientation and the PC Skills course: Being unprepared created anxiety and these nurses subsequently left the Nursing Station for presumably less stressful employment, thus contributing to the high turnover rate among nurses. Nurses who worked relief rather 131 than being fulltime FNIHB employees, including relief nurses who worked through an agency, were also unlikely to have taken the PC Skills course, although they received orientation. In return for being sponsored to the PC Skills course, relief nurses were expected to provide one year of service within one year of completing the course.24 However, relief nurses were in most cases also employed by another employer in addition to FNIHB, and were unable to commit either to taking the three months' long course or to being available for the return-of-service commitment. While my observations suggested that relief nurses were often inadequately prepared in terms of the core knowledge and skills required to work in the Nursing Station, nurse participants tended to address the issue of preparation for relief nurses in terms of experience rather than competency. However, of the relief nurse participants in this study, only one had significant previous experience working in Nursing Stations, almost all of it as a relief nurse. This participant was a diploma-prepared nurse who had worked as a relief nurse in Nursing Stations for more than 10 years yet had never had the opportunity to take either the PC Skills course or the C H N upgrade course. The following excerpt from an interview with this nurse revealed that she had not had the advantage of additional formal education related to immunization but had learned "on the job": R: Where do nurses get those public health skills? P: On the job. R: On the job? P: Yeah, I mean they, they give you a very intense exam, you know, and, uh... R: The immunization exam? P: Yeah. Discussing nursing competencies related to providing immunizations, she stated: "I don't think immunizations are as hard as everybody makes it out to be... it's basically giving a 2 4 Under special circumstances relief nurses could arrange to work back the 12 months over a two-year time period. 132 needle at the right time". This comment reflected understanding immunization as primarily a psychomotor skill. Formal advanced education related to epidemiology and the theoretical basis of immunization is necessary to appreciate the complexity involved in providing an immunization program. This excerpt also illustrated how the complexity of the public or community health aspects of the Nursing Station role were sometimes unintentionally devalued or minimized by some nurse participants. Notably, these participants tended to be non-baccalaureate-prepared nurses with relatively more previous experience in acute care nursing roles. To be clear, relief nurses provided a very necessary service in the study site, in that the Nursing Station relied heavily on relief nurses both to provide relief when fulltime nurses went on vacation or away to courses (e.g., the PC Skills course), and to fill in nursing positions that weren't fdled by fulltime FNIHB nurses. With the help of relief nurses, the Nursing Station operated short-staffed much of the time; without relief nurses, the operation of the Nursing Station would have been even more compromised. Moreover, participants believed that Nursing Stations in remote communities would always be dependent on relief and short-term staffing solutions. Thus, while fulltime nurses perceived the work of relief nurses as a group as not contributing to the overall quality of patient care, and the reality of having over 70 different nurses rotating through the Nursing Station in a two-year span clearly could be construed as having implications for at least relational continuity of care, the Nursing Station was nevertheless forced to rely on relief nurses to maintain even semi-adequate staffing levels. Considering the turnover of staff and the number of relief nurses that revolved through the study site, having relief staff that were adequately prepared to work in the expanded Nursing Station role would appear to offer obvious benefits. However, as fulltime participants suggested, and my field observations substantiated, relief nurses were 133 often not adequately prepared for the role, a phenomenon also noted by researchers Minore et al. (2005) in their recent study of health care in remote First Nations communities. An important paradox existed between most relief nurses' relative lack of preparation for the Nursing Station role and the extreme degree to which the study site depended on relief nurses. It was also paradoxical that working for short periods of time in different Nursing Stations was viewed as posing challenges that were unique to relief nurses, that were different to the challenges faced by nurses who worked in one site, yet relief nurses were more likely to slip through the cracks and to be less adequately prepared than were the fulltime nurses. The Interface of Being Too Busy and Being Unprepared: Inefficiency and Ineffectiveness Observational data collected during nurse-patient encounters and as general observations of the day-to-day work of nurses, as well as interview data with nurses, administrators and physicians suggested that when nurses were inadequately prepared for the work they were expected to do, it took longer for the work to be accomplished. For example, when nurses were not familiar with the physical layout or the routines of the work environment, it took longer for them to locate needed equipment or supplies. When nurses were unfamiliar with clinic routines, patient visits were delayed while the nurse read a guideline or found someone to ask. Finding an available resource person, and also searching through patient charts for needed information, were structural issues that contributed to extending the length of time it took a nurse to complete a patient visit. However, a nurse would also have to find a resource person (i.e., another nurse or a physician) to refer to when unsure of a diagnosis or how to best manage a particular patient condition. The work of assessing, diagnosing and managing patient health conditions was often unnecessarily drawn-134 out when nurses lacked adequate primary care knowledge and / or