UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

A randomized controlled trial to test the effect of an antenatal educational intervention on breastfeeding… Kluka, Sandra Maria 2004

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

Item Metadata

Download

Media
831-ubc_2004-931412.pdf [ 10.28MB ]
Metadata
JSON: 831-1.0091835.json
JSON-LD: 831-1.0091835-ld.json
RDF/XML (Pretty): 831-1.0091835-rdf.xml
RDF/JSON: 831-1.0091835-rdf.json
Turtle: 831-1.0091835-turtle.txt
N-Triples: 831-1.0091835-rdf-ntriples.txt
Original Record: 831-1.0091835-source.json
Full Text
831-1.0091835-fulltext.txt
Citation
831-1.0091835.ris

Full Text

A RANDOMIZED CONTROLLED TRIAL TO TEST THE EFFECT OF AN ANTENATAL EDUCATIONAL INTERVENTION ON BREASTFEEDING DURATION AMONG PRIMIPAROUS WOMEN by SANDRA MARIE (COSGROVE) KLUKA B.N., The University of Manitoba, 1978 M.N., The University of Manitoba, 1992  A THESIS SUBMITTED IN PARTIAL FULFILMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY IN NURSING In THE FACULTY OF GRADUATE STUDIES SCHOOL OF NURSING  We accept this thesis as conforming to the required standard:  THE UNIVERSITY OF BRITISH COLUMBIA July 2004 © SANDRA MARIE (COSGROVE) KLUKA, 2004  ii  ABSTRACT  Although the health benefits of breastfeeding have been well established, the majority of Canadian women wean prior to six months after birth. A randomized controlled trial was conducted to test the effect of an antenatal educational intervention on breastfeeding duration. The study examined how long women assigned to the intervention group (n=l 11) breastfed their infants compared to women who received usual care (n=98). Secondary outcomes of the study (i.e. maternal competence, breastfeeding confidence, maternal satisfaction, fatigue and onset of depressive symptoms) were analyzed in relation to breastfeeding duration. Primiparous women, who attended a series of antenatal classes and had decided to breastfeed their infants, were recruited. The intervention, which included completion of a pre-workshop guide and attendance at an interactive group workshop, emphasized strategies expectant mothers and their partners could use to prevent or minimize the problems most often reported as the reasons women quit breastfeeding. Strategies also were provided that have been regarded by mothers and health experts as being helpful to prolong lactation. The analysis was conducted on an intention-totreat basis. Results indicated there was no statistically significant difference between the groups regarding the proportion of women still mainly breastfeeding at 24 weeks after birth. Multivariate analyses yielded three significant predictive variables of breastfeeding duration including higher breastfeeding confidence (p=.001), non-smoking status (p=.017), and the mother receiving a visit by a community health nurse within two weeks after the infant's birth (p=.023). Implications for practice and research are discussed.  iii  T A B L E OF CONTENTS  Abstract  ii  Table of Contents  iii  List of Tables  vii  List of Figures  ix  Acknowledgments  x  CHAPTER ONE  BACKGROUND  1.1 1.2 1.3 1.4  Statement of the Problem Study Purpose Theoretical Framework Literature Review 1.4.1.1 Systematic Reviews of Breastfeeding Duration 1.4.1.2 Antenatal Studies to Extend Breastfeeding Duration 1.4.2 Variables DerivedfromTheoretical Framework 1.4.2.1 Maternal Competence 1.4.2.2 Breastfeeding Confidence 1.4.2.3 Maternal Satisfaction 1.4.2.4 Fatigue 1.4.2.5 Postnatal Depression 1.4.3 Factors Associated with Breastfeeding Duration 1.4.3.1 Maternal Antecedents 1.4.3.2 Internal Resources 1.4.3.3 External Resources 1.4.3.4 Interferences 1.5.5 Summary  CHAPTER TWO 2.1 2.2 2.2 2.3 2.4 2.5  1 5 8 9 11 11 13 27 27 28 29 30 32 33 34 34 36 38 40  METHODS  42  Design Primary Research Question Secondary Research Questions Procedures Sample Participant Inclusion Criteria  42 42 43 43 44 46  T A B L E OF CONTENTS (Continued) 2.6 2.7 2.8 2.9  2.10 2.11 2.12  Participant Exclusion Criteria Recruitment Procedure The Intervention Instrumentation 2.9.1 Duration of Breastfeeding - Operational Definition 2.9.2 Parenting Sense of Competence Scale 2.9.3 Breastfeeding Self Efficacy Scale 2.9.4 Maternal Breastfeeding Evaluation Scale 2.9.5 Multidimensional Assessment of Fatigue Scale 2.9.6 Edinburgh Postnatal Depression Scale 2.9.7 Personal Information Data Collection Forms 2.9.8 Qualitative Data Collection Ethical Considerations Strengths and Limitations Plan for Data Analysis 2.12.1 Primary Research Question 2.12.2 Analytic Approach to the Primary Question 2.12.3 Secondary Research Questions 2.12.4 Analytic Approach to the Secondary Questions  CHAPTER THREE FINDINGS 3.1  3.2 3.3  3.4 3.5  47 47 52 53 54 54 56 57 59 61 62 63 63 65 68 69 69 70 70 72  Sample Characteristics 72 3.1.1 Participants' Demographic Data 72 3.1.2 Comparison of Baseline Variables by Group 72 3.2.3 Description of Variables for the Total Sample Following Delivery 76 3.3.3 Group Comparisons of Variables Following Delivery 77 Primary Research Question - Duration of Breastfeeding 78 Secondary Research Questions 83 3.3.1 Psychometric Properties of Major Study Instruments 88 3.3.2 Predisposing Factors Associated with Duration 99 3.3.3 Factors Associated with Increased Duration 101 3.3.4 Additional Variables Associated with Duration 104 Qualitative Outcomes 112 Summary of Findings 114  T A B L E OF CONTENTS (Continued) CHAPTER FOUR 4.1  4.3  4.4 4.5 4.6 4.7  DISCUSSION Discussion of Findings 4.1.1 Primary Outcome 4.1.2 The Study Design 4.1.3 The Intervention 4.1.4 Secondary Outcomes 4.1.5 Predictive Variables 4.1.6 Qualitative Outcomes Recommendations for Future Research 4.3.1 Methodological Considerations 4.3.2 Recommendations for Intervention Design Implications for Practice Implications for Education Implications for Health Policy Dissemination of Findings  4.8 Study Summary and Conclusion  117 117 117 122 124 127 134 136 138 138 140 143 147 149 151 153  References  154  Bibliography  176  Appendices  184 Appendix I. Pre-Workshop Guide  184  Appendix II. Interactive Group Workshop  196  Appendix III. Antenatal Data Collection - Personal Information  197  Appendix IV. Parenting Sense of Competence Scale  198  Appendix V. Breastfeeding Self-Efficacy Scale  199  Appendix VI. Maternal Breastfeeding Evaluation Scale  201  Appendix VII. Multidimensional Assessment of Fatigue Scale  203  Appendix VIII. Edinburgh Postnatal Depression Scale  204  VI  T A B L E OF CONTENTS (Continued) Appendix IX. Postnatal Data Collection - 2 week Personal Information and Feeding Data  206  Appendix X. Postnatal Data Collection - 12 week Personal Information and Feeding Data  208  Appendix XI. Postnatal Data Collection - 24 week Personal Information and Feeding Data  210  Appendix XII. Content of Intervention versus Usual Care  212  Appendix XIII. Letter of Information  213  Appendix XIV. Consent Form  217  Appendix XV. Agreement of Research Assistant(s) to Maintain Confidentiality  219  VII  LIST OF TABLES Table 1. Antenatal Intervention Studies to Prolong Breastfeeding Duration  16  Table 2. Design Characteristics of Antenatal Education Intervention Studies  