procedural skills to do this work efficiently. Inadequate preparation as a primary care provider not only compromised nurses' ability to provide efficient and effective informational and management continuity of care but also contributed to nurses' perception that they were "too busy". Interview data indicated that nurses perceived that one reason they were too busy was because they sometimes saw as many as 13 patients in a day: "it's a busy community, like, we would see 13 people a day". Notably, in the study site, as in Nursing Station settings generally, nurses might necessarily spend more time to complete a patient visit than would primary care providers in other settings because the nurse in the Nursing Station carried out all aspects of service provision herself. For example, the nurse dispensed medications because there was no pharmacist and collected lab specimens (such as bloodwork) because there was no lab technologist. This suggested that one aspect of nurses' busyness was related to the lack of personnel resources. However, this observation notwithstanding, in other primary care settings, including other Nursing Stations (based on personal clinical experience and anecdotal evidence from experienced Nursing Station nurses), nurses frequently see many more than 13 patients in an eight-hour workday. While not wishing to imply that patient volume equates to quality of care, 13 patients in even a seven-hour workday (accounting for breaks) worked out to more than 30 minutes per patient visit. This compared to one Ontario-based study of physician-provided primary care where physicians spent on average 16 minutes per patient (Boone, Stewart, Kennard, & Guimond, 2003), and a study of nurse practitioners in Ontario, Saskatchewan and Newfoundland where 45% of patient visits were completed in less than 15 minutes and 84% of visits were completed in less than 30 minutes (Way, Simms, & Greene, 2001). These findings suggested that more adequate clinical preparation might enable nurses to deliver primary care more efficiently and 135 effectively in the study site, thereby enabling them to cope more effectively with the volume of work, and consequently decreasing the perception that nurses were "too busy". To illustrate how inadequate clinical preparation for the primary care role contributed to inefficient patient care I present a brief case scenario based on an observation of a nurse-patient interaction: Case Scenario 5.1 I observed a nurse examine a five-month old infant who had been brought to the clinic because of a cough. The nurse took a history from the infant's mother that suggested a mild upper respiratory infection (URTI). Examination of the infant revealed no significant findings other than a mildly elevated temperature (i.e., 37.8°C, consistent with a mild URTI), and no respiratory signs. Without establishing a clear working diagnosis, the nurse then gave the infant an aerosol "mask" treatment with oxygen.2 5 Not only was this treatment not indicated, it was also inappropriate, as current clinical recommendations do not support the use of aerosol masks for URTI, or of masks given with oxygen. The nurse then dispensed saline nasal drops to the mother. This patient visit took one hour to complete, whereas I estimated that 15 to 20 minutes would have represented a more reasonable length of time, given a primary care provider who was adequately prepared to conduct a pediatric assessment, make an accurate and clear diagnosis and offer appropriate management of what was essentially a well infant with a common "cold". Where the relief nurse in this scenario took an hour to complete a patient visit, a better-prepared provider might feasibly have completed three or four patient visits. Since on most days there was a continuous backlog of patients waiting to be seen, lack of clinical preparation for the Nursing Station role contributed to nurses' being "too busy". Moreover, when patient participants voiced dissatisfaction with health services in the study site, it was most often related to clinic waiting times, which were often lengthy even for patients who had made an appointment. 2 5 Although this was the only instance when I observed a nurse offer an aerosol treatment in the study site, my clinical practice experience in similar sites suggests that it is not uncommon for nurses to offer a non-medicated aerosol treatment by "mask", in the belief that this "loosens up" airway congestion; however, this practice is not supported by the clinical guidelines or an evidence base. 136 By providing treatment that was not indicated and was thus inappropriate, the message that was conveyed to the infant's mother in this scenario was that this essentially well infant with a common cold required nursing intervention, thereby reinforcing the same dependency on nurses that other participants in this study viewed as problematic, and that contributed to the volume of patient visits and after-hours calls for non-emergent concerns, and that therefore also contributed to nurses' being too busy. Moreover, while the nurse did determine during the visit that the infant was not up-to-date with immunizations, she did not take this opportunity to immunize the infant, necessitating the need for a second visit at a later date to obtain the immunization. Not providing complete care at one visit necessitated a subsequent visit, which contributed to the volume of patient visits and nurses' busyness. Moreover, as the low infant immunization rates found in the chart review arm of this study suggested (see Chapter Six), infants were "slipping through the cracks" and were not receiving immunizations, sometimes despite multiple visits to health care professionals. This finding held clear implications for management continuity of care. The outcome of this scenario may have been different if the nurse had been better prepared to work in the broader PHC role; that is, i f she had demonstrated competencies related to giving immunizations. Interestingly, the nurse participant in this scenario appeared unaware of the inconsistency that was implied by taking an hour to complete this particular "sick visit" yet stating there had been insufficient time to conduct a "well child assessment" and provide the needed immunization, further suggesting the need for nurses to possess competencies related to pediatric assessment and immunization. 137 Summary Both the broader community context and the structure of health service delivery in the study site were characterized by inequitable social conditions and a lack of access to resources and support services, relative to communities and health services in mainstream Canadian settings. The community and health service system were further challenged by remoteness and by the historical and socio-political context of aboriginal health. Living as they did on both the margins of the local community and working on the margins of mainstream health services, nurses in the study site also experienced inequitable social and work conditions relative to nurses in other health care settings. Nurses' work was demanding, and contextual and structural factors influenced nurses' practice and their ability to influence continuity of patient care and health outcomes. Two influences in particular emerged as key to shaping nurses' work. One was nursing workload issues: nurses were "too busy". A number of factors contributed to nurses' being too busy, including the high rate of turnover among nurses, working short-staffed and lack of resources and supports. The second influence was nurses' preparation to work in the expanded and advanced practice Nursing Station role. Nurses often arrived in the study site with little formal education or experience specific to prepare them to work in either the primary care or PHC roles, despite the finding that new nurses received a three-week orientation from FNIHB. Nurses who worked fulltime as FNIHB employees were further supported by eventually being sponsored to the three-month Primary Care Skills course, and diploma-prepared nurses had sometimes been sponsored to the C H N upgrading course. However, despite the particular challenges that relief nurses faced, relief nurses fell through the cracks of preparation and tended to be the least well prepared for the PHC role, particularly in regard to the primary care and community nursing aspects of the role. 138 Nurses' lack of preparation was closely linked with nurses' being too busy, in that lack of adequate preparation for the Nursing Station role contributed to inefficient and ineffective practice, which exacerbated nurses' perception that they were too busy. Thus, the structure and context of nurses' practice contributed to the challenges of providing continuity of patient care, in that nurses were often too busy or lacked adequate preparation to provide health services that supported relational, informational and managerial continuity of care. The themes related to the context and structure of nurses' challenging work conditions, and nurses' working at the margins of mainstream nursing practice, offer a backdrop to the presentation of findings related to health outcomes in Chapter Six. 139 C H A P T E R SIX D A T A F R O M C H A R T REVIEW: A WINDOW INTO H E A L T H O U T C O M E S A N D CONTINUITY O F C A R E Introduction A primary hypothesis underlying and motivating this research was that residents of some remote aboriginal communities experience fragmented health care and poorer health outcomes relative to other Canadians. This question was based in the national health status indicators data (i.e., INAC, 2003; CIHI, 2004a) and the literature presented in Chapter Three that described aboriginal health and health outcomes at the national level. The purpose of collecting health outcomes data that were specific to the study site was to determine whether health outcomes in this community were consistent with the national aboriginal health indicator data. Exploring health outcomes specific to the study site also offered an anchor for examining the contextual, structural and process issues in the site that influenced continuity of patient care. In this chapter I first present data from patient chart review substantiating that study site health outcomes on the selected health indicator conditions were consistent with national indicators of aboriginal health. These findings suggested that health services in the study site are challenged by fragmentation. I then present findings from the contextual notes data collected from chart reviews (as described on p. 89), to situate the health outcomes data contextually, and to link health outcomes to continuity of patient care. The analysis of the contextual notes data in light of the health outcomes exposed some of the sources of fragmentation. The interpretation presented here is organized around the three dimensions of continuity of care identified by Reid et al. (2002): (a) relational, (b) informational and (c) managerial. To review the definitions of these dimensions, informational continuity refers to 140 the transfer of information and the accumulated knowledge a health provider has of a patient; relational continuity refers to both on-going patient-provider relationships and consistency of personnel; and management continuity depends on coordination and consistency in the approach to treatment management. Using this framework to explore fragmentation of health services exposed the gaps in service that were reflective of discontinuity of patient care. Exploring Health Outcomes in the Study Site Data related to clinical health outcomes were based on the two pre-selected health indicator conditions: (a) prenatal care and (b) care of the infant through the first year. Data were extracted from a manual review of 128 randomly selected patient charts, which represented 65 mother-infant dyads (including two maternal charts for which the corresponding infant's chart could not be located), or approximately 22% of the births that occurred to community members during the defined study period from September 1, 2001 through August 31, 2004 (as described in detail in the methods section). The data related to health outcomes were collected and analyzed (also as described in the methods section). The findings described in this chapter are organized first in terms of describing the sample of prenatal women and their infants, outcomes related to perinatal care, and outcomes related to infant care. Contextual notes data are then presented, using three patient case scenarios to illuminate the context and process of patient care and to link health outcomes with