24  Table 3. Study Timeline  48  Table 4. Comparison of Groups by Baseline Demographic Variables  73  Table 5. Comparison of Groups by Personal Factors  74  Table 6. Number of Women in each Group by Feeding Category or Study Withdrawal  79  Table 7. Proportion of Women Mainly Breastfeeding by Group at 2,12, and 24 weeks  82  Table 8. Repeated Measures ANOVA Results of Interval Variables  85  Table 9. Internal Consistency Values of Major Study Instruments  88  Table 10. PSOC Items with Principal Components Varimax Factor Loadings  92  Table 11. BSES Items with Principal Components Varimax Factor Loadings  93  Table 12. MBFES Items with Principal Components Varimax Factor Loadings  96  Table 13. MAF Items with Principal Components Varimax Factor Loadings  98  Table 14. Cross-tab Results of Categorical Variables and Mainly Breastfeeding  100  Table 15. Logistic Regression to Model Mainly Breastfeeding at 2 weeks Full Model  106  Table 16. Logistic Regression to Model Mainly Breastfeeding at 2 weeks Final Model  107  LIST OF TABLES (Continued) Table 17. Logistic Regression to Model Mainly Breastfeeding at 12 weeks Full Model Table 18. Logistic Regression to Model Mainly Breastfeeding at 12 weeks Final Model Table 19. Logistic Regression to Model Mainly Breastfeeding at 24 weeks Full Model Table 20. Logistic Regression to Model Mainly Breastfeeding at 24 weeks Final Model Table 21. Cox's Proportional Hazard Regression - Full Model Table 22. Cox's Proportional Hazard Regression - Final Model  LIST OF FIGURES Figure 1. Percent of all Manitoba Infants Breastfed  45  Figure 2. Study Design  49  Figure 3. Flow Diagram -Trial Recruitment Outcomes  50  Figure 4. Proportion of Women Continuing to Mainly Breastfeed by Group  82  Figure 5. Proportion of Women Continuing to Mainly Breastfeed by Smoking Status at 2 weeks  102  Figure 6. Proportion of Women Continuing to Mainly Breastfeed by Smoking Status at 12 weeks  103  Figure 7. Proportion of Women Continuing to Mainly Breastfeed by Whether Mother was Breastfed  103  Figure 8. Proportion of Women Continuing to Mainly Breastfeed by Whether Mother Received CFTN Visit by 2 weeks after birth  104  ACKNOWLEDGMENTS  With special thanks, I would like to acknowledge a number of individuals who have provided assistance and encouragement throughout my doctoral journey. First, I would like to thank the members of my supervisory committee: Professor Elaine Carty who sparked my ideas and thoughtfully guided my progress; Dr. Wendy Hall who inspired my thinking and carefully critiqued my work; Dr. Roberta Hewat who provided substantive expertise and challenged my thinking; and Dr. Annette Gupton who offered ongoing direction and support. I also would like to thank those who contributed directly to the completion of the thesis: Michelle LeDrew who helped me design the intervention and facilitated the group sessions, Cathy Jacobson and Tamara Hess who contacted participants and recorded the data, and Mary Chang for her statistical expertise and wonderful sense of humor. Financial supportfromthe Canadian Nurses Foundation/ Canadian Health Services Research Foundation and the Health Sciences Center Foundation, Winnipeg, Manitoba is gratefully acknowledged. I also would like to commend those women and their partners who participated in this research; it was a joy to meet you and satisfying to learn of your interest in my work. Pursuit of this degree as an inaugural 'off-campus' doctoral nursing student would not have been possible without the vision and commitment of Drs. Janet Beaton, Kathryn May, Betty Davies, Cynthia Cameron, and Sonia Acorn. Completion of the degree is attributed to the wealth of support and patience receivedfrommy: family (husband Joe and son Chris, parents Grant and Alice, and sister Carolyn); study group (Kathryn, Donna, Elaine, and Bev); University of Manitoba colleagues (Marilyn, Peri, and Maureen); mentors (Mrs.Margaret Mackling and Dr. Linda Kristjanson), and manyfriends.Your love and support have been an indispensable source of strength.  CHAPTER ONE BACKGROUND Breast milk is considered to be the best source of nutrition for human infants (Canadian Pediatric Society, Dieticians of Canada, & Health Canada, 1998). Breastfeeding is recommended by the American Academy of Pediatrics (1997) because of the associated "health, nutritional, immunologic, developmental, psychological, social, economic, and environmental benefits" (p. 1035). Research findings have demonstrated significant advantages of breastfeeding for infants, mothers, families, and societies including reduced health care costs (American Academy of Pediatrics). Extending the duration of breastfeeding has been recognized as a national health priority (Health Canada, 2000). The nutritional benefits of breast milk for the infant have been extensively documented in the literature (Heinig & Dewey, 1996; Kunz, Rodriguez-Palmero, Koletzko, & Jensen, 1999; Rodriguez-Palmero, Koletzko, Kunz, & Jensen, 1999). Some of the many benefits include: a reduced risk of sudden infant death syndrome, asthma, childhood cancers, insulin-dependent diabetes, and rheumatoid arthritis (UNICEF UK Baby Friendly Initiative, 2003). After completing a systematic review, Kramer & Kakuma (2002) reported infants experience less morbidityfromgastrointestinal infection and no deficits in growth when exclusively breastfed for six months. Breastfeeding also is favorably linked with childhood development in that breastfed infants have scored significantly higher than formula fed infants when tested for cognitive and motor skill development (Rogan & Gladen, 1993). The health benefits of breastfeeding extend to the mother and include enhanced maternalinfant attachment, less postpartum bleeding, more rapid uterine involution, weight reduction, less iron-deficiency, improved self-esteem, and the satisfaction of knowing one is able to supply the  2  best source of nutrients for one's infant (Bick, 1999; Heinig & Dewey, 1997; Kelly, 1999; Williams, 1997). Newcombe et al.'s (1994) study (n=6,888) demonstrated that the premenopausal breast cancerriskwas 22% lower among women who had breastfed. Following a critical review of clinical papers, Labbok (1999) reported "the initiation and physiologic completion of breastfeeding over the first two to seven months postpartum was associated with a significant reduction in ovarian cancer risk" (p.495). She also summarized evidence of other health benefits for women in that breastfeeding reduces the incidence of bladder or other infections, as well as theriskof spinal or hip fractures. Labbok strongly recommended that women be advised of the many short and long term health benefits for mothers who have breastfed, rather than limiting the information conveyed to the health benefits for the infant. Given the combined potential health benefits for families, and the broader economic benefits for society, guidelines prepared by Health Canada (2000) urged the development of new strategies to support the continuation of breastfeeding for at least six months after the infant's birth. Reasons why women are not continuing to breastfeed for as