continuity of care. Description of the Sample Population A l l of the women and infants (n = 128) that comprised the sample for chart review were "treaty" members of a local First Nations Band and were deemed residents of the study site community. The charts of 65 infants and their mothers were randomly selected for review. Each of the women (n = 65) had an "active" patient chart in the nursing station and 141 had received some or all of her prenatal care at the nursing station. Similarly, each of the infants had an active patient chart in the nursing station and had received some or all routine infant care at the nursing station, with two exceptions. One infant's chart could not be located and another infant had remained hospitalized in Winnipeg during the entire first year following birth, thus the charts of 63 infants were reviewed. Of this sample about half of the infants were male (n = 33) and half were female (n = 30). The mean age of women in the sample was 22.5 years (n = 65). Ages ranged from 14 to 39 years, with a median age of 22 years and a mode of 17 years. Seventy two per cent of the sample was aged 25 years and younger. Only 6% of the sample was aged 30 years and older. Almost 31% of the sample was aged 19 years and younger (i.e., were adolescents). The rate of births to adolescent women in the sample was high compared to approximately 6% of all live births to adolescents in the Canadian population overall (based on 1997 data, Health Canada, 2000a) and approximately 22% in the Canadian First Nations population (based on 1999 data, Health Canada, 2003). Almost 28% of the women were primigravid. Gravida status in the current pregnancy ranged between one (i.e., nulliparous) and 12; the average was 3.3 and the median number of pregnancies was two. Para status in the current pregnancy ranged between none and eight; the average and the median were 1.8 previous live births, which is less than the estimated fertility rate of 2.7 among on and off reserve aboriginal women, but greater than the rate of 1.5 for Canadian women overall (Health Canada, 2003). Six percent of women were "grand multips", defining grand multiparity as five or more previous viable births (Humphrey, 2003). The charts of five women were reviewed twice, for two different pregnancies during the study period. Four of the pregnancies included in the chart review were twin pregnancies; in these cases, only the chart of the single infant that had been randomly selected was 142 included in the infant chart review. Data describing age and gravida status are presented in Table 2. Table 2 Maternal Age and Gjravida Status Variable Maternal age (N = 65) Percent Range 14 -39 years Median 22 14 - 19 years 20 - 25 years 26 - 29 years 3 0 - 3 9 years 30.8% 41.5% 21.5 % 6.2% Gravida status (N = 65) 1-12 pregnancies 2 Primiparous Para 2 - 4 27.7% 66.1% 6.2% Perinatal Health Outcomes Descriptive statistics related to perinatal health were collected on four specific clinical health outcomes: (a) gestational age at delivery was between 38 and 42 weeks, (b) infant birth weight was between 2500 and 3999 grams, (c) whether the delivery was free of complications that might reasonably be directly attributed to prenatal health, and (d) a visit for routine postpartum follow-up care had occurred between six and eight weeks following delivery. These data were collected nominally; that is, the criterion for each outcome was met or not met. The number of outcomes that were satisfactorily achieved by each woman reflected the outcome of each pregnancy in general terms. Almost three quarters (73%) of the women in this sample achieved satisfactory perinatal outcomes on 50% or more of the criteria. Results are tabulated in Table 3. 143 Table 3 Perinatal Health Clinical Outcomes Clinical Health Outcome Frequency Percent Gestational age at delivery between 38-42 weeks 50/65 75% Infant birth weight between 2500 - 3999 grams 49/65 75% Delivery free of complications 39/65 60% Postpartum visit made 10/65 15% Number of Perinatal Outcomes Achieved by Each Woman: 4 Outcomes achieved (100%) 4/65 6% 3 Outcomes achieved (75%) 32/65 49% 2 Outcomes achieved (50%) 12/65 19% 1 Outcomes achieved (25%) 13/65 20% 0 Outcomes achieved (0%) 4/65 6% Data were also collected on gestational age at delivery and infant birth weight. Gestational age at delivery ranged from 34 to 42 weeks, with a mean of 38.7 weeks (n = 65; SD = 1.98). The median and the mode were each 39 weeks. Premature births were defined as occurring before 37 completed weeks' gestation (Baskett, 1991; Wenman, Joffres, Tatryn, et al., 2004). Premature births accounted for 22% of births in the sample, compared to 7.1 per 100 live births in Canada overall in 1997 (Health Canada, 2000a). More than 75% of births in the sample occurred between 37 and 41 completed weeks of gestation. Births past 42 weeks' gestation accounted for 3% of births, compared to the Canadian national rate of 1.8% (Health Canada). However, it is important to note that planned interventions such as induction of labour or caesarean section were routinely implemented past 42 weeks' gestation. It should also be noted that since sample selection was dependent on recorded live births, any premature or post-dates births that resulted in fetal death would not have been included in the chart review; therefore, the number of pregnancies that terminated in premature or post-dates delivery may actually be higher than indicated by this data. Infant birth weights ranged from 1347 to 4936 grams, with a mean of 3498 grams (SD = 642) and a median of 3577 grams. Six percent of birth weights in this sample were low 144 birth weight (i.e., less than 2500 grams), compared to 4.7% of all First Nations on-reserve births in 2000 (FINIHB, 2005). Nineteen percent of birth weights in the sample were high birth weight (i.e., greater than 3999 grams), compared to 20.4% of First Nations births (both on and off reserve) nationally (Health Canada, 2003). Thus, while 75% of birth weights in the sample fell within the desirable range of 2500 to 3999 grams (i.e., the range associated with better long-term health outcomes [Health Canada, 2003]), 25% of birth weights fell outside of the desirable range. The descriptive data related to gestational age and birth weight are presented in Table 4. Table 4 Gestational Age at Birth and Birth Weight Variable Gestational age at birth (N = 65) < 37 completed weeks 3 8 - 4 1 completed weeks > 41 completed weeks Range 3 4 - 4 2 Median 39 SD Frequency Percent 1.98 14/65 21.5% 49/65 75.4%o 2/65 3.1% Birth weight (N = 65) < 2500 grams 2500 - 3999 grams > 3999 grams 1347.4936 3577 642 4/65 49/65 12/65 6% 75% 19% Only 10 of 65 women (15%) in this sample experienced routine postpartum care with a nurse or physician. The charts of 55 of 65 women (85%) offered no evidence that the women had attended a postpartum visit. This compares to 85% of women who did attend a postpartum visit in a large (n = 9,953) study of American women using data from the 1988 National Maternal and Infant Health Survey (Lu & Prentice, 2002). The Prenatal Data Abstraction Protocol that guided data collection during chart review (see Appendix E) defined the occurrences during pregnancy, labour and delivery that constituted "complications" for the purposes of data collection. Although data regarding complications of pregnancy, labour and delivery were collected at the nominal level, the 145 contextual notes provided additional data to indicate the type and extent of complications. Review of the descriptive data presented in Table 5 revealed that pre-existing health conditions (e.g., diabetes, hypertension) and conditions that developed during pregnancy (e.g., pregnancy-induced hypertension [PIH], gestational diabetes mellitus [GDM], macrosomia) were frequently associated with premature labour and delivery and with interventions such as induction of labour and caesarean section. Interventions in turn carried a higher degree of risk of poor outcome, such as wound infection and anesthetic complications. Twenty-nine out of the 65 women in this sample (44.6%) experienced a complication of pregnancy, labour or delivery. Two of the patients included in this chart review experienced major complications of pregnancy, labour and delivery; one woman underwent hysterectomy secondary to uncontrollable postpartum hemorrhage and another woman experienced complications during a caesarean section that resulted in ongoing loss of function.26 Caesarean section was included as a complication only if there was reasonable evidence in the patient chart to suggest the procedure was due to a complication of pregnancy or labour; for instance, elective c-sections or c-sections for certain malpresentations (e.g., breech, transverse lie) were not included as complications. Caesarean sections for macrosomia or fetal distress were included as a complication. Table 5 suggests that 11 out of 65 women underwent c-section, or just less than 17%, which is low compared to the figure of 22.5% estimated for c-sections nationally (CIHI, 2004b). However, by including the c-sections that were not counted as complications (i.e., planned c-sections), the c-section rate in this sample increased to 24.6%, or slightly higher than the national rate. Unspecified to safeguard patient anonymity. 146 Table 5 Complications of Pregnancy, Labour and Delivery Description of Complication of Pregnancy, Labour and Delivery Age Para Postpartum hemorrhage, anemia postpartum 22 * Multiple gestation, premature delivery at 36 weeks 20 * Multiple gestation, premature labour, c-section at 34 weeks 20 2 Postpartum hemorrhage 23 * D M , macrosomia, 2 degree tear, hemoglobin 55 postpartum 21 0 Type I Diabetes, multiple gestation, c-section at 34 weeks 24 * Uncontrolled HTN, Type 1 diabetes, macrosomia, c-section at 37 weeks 26 2 Persistent anemia, low birth weight infant, c-section for failure to progress 19 0 Anemia, HTN, induced at 37 weeks 17 0 Renal failure, multiple gestation, HTN > PIH, SVD at 37 weeks 27 4 Persistent UTI's, septic infant, NICU x < 1 week 17 1 Anemia, elevated BP, macrosomia, c-section at 41 weeks for failure to 24 1 progress Grand multip, anemia, G D M , induced 2° G D M , postpartum hemorrhage, 27 8 infection Uncontrollable postpartum hemorrhage > hysterectomy, anemia, wound 18 1 infection Untreated UTI > premature delivery at 34 weeks in nursing station 17 1 C-section at 34 weeks for PIH and G D M > infant NICU x < 1 week 34 4 Primip SVD in nursing station at 37 weeks (39 weeks per fetal assessment) 17 0 Anemia, D M , macrosomia, induced at 41 (42) weeks 2° oligohydramnios, c- 27 0 section for failure to progress, CPD PIH, induced at 39 weeks 18 1 Anemia, PROM at 36 weeks, induction 29 1 G D M , elevated BP, c-section (breech) 18 1 Type 1 diabetes, delivery at 35 weeks, infant had significant congenital 30 3 abnormalities D M , premature delivery at 36 weeks, c-section 23 0 Anemia, PROM, premature delivery at 37 weeks, medevac 20 0 Anemia, D M , elevated BP 29 3 Anemia, spotting at 34 weeks, labour at 41 weeks, c-section for failure to 23 3 progress, fetal distress > anesthetic complications (GA) > ongoing maternal sequelae Postpartum hemorrhage 17 0 Persistent UTI's, PIH, pre-eclampsia, failed induction, C-section at 36 weeks 20 0 Grand multip, renal failure, anemia, premature delivery at 36 weeks 29 6 * Missing data Complications of pregnancy were often found in conjunction with pre-existing conditions or illnesses such as diabetes or renal disease. Nine out of 65 women (13.8%) in this sample had a diagnosis of diabetes during pregnancy. Notably, the contextual notes and 147 observational data demonstrated inconsistencies both in implementing clinical guidelines for gestational diabetic screening and for following up with further diabetic testing postpartum, as well as with documentation of testing results; therefore it is likely that rates of diabetes in the sample group were actually higher than Table 5 suggests. The Canadian Diabetes Association (CDA) (2003) estimates the prevalence of gestational diabetes to be between 8% and 18% in the Canadian aboriginal population. Similarly, nine out 65 women (13.8%) had been identified in pregnancy as having an elevated blood pressure, either pre-existing or pregnancy-induced. Anemia during pregnancy was not included as a complication unless it appeared to be persistent or profound. Thus the prevalence of anemia in the sample group was higher than indicated in Table 5. Infant Health Outcomes Descriptive statistics related to outcomes of infant care through the first year were collected on three clinical health outcomes: (a) whether routine immunizations were up-to-date at age 12 months (or at the time of the chart review if an infant was aged less than 12 months at the time of the review), (b) whether infant weight gain was satisfactory at age 12 months (or at the most recent documented weight i f an infant was aged less than 12 months at the time of the chart review), and (c) whether the infant's hemoglobin had been documented and was within normal parameters between six and 12 months of age. These data were collected nominally; that is, the criteria for each outcome was met or not met. The number of outcomes that were satisfactorily achieved by each infant reflected health outcomes in infancy in general terms. At age 2 months, 36.5% of infants were fully immunized. Immunization rates then dropped for the next two immunizations: fewer than 12% of infants were fully immunized at age 4 months, and only 5% of infants were fully immunized at age 6 months. While 148 immunization rates increased again by age 12 months, over two thirds of infants were not fully immunized at the time of their first birthday. Infant health outcomes and immunization rates are summarized in Table 6. Table 6 Infant Clinical Health Outcomes Clinical Health Outcome Weight gain satisfactory at 12 months (or for age) Hemoglobin within normal limits (6-12 months) Immunizations up-to-date for age 2 7 Frequency 37/63 23/58 24/63 Percent 59% 40% 38% Number of Infant Health Outcomes Achieved by Each Infant: 3 Outcomes achieved (100%) 7/63 11% 2 Outcomes achieved (67%) 21/63 33% 1 Outcomes achieved (33%) 21/63 33% 0 Outcomes achieved (0%) 14/63 22% Infant Immunization Rates: Immunized Unimmunized 2 month 4 month 6 month 12 month 23/63 7/60 3/59 17/51 36.5% 11.7% 5.1% 33.3% 63.5% 88.3% 94.9% 66.7% The immunization rate of 33.3% at age one year among infants in the study site compared to a rate of 57.3% among all on-reserve First Nations infants in Manitoba between 1995 and 1998, 62% of all First Nations infants in Manitoba (i.e., on and off-reserve), and 88.7%o of all other Manitoba infants (Manitoba Centre for Health Policy [MCHP], 2002). Thus immunization rates among one-year-old infants in the study site were lower than NB: The seeming discrepancy in the data presented in Table 6 related to "Immunizations up-to-date for age" (i.e., 38%) and "Infant immunization rates at 12 months" (i.e., 33.3%), presented at the bottom of Table 6, is explained by the fact that a few infants in the second round of sampling had not yet reached their first birthday at the time of the study. In these cases, infants were deemed to have an up-to-date immunization status if their immunizations were up-to-date for their age at the time of the chart review (also, see Appendix E). 149 immunization rates for other First Nations infants in Manitoba and lower than the overall Manitoba infant immunization rate. Of the 58 infants who had passed six months of age at the time of chart review and who ought to have received a hemoglobin screen in accordance with the FNIHB clinical guidelines, 23 (40%) had a documented hemoglobin value that was within normal parameters for the infant's age at the time of screening (i.e., 110-150 g/L at age 12 months, or 105-145g/L if screened at age 6 months only). Thus, 60% of infants were either screened and identified as having low hemoglobin, or had not been screened and therefore their hemoglobin value was unknown. Data extrapolated from the contextual notes suggested that low hemoglobin was prevalent in this sample of infants. Of the 58 infants who were age-eligible for screening, 21 (36%>) had not been screened by age 12 months. Of the 37 infants who had been screened, 14 infants (38%) had a low hemoglobin value. The lowest infant hemoglobin value noted was 82 g/L. Several values were noted to be less than 100 g/L. In a study of the prevalence of anemia among Cree First Nations infants, Willows, Morel, and Gray-Donald (2000) found that 31% of infants had a hemoglobin value of less than 110 g/L at age nine months. The value of 110 g/L is the 2.5 percentile for healthy infants, while the value of 100 g/L "corresponds with at least moderately severe anemia" (Willows et a l , p. 324). Willows et al. noted that the prevalence of hemoglobin value lower than 110 g/L in Canadian non-aboriginal infants is 8%. Linking Health Outcomes to Context and Continuity of Care The clinical health outcomes findings described in this chapter reflected less than ideal clinical health outcomes in maternal and infant health. In remote aboriginal communities in Canada the responsible federal health authorities recognized early on the 150 need to offer a strong program of maternal-infant health care. The hiring preference for early outpost nurses to have midwifery training, therefore proficiency in maternal-infant health, attests to this recognition. The responsibility for providing maternal-infant health care in remote aboriginal communities has traditionally fallen to nurses, as the main primary care providers. The emphasis on nurses having midwifery or strong maternal-infant nursing skills has decreased since the time FNIHB (then MSB) began to discourage childbirth in remote aboriginal communities by "evacuating" aboriginal women to southern obstetrical facilities several weeks prior to their delivery dates (O'Neil, 1986). Despite this change, there has continued to be an implicit belief that maternal-infant health is an area of primary care that nurses continue to "do well" in remote aboriginal communities. But the maternal-infant health outcomes described in this chapter suggested that this is an area both women and infants may be falling through the cracks of the health care system. That is, less than ideal outcomes suggested that mothers and infants experienced less than optimum continuity of care. But precisely where did the gaps exist, and what factors contributed to the dynamics that created gaps? Nurses' preparation as primary care providers of women and infants is one factor that may influence continuity of care as experienced by women and infants in this study; in particular, the dimensions of management and informational continuity. If providers lack the knowledge and skills necessary to effectively and efficiently manage and follow-up the health of women and infants, then management and