long as recommended are complex (Basire, Pullon, & McLeod, 1997). Historical, socio-political, and cultural influences have shaped the context for women who are deciding whether to initiate breastfeeding or how long to continue to breastfeed (Nichols, 2000). Twenty-first century lifestyles that may contribute to premature weaning include increasing numbers of employed women despite societal expectations of women to continue as primary child-care providers (Kirkley, 2000; Ruchala & Halstead, 1994; Singh, 2000). Furthermore, industry and the mass media have perpetuated views of women's breasts as sex objects, potentially resulting in women or their  3  partners feeling discomfort when breastfeeding in public places (Dykes & Griffiths, 1998; Sheeshka, Potter, Norrie, Valaitis, Adams, & Kuczynski, 2001). Frequently women do not want to wean but feel compelled to do so because of difficulties with the breastfeeding process. According to the Statistics Canada (1994/1995) report titled the National Population Health Survey, women who stopped breastfeeding before three months after birth reported the following reasons for their decision: perceived insufficient milk supply, difficulties with technique, sore nipples, inconvenience, low levels of breastfeeding confidence, and fatigue. Thus, some women find the demands of breastfeeding can exceed the perceived benefits, especially if the woman is already feeling overwhelmedfromthe demands of new motherhood (McVeigh, 1997). In addition, women are more likely to wean earlier if they are younger, unmarried, smoke, have less education, or have lower socioeconomic status (Statistics Canada). Demographic variables cited in previous breastfeeding research, such as age or ethnicity, are, however, non-modifiable, and less predictive of breastfeeding duration than psychosocial variables such as prenatal commitment to breastfeeding (Donath et al., 2003; Dungy, Losch, & Russell, 1994b). In contrast, women who continue to breastfeed have perceived the breastfeeding experience as satisfying, the infant to be enjoying the breastfeeding sessions, and their partners to be supportive of breastfeeding (Bar-Yam & Darby, 1997; Freed, Fraley, & Schanler, 1992). The latter consideration includes the partner's support of his wife/partner breastfeeding when others were present (Arora, McJunkin, Wehrer, & Kuhn, 2000). Hoddinott and Pill (1999) reported that women who believe they have managed well with prior challenges, and who anticipate being  4  able to learn to breastfeed, are among those most likely to achieve their breastfeeding goals. Feeling confident about one's ability to become competent and comfortable with breastfeeding has been associated with longer duration of breastfeeding (Dennis, 1999). The timing of the decision to breastfeed also is associated with continued breastfeeding (Lawson & Tulloch, 1995; O'Campo, Faden, Gielen, & Wang, 1992). It has been noted that women who decide early or prior to the pregnancy will try breastfeeding and are more likely to continue to breastfeed longer (Dennis, 2002a; Lawson & Tulloch). In a recent study, it also was suggested that the intended duration of breastfeeding is a strong predictor of actual duration (Donath et al., 2003). New mothers are coping with a multitude of physical and emotional changes during the postnatal period. Family and partner relationships must be realigned and the realities of motherhood are often very different than anticipated (Montgomery, 2000). A breastfeeding mother has specific skills to acquire with her infant to establish a feeding pattern, a task that takes time and practice for the mother and the baby. A mother's perception of her infant's response to breastfeeding is pivotal to how long she may continue to breastfeed. New parents cite fatigue as an anticipated, yet overwhelming, aspect of this transitional period (McVeigh, 1997). As well, according to Beck (2001), up to 80% of all new mothers experience "postnatal blues" and approximately 13% sufferfrompostpartum depression (Beck, 1998; O'Hara & Swain, 1996). Mothers who are tired and/or depressed may be less likely to continue to breastfeed (Bick, MacArthur, & Lancashire 1998; Campbell, Cohn, Flanagan, Popper, & Meyers, 1992; Cooper, Murray & Stein, 1993; Galler, Harrison, Biggs, Ramsey, & Forde, 1999;  5  Hellin & Waller, 1994; Milligan, Parks, Kitzman, & Lenz, 1997; Papinczak & Turner, 2000; Tulman & Fawcett, 1988). Statement of the Problem While breastfeeding initiation rates are reasonably high in Canada, the length of time a woman continues to breastfeed falls far short of the recommendations by major health promotion organizations. In Manitoba, although up to 85% of infants are initially breastfed, only 36% continue to receive breast milk by six months after birth, and 15% by one year after birth according to a report completed by Statistics Canada (1996) titled the National Longitudinal Survey of Children and Youth (NLSCY). Therefore, the mean duration of breastfeeding is far less than targets set by the Joint Statement on Nutrition for Healthy Term Infantsfromthe Canadian Pediatric Society, Dieticians of Canada and Health Canada (1998), who urge exclusive breastfeeding until the infant is four months of age. The duration rates also are less than targets set by the American Academy of Pediatrics (1997), and the World Health Organization (1999), who recommend exclusive breastfeeding until six months of age followed by the introduction of complementary foods and continued breastfeeding for at least two years. Around the world, many antenatal, early postnatal, and long-term programs have been used to try to increase the numbers of women who breastfeed, and to improve breastfeeding duration. Multilevel strategies that include education, media promotion, and community support programs have been proposed to increase breastfeeding initiation and continuation. Three systematic reviews have been conducted to evaluate the effectiveness of interventions to promote continued breastfeeding (Fairbank, O'Meara, Renfrew, Woolridge, Sowden, & Lister-Sharp,  2000; Green, 1999; Sikorski & Renfrew, 1998). Fairbank et al. and Green examined any type of intervention designed to promote breastfeeding. Sikorski and Renfrew limited the scope of their review by excluding interventions that were primarily antenatal and educational. The Fairbank et al. review concluded... "Packages of interventions have been shown to be effective at increasing the initiation and, in most cases, the duration of breastfeeding in developed countries. Effective packages appear to include a peer support programme and/or a media campaign combined with structural changes to the health sector and/or health education activities"(p. vii). Green's review also concluded that "results suggest prenatal education, when combined with other interventions... is associated with improved breastfeeding practices" (p. vii). However, again according to Green, only a few studies have focused on the impact of antenatal education in isolation from other interventions (Pugin, Valdes, Labbok, Perez, & Aravena, 1996; Sciacca, Phipps, Dube, & Ratliff, 1995; Wiles, 1984), leading Green to call for further research to help identify which interventions, or combinations of interventions, could change breastfeeding behavior. Organization and management of antenatal and postnatal programs require substantial human and fiscal resources. Therefore, in order to provide salient, cost-effective interventions, health care and social resources should be streamlined. Fairbank et al.'s (2000) review evaluated the effect of stand-alone interventions and concluded that... "small, informal group health education classes appear to increase initiation and duration rates of breastfeeding in developed countries. This form of intervention can be effective when implemented among women of all income groups andfromminority ethnic groups." (p.48). Green (1999) also concluded that women who received information about breastfeeding during the antenatal period were more  7  likely to exclusively breastfeed longer than women who did not receive such information. Small group antenatal educational interventions may be offered at relatively limited expense to facilitate opportunities for several women to receive information and support, and to share common concerns or ideas. Most communities have nurses or community women, often trained as lactation consultants, who could carry out an effective educational program. In addition to the systematic reviews, the literature yielded a limited number of studies (n=l 1), conducted in countries other than Canada, that have tested the effect of an antenatal, educational intervention on breastfeeding duration. Of the seven randomized controlled trials, four reported interventions that significantly prolonged breastfeeding duration, but only two of the studies reported a projected sample size necessary to obtain the power to detect a difference between the groups (Duffy, Percival, & Kershaw, 1997; Hill, 1987; Kaplowitz & Olson, 1983; Kistin, Benton, Rao, & Sullivan, 1990; Loh, Kelleher, Long, & Loftus, 1997; Pugin et al., 1996; Rossiter, 1994; Sciacca et al., 1995; Serwint, Wilson, Vogelhut, Repke, & Seidel, 1996; Sheehan, 1999; Wiles, 1984). Both systematic reviews, conducted by Fairbank et al. (2000) and Green (1999) cautioned that the studies they reviewed often lacked the methodological rigor needed to determine which interventions could extend breastfeeding duration. Consequently, a randomized controlled trial was designed using sufficient power to test the effect of an antenatal educational intervention on breastfeeding duration.  8  Study Purpose The purpose of this study was to determine if an antenatal educational intervention in primiparous women randomized to an intervention group breastfed longer compared to primiparous women receiving usual care. The study was undertaken in the context of Mercer's (1981; 1995) theory of maternal role attainment which embraces the multifactorial experience of breastfeeding and its relation to the transition to motherhood. Contents and structure of the intervention were based on existing research about factors related to the duration of breastfeeding, prior interventions that had demonstrated a significant effect on breastfeeding duration, and the researcher's clinical experience. Building on the strengths of previous work, the intervention was designed to commend women for their decision and commitment to breastfeed, to promote breastfeeding competence, confidence, and satisfaction, and to minimize or prevent the problems most frequently reported by women underlying their decision to stop breastfeeding. The intervention included a pre-workshop guide and attendance at an interactive group workshop during the third trimester of pregnancy (Appendices I and II). The guide was designed to stimulate thinking about the many aspects of becoming a first-time mother planning to breastfeed. The guide encouraged participants, and their partners where appropriate, to consider the many challenges of the early postpartum period, and to identify individual strategies they might find helpful. The group workshop was designed to provide further information, support, and discussion opportunities with other expectant women. During the session, the participants, (and their partners where appropriate), were urged to share the strategies they had identified and  to add to their own guides for future reference. A nurse lactation consultant, who had breastfed her own children and assisted the principal investigator in the design of the intervention, facilitated the workshop. Theoretical Framework Mercer defined maternal role attainment as an interactional and developmental process (1981; 1995). Based on Rubin's (1967a; 1967b) seminal research, Mercer and colleagues studied the process of maternal role acquisition during the first year after birth (Mercer 1985; 1986; Mercer & Ferketich, 1988; 1994; Mercer, Ferketich, & DeJoseph, 1993; Mercer, Ferketich, May, DeJoseph, & Sollid, 1988; Mercer, Nichols, & Doyle, 1988). Components of maternal role attainment include acquisition of skills and a sense of mastery. Information that strengthens a mother's confidence and competence, including information about the responsiveness of her infant, is directly linked to her overall perception, and sense of satisfaction in the new role (Bullock & Pridham, 1988; Mercer 1995). However, common postnatal experiences like fatigue or depression have been demonstrated to interfere with a mother's ability to cope resulting in premature weaning (Cooper et al., 1993; Galler et al., 1999; Mercer, 1995; Milligan, 1989). Guided by Mercer's theoretical framework, the variables selected for examination as secondary outcomes in this investigation included maternal competence, confidence, satisfaction, fatigue, and depression. Referring to the earlier work of Thornton and Nardi (1975), Mercer (1981; 1995) identified four stages in the acquisition of the maternal role: the anticipatory, formal, informal, and personal. The anticipatory stage begins during pregnancy as the woman begins a  10  relationship with the unborn child. The expectant mother identifies with the maternal role through fantasy and role-play. She may seek role models as she anticipates becoming a parent. The formal stage commences with the birth of the infant, and continues for approximately two months while the new mother works to interpret her birth experience, recover physically, affirm the normalcy of her baby, and develop skills in infant care. When the mother has learned appropriate responses to her infant's cues and responds in her own unique way, she is considered to have progressed to the informal stage. The final personal stage occurs when the mother begins to internalize her role, and feels a sense of harmony, while acknowledging the infant as a central person in her life. The four stages can be applied to the experience of a new mother learning to breastfeed (Bocar & Moore, 1987). While pregnant, the expectant mother begins to plan how she will address aspects of motherhood including infant feeding. Mercer's (1981) theory cites the antenatal period as an ideal time to introduce information and strategies