informational continuity will be compromised. For example, one important observation I made during a nurses' "in-service" educational session was the general lack of knowledge nurses had regarding diabetic screening in pregnancy. This lack of knowledge was also reflected in the data I collected from reviewing women's charts that demonstrated inadequate or inappropriate screening for 151 gestational diabetes, failure to identify patients at risk for gestational diabetes (e.g., women who had previously delivered a large-for-gestational age infant), or failure to proactively follow-up women at risk. Similarly, during fieldwork I observed the challenges several nurses experienced regarding the diagnosis, management and follow-up of anemia in both women and infants, which was substantiated by the chart review data. Relational continuity, as defined in Chapter Two, refers to on-going patient-provider relationships and consistency of personnel (Reid et al., 2002). Relational continuity may also be defined as "'practice' or 'site' continuity" (Haggerty et al., 2001, p. 6), for example, when a patient consistently attends for health care at one facility, even though the patient may not be consistently assigned to a single provider. Although there were a few exceptions in this study where patients moved between the study site and either a neighbouring community or one of the urban referral centers (such as Winnipeg), most patients received all or most of their primary care at the Nursing Station in the study site, thus experienced relational continuity in terms of health facility. Because the same physician had been visiting the study site for several years, review of patient chart data suggested it was likely that many patients experienced some degree of relational continuity in terms of their physician provider. However, because nurses provided most of the routine primary care, the high turnover rate among nursing staff meant that patients were vulnerable to poor relational continuity in terms of their primary care provider. In the remainder of this chapter, I present an interpretation of the findings that emerged from the contextual notes data derived from chart review. These findings are framed as case scenarios. Each case scenario demonstrated the links between outcomes, context and the three dimensions of continuity of care. In some cases, these links were self-evident. However, it would be inaccurate to suggest that these links were clear-cut in all cases, or that 152 other factors did not often, i f not usually, also play a role in the dynamics that underlay continuity of care. These other factors most often originated within the broader community context, the local context and structure of health services, and occasionally, the context and structure of the larger health services delivery system (i.e., referral services located in the larger urban centres). Cases where links were less apparent demanded a degree of speculation. Speculation, as offered in this interpretation, was suggested and supported by findings from the observations, interviews and artefacts data. The interrelationships among the factors presented in the following case scenarios were complex and sometimes challenging to unravel. However, the contextual notes data from chart review helped to illuminate and provide additional insights into both the health outcomes that were a finding of this study and the links between outcomes, context and the three dimensions of continuity of care.28 Case Scenario 6.1 27-year-old grand multip. Pregnancy diagnosed about 7 weeks' gestational age, initial work-up done at 18 weeks. Seen by four different RN providers on a total of four visits for prenatal care. Gestational diabetic. Anemic throughout pregnancy: hemoglobin less than 80 g/L at 30 weeks despite iron therapy. Induced vaginal delivery at 40 weeks due to gestational diabetes. Babe weighed less than 3500 grams. "Small" postpartum hemorrhage: hemoglobin 86 g/L after delivery. Developed unspecified postpartum infection requiring IV antibiotics prior to hospital discharge [to home community]. No evidence of postpartum follow-up re: infection or anemia, or of a routine postpartum check at 6 weeks. This patient had requested a referral for TL [tubal ligation] prior to this pregnancy. The community M D sent a referral to a gynecologist at about the same time that the patient became pregnant again. There is no evidence that options to terminate the pregnancy were discussed with the patient at the time she had the pregnancy M It must be noted that previous studies have suggested that chart documentation data tend to lack completeness (Buhler et al., 1988; Sheps & Robertson, 1984), in that providers may either carry out interventions or discuss interventions with patients that are not subsequently recorded in the patient chart. These authors also noted that incomplete charting had a detrimental influence on informational continuity. The present study was conducted in a setting where it was anticipated that multiple providers might see a patient over subsequent visits, therefore completeness of chart information was viewed as an important factor that contributed to informational continuity of care. Thus, for the purposes of this research, the stance taken was "not charted, not done". 153 diagnosed at 7 weeks' gestation. An R N documented during a routine prenatal visit that the patient stated she did not desire further pregnancies. The referral gynecologist was also the attending obstetrician [i.e., at delivery]. There was no evidence in the chart of postpartum follow-up or of a postpartum check at 6 weeks. [At the time of chart review] this woman is currently pregnant again, having conceived about 9 months post delivery. Did the communication re: desired TL fall through a gap and if so, where and how? INTERPRETIVE M E M O : This case is an exemplar of fragmented care between providers, with very significant consequences. This woman was at-risk for further [high risk] pregnancies due to several risk factors, including diabetes, anemia and parity, and had requested a TL. Yet she has gone through at least 2 further pregnancies since her initial request, due to fragmented care. Why was no one listening to her request and acting to expedite her request? This case scenario described