about what a woman can anticipate about becoming a mother. Hauck and Dimmock (1994) concur and urge providers to maximize learning opportunities about what to expect before birth when parents may have more time and energy. Antenatal education offers opportunities to acquire information about what to expect, and to meet other women who may have breastfed. After birth and during the formal stage, the mother relies on social supports to acknowledge and reinforce her efforts in learning the new tasks of breastfeeding such as recognizing her infant's hunger cues. If breastfeeding is progressing well, the mother and infant dyad may then begin to experience a sense of harmony with the breastfeeding experience. Furthermore, the mother acknowledges the contribution of  11  her infant to the success of the breastfeeding experience. By the personal stage, it is anticipated a woman becomes more comfortable with her style of mothering and breastfeeding. Therefore, guided by the theoreticalframework,it is suggested by the researcher that women who fulfill their breastfeeding expectations are likely to experience a sense of physiological and psychological well being. A woman who feels confident will acquire the competence to breastfeed successfully and may be more likely to initiate breastfeeding and to continue to breastfeed. A woman who feels satisfied and perceives her infant is responding positively to breastfeeding also may be more likely to breastfeed longer. Alternatively, a woman who experiences fatigue or postnatal depression may be less likely to report a state of physiological or psychological well being. A woman who is persistently tired or depressed may feel less satisfied or feel less certain about the infant's response to breastfeeding, and therefore be more likely to wean sooner. Analyses of these secondary outcomes, as derivedfromMercer's (1981; 1995) theory, were used to understand differences in breastfeeding duration. Literature Review The literature review was completed to analyze the outcomes of research that had tested antenatal interventions to extend the duration of breastfeeding. DerivedfromMercer's (1981; 1995) theory of maternal role attainment, variables were selected and further examined in the literature to determine their relationship to breastfeeding duration. To establish data collection parameters, and to inform the development of the intervention, factors associated with breastfeeding initiation and duration were identified. Systematic Reviews ofBreastfeeding Duration  12  Green's (1999) systematic review of 51 intervention studies spanning two decades of breastfeeding research concluded antenatal interventions enhance breastfeeding. However, she also cited major methodological flaws, such as small sample size or the lack of a control group, as limitations of studies undertaken to improve breastfeeding. Inclusion of multiple interventions to promote breastfeeding limited the ability to adequately detect and test which intervention was the most useful among various groups of breastfeeding women at different times or in various circumstances. Green recommended larger scale trials be conducted to control for confounding variables and selection bias. Fairbank et al. (2000) completed a systematic review of 59 studies looking at the effectiveness of a variety of interventions (health education, institutional services, peer support programs, media campaigns, and multifaceted interventions) aimed at increasing breastfeeding. It was concluded that in developed countries, small, informal, group health education classes appeared to increase initiation and duration rates of breastfeeding. In particular, the authors stressed the importance of interactive methods of health education, rather than the use of breastfeeding literature, or more formal classes. The reviewers encouraged inclusion of the women's partners, mothers, or peers as part of any antenatal educational intervention approaches. Furthermore, the authors recommended randomized controlled trials should be conducted in future inquiries pertaining to breastfeeding initiation and continuation. A Cochrane review of 13 randomized trials reported multifaceted interventions offered in both the antenatal and postnatal period were able to extend breastfeeding duration until 2 months after birth, but found no difference at three, four, or six months after birth (Sikorski & Renfrew,  13  1998). The authors concluded there continued to be uncertainty as to the most effective way to prolong breastfeeding duration beyond two months after birth. Face-to-face interventions were deemed helpful, whereas telephone contact was not. Furthermore, the review did not include an examination of interventions that were conducted only in the antenatal period in order to limit the scope of the review (J. Sikorski, personal communication, September 23, 2000). Antenatal Intervention Studies to Extend Breastfeeding Duration A literature review (1980-2000) yielded 11 randomized controlled trials or quasiexperimental studies that had examined the effect of an antenatal education intervention on breastfeeding duration (Table 1. Antenatal Intervention Studies to Prolong Breastfeeding Duration), and (Table 2. Design Characteristics of Antenatal Intervention Studies) (Duffy et al., 1997; Hill, 1987; Kaplowitz & Olson, 1983; Kistin et al., 1990; Loh et al., 1997; Pugin et al., 1996; Rossiter, 1994; Sciacca et al., 1995; Serwint et al., 1996; Sheehan, 1999; Wiles, 1984). Six of the studies were completed in the USA, three in Australia, one in Chile, and one in Ireland. Sample sizes ranged from 40-194 participants; the number of women assigned to an intervention group rangedfrom20-108. All studies included comparisons of breastfeeding duration rates with women in a control group. Four of the studies were categorized as quasiexperimental because participants were not assigned randomly to either the intervention group or the control group but were assigned according to which day they attended an antenatal clinic or which series of childbirth education classes they had registered to attend (Loh et al., 1997; Pugin et al., 1996; Sheehan, 1999; Wiles, 1984). Three of the studies aimed to increase breastfeeding duration among low-income women (Kistin et al., 1990; Sciacca et al., 1995; Serwint et al.,  14  1996). Only the studies that included a group session intervention reported a positive effect on breastfeeding duration (Duffy et al., 1997; Kistin et al.; Rossiter, 1994; Sciacca et al.; Wiles). For purposes of this literature review, breastfeeding at 4, 6, 12, or 24 weeks after birth was recorded as "any" breastfeeding (Table 1). Results of the randomized controlled trials rangedfromno statistical difference between intervention and control groups (Hill, 1987; Kaplowitz & Olson, 1983; Serwint et al., 1996) to observations of statistically significant differences between groups at 4 weeks after birth (Rossiter, 1994), 6 weeks after birth (Duffy et al., 1997), 12 weeks afterbirth (Kistin et al., 1990; Sciacca et al., 1995), and 24 weeks afterbirth (Pugin et al., 1996). Sheehan (1999) observed no difference between groups in terms of the proportion of women who continued to breastfeed by 25 weeks after birth. She also observed that many of the women in her study breastfed well beyond national norms irrespective of group assignment. Eight of the antenatal interventions that were reviewed included attendance at a group session that ranged in lengthfrom20 to 120 minutes. Content most often was related to information about breast care, breastfeeding techniques, problems, and resources. The intervention for Rossiter's (1994) study in Australia was designed specifically for use with a sample of Vietnamese women in that the sessions were designed to be culturally and language appropriate. One study intervention included prizes as incentives to increase the likelihood of participation and decrease attrition rates (Sciacca et al., 1995). Three studies (Kaplowitz & Olson, 1983; Loh et al., 1997; Serwint et al., 1996) tested an intervention other than a group session to provide breastfeeding information in addition to the information received during usual  15  antenatal care in the community. In the Kaplowitz and Olson study, a series of pamphlets were mailed to participants. Loh et al. gave out a fact sheet and questionnaire during a clinic visit and in the Serwint et al. study, participants met antenatally with the infant's future pediatrician who had received additional preparation regarding breastfeeding promotion and techniques. Methodological strengths of the studies reviewed included: 1) randomization of participants, 2) clear inclusion and exclusion criteria, 3) completion of a power analysis and sample size projection a priori, 4) comparisons of different modes of teaching using common content, 5) comparisons of groups to determine homogeneity at baseline, 6) blinded outcome assessments, 7) reasons for withdrawal, and 8) intention-to-treat analyses. However, the studies also had a number of weaknesses including: 1) limited ability to generalize thefindingsdue to small sample size, 2) inadequate design or analysis procedures to control for confounding variables, 3) no standardized operational definition of breastfeeding duration necessary to facilitate meta-analyses, replication, or comparison with other studies, and/or, 4) no controls for parity status, 5) administration of the intervention by the principal investigator, 6) incentives such as prizes offered to participants which may have biased the findings. The research, therefore, suggested that antenatal interventions may be effective, but results could not be generalized because of methodological limitations. In response, this project was designed to adequately test an antenatal intervention using a randomized controlled trial, with sufficient power, and an intention-to-treat analysis.  oo  5  oo S£  &  ^ o;  *8  O 3  >  ST  0  r >  lr P & o o 0"- 2 P its o ^> S". B O B  O (3  K)  £  13  6*8, o 5 g  o o  O H  5 £ o e O CD O " S £ B co C CD  '1  •O  O  -  •  3 R  c P  2  OQ B  B.  1. 5- S 2. ^ S oq S- r* ^ o3 S P a &. o o.  CD*' P . "O  CD  I g  —.  CD o 2 §  ^  co  K- p CfQ CD CD B . < w f> CD  t» S g.  CD B  B S  h a  CD  O  B  CD  O B,  CD CD CD  g  "  Cd  2  •fl  cv  < tj fD  g  D a  P  ^ I  2  CD  P. B"  OQ  p CO 3 5 o 2 S g. O Q '  O  O o ^  CO  DO CJ*  <  3  CD CD  0  OQ M  o  CD c co  i  3  p o  1 aI e I  O '1  C  CO  o o a  CO  3  CD  a s:  P  5*  «  f e* o  ET  CD  Cu OQ  CD X  Xi  CD  a*  3.  CD  a*  CD CA CA  CD O CD  v;  $  P  s-  o o  00  CD  rt. l*>  2  S. 8. & OQ  CO  CT p  I  B* <»  CD X)  I  CD  | I  CO 2 g to 5* •n 3 B o' B CD •o CD S* CD  a.  CD CD  3 g  "O  cn  cv. CD  rt>C/3rt 3 B. O  O  £- e I S 2 <»  OQ  CO  >-*  l-S P I fo  13 L_rt  <  It a. 2 *§ I CP to co  B-  CD  B  s ^  CD* O  CD CD R  ES  * s§  Tl S3.  p* 2. a* Er & g. 5 ' c r  I s a. oi o- 8 B o S> CD P o 2 3 § B.  tt)  o o  (a CX. 3 0  s. a.  5 O* vo »' o  >  a. a  5 or > g, £ O c Cr ^- JJ O O p ST*a  ct  o  p  O T3 O i-t, » o  3  co  2  B L,5 B fr cr. ct C o CL  S o o  I o  O o  2 o o 7  3  cv> u> u> >-* OS OS oo w  o  C cn  CL  CH  CL O  o cr o. tf> ^ sP P 3' o  EL o  3 D  ct era' CO  a  o  5  s  a*  B<  CL  I-8 a* ct  fa ct  3  o  3  O S  t(<0  rt TI e. £ «  B era  1-  >  C-  <fi  I f  CC  o p  o r> P  I: o o  a  § S g g a § *o oo « ^  I  at 3  s  CD  «. P o H  s. a  cc  B-  CC B E , . .  rt cc w cc p P  CL  xs  O  P P.  CD  I  5  5§ 2§ 3  CD CC T! CC CC <C T3  !  to 3 p  d3 2 S | 8  s 8  rt  »  Si 3 c 3* P  >5 ~  SN^)  rt »  a*  4*  cl cc  5*  o p  4*.  Ol-t Q  » < ro  3  i-t  Q>  O2 O  00 &. B" o° s 1 s. a. M Pr^ P-i asrgso ~"cro §• B — 2 t±l rt2Op p 2 ^  4^ P  §  &  ' O P  cc  B'  00  §  *<3 CC g  CL  CD  5» a r >  Beg" CD  ^  ? o * S n o 2 o 2 a S' £. &' a ^ * E? 8. 3o 3. ° * CD 6  O  G  II  II  rt  C& p  o' O  fj*  I:  2  rt ^ 2 O B  O o  &.  f  g  00  CD  22. 2 8 O T3  O CD CO  oo' B  Z °  < a  o «  5" ° CD  CD  3  O  e &. 3. CD  a I CD'  2  T3 2 3. a.  H T  CD  B CD  O .3  5 CD  3' o 3"  CD  2 2  5'  CD  ^2 CD  O  3 "a O  <r  5' a. 2 e. a. 2 r*  a oo o CO  CD  < a  8 I a 2  o 1 ?  H  B*  oo 3 o I 2 ° rt  3  B ?£ •S3 5 c5'p  e? oo  CD CD  B-  a. 5;  |  S'l  I  CD  CD CL O  CD  Er  3 o o  CD  50 O  w. rt CO O B  to  o  >- O ^ B. — °3CD Cj" CD  rt  TJ  or >  ^ 5? C^ rt rt» CD p B^ O 3  a  3 1 F> 3 p O o Cr* n n O ••  CD . CL  w2i| : ft TO P . 2. O W  8  P  cs a a «  E.  ~ l f l  P  CD  I  I  3  CL  • . CD p T3 CD CD  P  a  CD  I o  O  5  p.  CD  3 3 w  cB B  Tl  CO  -a «^ 3 t3  M  C.  o p  o |  C7\  I -  Oto u> ^ |  H  —  I DO  Tl  *  ON  o  I" o a  t o  8 i' t  s I SP 5. Ch  o CL  p  CO  I I  d. cST t> a  2, S  G)  CO 3  o  C>  t Tl  Os 00  ir  2? £  to cl  3 I  Tl  Tl  CS  o  •a  0 5  1  w  P  rt  0  c3  CD rt CL B  CD  ST  cT  O o  P  O  I:  5  00  £ 3  vo ^  £2 S3  tr  g  «  a- -*• R.  Pt C D P ~~  B*  O  O  B  O o  R  g 3 o o 7 z R 8 o || II vof p ftCD  0  3. g  0  - T C P  CD  13  2: CD  CD CD CD 8  O C D CO  O H  oo"  • o 2- 5" 3 |  c R p  5 8.  3 ~ CD co * T3 2> Q B'<S'i 8 a a. P p  «. p o  0". -l  I  3  M,  3-  O  CD  co  c R  s. p  g §  2.  O CD  I  CD d  P. co CD 2  ET> o  p  c B wp  o  3 g B £3 -3 a i - 1 P CD o' S ^U ff§T H 2- a. . g? i ? 8* 8 •» 3 P C D ST CD O g. 0 &• C D a o P* ,_.  y B  *"  p -o cr " CD 3 a.  CD CD CD  » 63 .< S o & -j- T l _  p  CO  ft  to  i-h  o  u  CD B  o  ^  c p oo 3 2 2 a E » C D P" £ • P.' P p co C D *' B 3 55 1 S. cr C o P p 5* p cr B S  P  ^- B  3  0  v< S. CD  &- Q B* Z 3 3 » o  2-1°  2  o  (  R i o £  CO  Pi  p  Tl  SO  o * '3 N  CD C ,D  I  B  50  § p  3. CD tt  ^  CO  =• g  o* S  CD 5  o  CD o „ « R 2 T)  VO C  X  x  ^  o '  cr < OV cl CD er  Tl  CL  <- 3 e °  OQ  CD _  A  S  B  ^ 2 P £  * co G-  5-  "B R I 3 o  S  3'  2 as " 3 rT* 2 3 «  B 2 -i B CD  KJ  C/3  >— >  so ET cr P -P  O o 12 2, 3 n o z e o o II II T ft. EL 3 g oo as ~ H  & P £s «> ft- <?  ••§••8 £ i .  ±  < I cc  B*  e  p TJ  2- n*  < B  o o- CSs ^  B.8-  ft gt BT do' a cc &  p T3 0)  £. °* o 2 a* -  a> B  o B  o 3  ^ -u  gs W  00  as  as  00  Tl  5- EL S  311 & 2 3 B"Ip I I a O o' do'  VJ  Ch  &• O  <T 8  a  I  CD a B  5  Tl  B  vo oo CD '  ••s  I  )o  vo oo  rt  o  N  z o  w o CD ^  CD  CD 'C 3. „ •» 5  2 2 E* g-  2. o 3  ion  3 C 3D  CO  •o ut  rt rt CD  Cu 9-  n  CD  o'  o  3  t» >-t P O  z o  uidoi zation oce not  3  f  Cfl  rt  n  3.  