a woman who despite an expressed wish for tubal ligation and personal health factors that placed her at increased risk during pregnancy (i.e., high parity, gestational diabetes and anemia) went through two further undesired pregnancies. While this woman delivered an infant of healthy weight at term, she also experienced complications of labour and delivery (i.e., postpartum hemorrhage and infection) and she did not attend a visit for routine postpartum care. Thus, only two of the four health outcomes criteria (i.e., 50%) were met. The outcomes of the subsequent pregnancy were not part of the data collected for this study and remain unknown. Chart data suggested compromise of all three dimensions of continuity of care in the prenatal care provided to the patient in this scenario. Although the patient received primary prenatal care at a single facility, a different nurse saw her on each of four visits; a finding that suggested a consistent patient-provider relationship was not part of the dynamic of care. Informational continuity was compromised in relation to passing on the patient's desire for tubal ligation, in that the referral and consultation to the gynecologist did not occur before the patient became pregnant again. Illustrating one of the limitations of chart data, it is unknown what discussion the gynecologist may have had or not had with the patient regarding tubal 154 ligation at the time of delivery and what decision may have been agreed on at that time, as that information did not appear in the patients' chart. This in itself further compromised informational continuity. For example, i f the patient and gynecologist had agreed that the patient would return at six weeks' postpartum to undergo tubal ligation, that information was not conveyed back to the nursing station and placed on the patient chart, where nurses may have been more likely to follow-up the plan with the patient and facilitate the referral and procedure. Nine months elapsed between delivery of this pregnancy and the patient's subsequent pregnancy, but there was no indication of follow-up of the original referral during that time. This illustrates the basis of the "not charted, not done" stance taken in this study. Management continuity compromised several aspects of care in relation to the current pregnancy as well as a future pregnancy. For example, despite the patient's expressed desire for no further pregnancies, there was no indication that options related to pregnancy termination were discussed. Although pregnancy was diagnosed at about seven weeks' gestation, the initial prenatal workup was not done until 18 weeks' gestation, thus this patient received little prenatal care or assessment during the critical first trimester of pregnancy. There was no evidence of postpartum follow-up of either anemia or infection, or evidence that the patient was encouraged or reminded to attend for routine postpartum care at six weeks' postpartum. This last aspect of care was perhaps the most significant in terms of facilitating the patient's referral for tubal ligation; the routine postpartum visit would have offered an opportunity for a nurse (or physician) to review the patient chart, follow-up the previously made referral, and discuss interim contraceptive options with the patient. Contextual factors that may have also contributed to the dynamics of continuity in this case scenario included the finding that high parity and multiple risk factors in pregnancy were common within the study site, to the extent that these had become normalized to some 155 degree. Over six percent of women in the sample for this study were "grand multips" (see Table 2) and diabetes (both gestational and pre-existing) and anemia were also relatively prevalent in the sample (see Table 5). The effect of normalizing risk factors in pregnancy was that even when patient risk factors indicated that a more proactive or aggressive approach to management and follow-up may have been appropriate, these factors did not prompt nurses to modify care management. Thus, normalizing risk factors held implications for management continuity of care. This was a key finding of this study and will be discussed at greater length in Chapter Seven. Other contextual factors that may potentially have contributed to the dynamic underlying continuity of care in this case scenario included the individual patient context. For example, the woman in this case scenario may have avoided undesired pregnancies by using contraceptive methods other than tubal ligation, and it might be argued that she had a responsibility to do so. However, acceptability of contraceptive methods is a highly individual choice among all people, men as well as women. In the study site, as in similar communities, politico-economic, community and family dynamics sometimes subtly encouraged large families and discouraged women from using contraception. The broader community context also held implications for the patient's ability to engage with self-responsibility for continuity of care. While the precise context of this individuals' life was not accessible as data, informal interviews with other women in the community, as well as interviews with health professionals and with para-professional staff who were members of the community suggested that the context of women's lives in the community often imposed challenges that as a grand multip, the patient in this case scenario may have shared: She may have had from five to ten or more children to care for, may have lived in one of the many houses in the community that lacked running water, may have had no resources for child care 156 that would allow her time to visit the Nursing Station for her own health needs, and may have had little access to transportation to attend prenatal and postpartum visits. Case Scenario 6.2 Teen multip. Presented at 31 weeks' gestational age and had initial prenatal workup. No complaints. Urine specimen was sent, results returned positive for a urinary tract infection (UTI) within 2 days. The lab result sheet was flagged for the patient to be called in for an antibiotic prescription. There was no documentation indicating the patient had been recalled. Patient presented 6 days later, with complaint of fever, chills, aching. Febrile on exam. The lab result was found on