CO  3  g  cn  C D Cu  O  o 3  X  X  X  O c« oo ^  £• a I^ o  CD  3 3  E X  X  X  5' 3  5' o Cd > 3 c7 8, p  CD  ar -rj g  03  re 3  m Cu  c  o p  #--*)•> — I* •  o 3  rtH  3 <-»-  g  T J CD  CD  "O O ft '1  CD  Op3 O g 2 -o  ©<EJ 3' g  3 f? S. 5 T3 ^ ! ^ 2 o o S» I 3 ? | 2 ' 3 CD P  cy> w  2. it  1  CD  5.  Z o  & rt CD  Cu  z o  z o  z o  S" Cu CD  S" CD Cu  Cu CD  X  X  CO  rt  CD  rt SJ rt  z o CD  Cu X  X  X  X  CD* P  3  rt-  o' 3  CZ3 .  p  s  CD'  > o co crt> o oo § 3 3  7i g Z o 5- 3  a cu cr CD Cu 3 2 Zo P  uf ? S p H 2  * c c? c2. o P o 3  Ju  P  00  IH  TJ  I  P . &°  2-  c  SO Os  so vo Os  3 „  ct 2! o  3  ft  CD  o p W 3 Q 3,- o a cT  Cl T3 O  3  C  3- g.  E-B-f; X  2 o  25 o  g  CO  ct  o 3 N  &.  o 3  X  xw  ^ >S CO  2. Li.  CO  * If "  Tl  5  2 £  Ti  X  ^ -  2 c5  B' co  (g  Tl  « 3 o  |  S •§ a £ ci P 2 S Cl s B* Cl  25 o r+  5  ct a.  X  o CO  Cl Cu  Cl CL  g  > O CO ci o 5  u  H  3  3  CD  Cu  8 ° 3  •8 o  a.  s  Cl Cu  Cl  o-  X  r> £T 3 cl H |  E2. p o Co  co Ct O °  3  3a.  O  g 6d T| C  r» O  g  o? 0  3. ^ 3. P  ct 3' 8 e a. £P fe ct o e B' ct Cl  M  p o o 3  CO o B>-O TJ P ct P-  £ ct ET §• 3 3*  2 o 2. Cl ts -a ft  Cu 13 T3  8  e  |!T  °" e-  3--0 % -< 2 H* O. 2 5 ct K- 5-  (H  Cu  "3ct 3  o a I' CD  a.  Cu  ta.  Cu  2 p.  to  [3  v o ctcr v vo o Crct  > — 00  t CD  vo £> vo sr EL  i c  vo  n  ~J rt  o  p m Bj cT Q  o  3-  I—• •  o  g  O 3  3  X  X  ft (t  Cu O  o EL  3  *3 CD  Ct Cu  X  o 3 O oo ^ BL I ^  X  a t gp Cl 3  X  s e c  O 8, cT  X  D  B* §.  to 3  CD  Cu^  II  O  a* a  ct  S  co  3  DO O 3 6,3 § g  8"  CD Cu  co  £ §s CD  2 o  X  CO  >  O DO  CD  8 I  CD 3 H3 B. CD  I Ct Cu  CD 3  •23? O Cu  X  X  X  C O B EL  3  co  p  s. a 5 O  !  S 5" g cl S— CO  cj oo oo cTS EL Bi 2. 3  co  o Ct Cu  3  g.cT  Ct J2. CU N Ct  27  Variables DerivedfromTheoretical Framework For purposes of this study, breastfeeding is viewed as one aspect of maternal role acquisition, for those women choosing to breastfeed (Fowles, 1998). Reliance on Mercer's (1981; 1995) theoretical framework for this study is consistent with the research findings of Flagler Virden (1988b) who reported a mother's perception of success with infant feeding was taken as an assessment of one's ability to mother, especially among primiparous women. Variables derivedfromMercer's work for use in this study included maternal competence, breastfeeding confidence, satisfaction, fatigue, and postpartum depression. Maternal Competence Maternal role competence, defined by Mercer and Ferketich (1994b, p.38) as "the mother's skills and interactions in the care of the infant that promote the infant's development," is acquired over time and is subject to the motivation of the learner, access to timely resources, and positive reinforcement. Many first-time mothers report feelings of inadequacy associated with inexperience or inconsistent preparatory information (Curry, 1988; Kirgis, 1988). In an effort to resume equilibrium, the mother must learn new tasks, and confront fears of how she will perform in her new role. According to Gibaud-Wallston and Wandersman (1978) new mothers who perceive their baby as "relatively easy to care for report more overall sense of competence as parents and more satisfaction in the role than those perceiving their baby as more difficult to manage"(p. 11). Variables that have been significantly related to maternal competence include age, previous infant care experience, self-esteem, social supportfrompartner, and the occurrence of maternal depression (Flagler Virden 1988a; Gibaud-Wallston & Wandersman;  28  Mercer & Ferketich). Mercer and Ferketich (1994b) studied high (n=121) and low (n=182) risk women at one, four, and eight months post birth to determine whether the two groups would differ in terms of perceived maternal competence, and whether predictors of maternal competence would differ for the two groups over time. By four and eight months after birth, maternal competence scores had increased compared to the one-month scores in both groups. Self-esteem and mastery were consistent predictors of maternal competence for both groups. Breastfeeding Confidence Confidence has been cited as an integral, affective component of maternal role acquisition (Zahr, 1991; 1993). In one study, a mother's past experience with infants and her prenatal expectations about her ability to understand her baby's signals predicted how confident and attached she was to her baby during the postpartum period (Williams, 1997). O'Campo et al. (1992) studied 11 psychosocial, demographic, and medical antenatal factors and found maternal confidence to be within the top five factors affecting the anticipated length of breastfeeding. Women with low levels of confidence about their ability to breastfeed were at 3.1 times greater risk of weaning early, compared to women with high levels of confidence about their ability to breastfeed. Furthermore, Buxton, Gielen, Faden, Brown, Paige, and Chwalow (1991) found that 27% of the women with low confidence in the antenatal period discontinued breastfeeding within the first postpartum week, whereas only 5% of the highly confident women stopped breastfeeding that soon after the birth. Bullock and Pridham (1988) conducted interviews at 30 and 90 days postpartum with 49 mothers of healthy newborns to learn about  29  about sources of confidence or uncertainty. Infant mood was most frequently reported as either a source of confidence, or uncertainty, at both test times. The infant's response to care as a source of confidence was positively related to perceived maternal competence (Bullock & Pridham). Following a review of relevant literature, Dennis (1999) developed a conceptual framework pertaining to breastfeeding confidence derived from self-efficacy theory. Selfefficacy theory suggests that positive perceptions about one's abilities results in one being more likely to successfully engage in a given behavior (Bandura, 1977). Self-confidence in one's ability to successfully undertake a health promoting task, such as breastfeeding is, therefore, likely to enhance breastfeeding duration (Dennis). Maternal Satisfaction Farber (1997) observed whether or not breastfeeding is perceived as satisfactory in the early postpartum period may be enhanced by antenatal education. Matthews (1991) concluded "a good start to the breastfeeding relationship often reinforces the mother in her decision; likewise, difficulties with breastfeeding at this time creates anxieties that undermine the mother's self-confidence, and ultimately decrease her motivation to continue breastfeeding" (p.49). Cooke, Sheehan and Schmied (2003) recommend further research to determine how breastfeeding satisfaction may be improved even if a woman experiences breastfeeding related problems. Cooke et al. studied physical breastfeeding experiences such as painful nipples or engorgement in relation to satisfaction but recognize infant behavior and social experiences may also influence breastfeeding satisfaction, and therefore ultimately, breastfeeding duration. Leff, Gagne, and Jefferis (1994) argued that if health care professionals continue to  30  define what constitutes "successful breastfeeding" only by the number of days a woman breastfeeds, they may undermine a woman's self-confidence and ultimately her self-esteem. To broaden their understanding of what constitutes successful versus unsuccessful breastfeeding from the perspective of breastfeeding women, the Leff team conducted interviews with 26 women. Their data revealed that women described breastfeeding as the natural means of sustaining the infant, and that when the experience is perceived positively, breastfeeding can be deeply satisfying and confidence building. Five major categories of successful breastfeeding were identified through use of constant comparative analysis including; infant health, infant satisfaction, maternal enjoyment, desired maternal role attainment, and lifestyle compatibility (Leff et al.). The researchers summarized the findings by stating, "...when the infant is not satisfied, breastfeeding success is relatively low, but if the infant is satisfied, high levels of success also may depend upon maternal enjoyment and attainment of desired maternal role" (Leff et al., p. 103). A sense of rhythm, or a "give and take" aspect of breastfeeding, requires cooperative efforts of each member of the dyad (Leff et al.). This description of mother/infant cooperation resembles findingsfromthe earlier work of Hewat and Ellis (1984) who proposed the concept of reciprocity or integration and negotiation between mother and infant as they pursued the task of concurrently learning the art of breastfeeding together. Fatigue Fatigue is one of the most common complaints of new mothers (Carty, Bradley, & Winslow, 1996; Gardner, 1991; Smith, 1989). Milligan (1989) reported that women who breastfeed report greater fatigue than women who bottle feed their infants and that finding ways  31  to conserve energy levels may prevent premature weaning. Since then, Milligan and colleagues have pursued both qualitative and quantitative inquiries to examine new mothers' characterizations of their fatigue, and to compare postpartum fatigue with postnatal depression (Milligan, Lenz, Parks, Pugh, & Kitzman, 1996; Milligan, et al., 1997; Pugh & Milligan, 1993). They concluded postpartum fatigue is multidimensional, and includes both physical and mental aspects that are different from tiredness or postpartum depression, although there may be common elements. Negative psychological states such as depression or somatic anxiety have been noted to exist among those who are fatigued. Caltabiano and Caltabiano (1996) claimed sleep deprivation coupled with the multiple roles of women, may lead to physical exhaustion among new mothers. Furthermore, they argued that while physical exhaustion and postnatal depression are related, they are mutually exclusive. Using a convenience sample of expectant couples (n=44), Elek, Hudson, and Fleck (2002) examined mothers' reports of fatigue prior to and after the birth of their first children. Women reported significantly more fatigue one month after birth compared to prior to birth. Gardner studied 35 women to determine fatigue levels until six weeks after birth and reported the initial two weeks after birth to be the most critical time for maternal recovery. McVeigh (1997) surveyedfirst-timemothers (n = 79) at six weeks postpartum. Mothers commented no one had prepared them for the unrelenting demands of infant care, level of fatigue they would experience, loss of personal time, and the realities of "around-the-clock" care. Following content analysis of the responses, McVeigh reported the mothers' written comments suggested that "a conspiracy of silence" seemed to exist regarding the (unspoken) realities of  32  motherhood. Twenty-two (28 %) of the women interviewed reported fatigue, or being tired, as a major descriptor of their postpartum experiences. Fatigue also was cited as a major physical concern by 38 (48 %) of the participants. When asked to elaborate how fatigue had affected them, the women reported an adverse effect on various aspects of their lives, including household responsibilities, and relationships with others. Fatigue had impacted their ability to continue with aspects of their life they had previously valued. Wambach (1998) examined the effect of maternal fatigue levels among primiparous breastfeeding women until nine weeks after birth. Using a prospective, correlational design to determine patterns that existed among select physiological, psychological, situational, and performance factors, she determined fatigue to be moderately significant for first-time breastfeeding mothers. Wambach reported fatigue positively correlated with breastfeeding problem severity when compared at four different times and that "...increased severity within specific categories of problems were associated with higher levels of fatigue" (p. 224). The finding that breastfeeding may be adversely affected by fatigue is consistent with the recommendations of Pugh, Milligan, Parks, Lenz, and Kitzman, (1999) who claimed that helping women to better understand their fatigue and how it may influence them, might aid in extending breastfeeding duration. Postnatal Depression It has been estimated that at least 13% of all mothers experience postnatal depression (Beck, 1998; O'Hara & Swain, 1996). According to Beck (2001), postpartum depression not only results in maternal suffering but, "can have more insidious consequences, such as disturbed  33  mother-infant relationships and impaired cognitive and emotional development of the children" (p. 275). Multiple studies have found a significant association between the onset of postpartum depression and breastfeeding duration (Bick et al., 1998; Papinczak & Turner, 2000; Susman & Katz, 1988; Taj & Sikander, 2003; Tammentie, Tarkka, Astedt-Kurki, & Paavilainen, 2002). Maternal depression limited how long a mother continued to breastfeed in a study conducted by Galler et al. (1999). Cooper et al.'s (1993) study observed a significant association between postnatal depression and premature cessation of breastfeeding by two months after birth (p < .001), but found no relationship between antenatal depression and breastfeeding outcomes. Fleming, Ruble, Flett, and Shaul (1988) found that a depressed mood at three months postpartum was associated with a significant reduction in breastfeeding. Similarly, Henderson, Evans, Straton, Priest, and Hagan (2003) reported in their cohort study (n = 1745) that women who experienced postnatal depression at any time during thefirstyear after their infant's birth had 1.25 times greater risk of having weaned earlier than women who were not depressed. Henderson et al. observed that the onset of depression occurred before cessation of breastfeeding as did Misri, Sinclair, and Kuan (1997). Hellin and Waller (1994) reported a combination of depression and anxiety at two weeks after birth to be predictive of breastfeeding cessation by two months after birth. In contrast, Jacobson, Jacobson, and Frye's (1991) correlational study did not find any relationship between breastfeeding and postnatal depression when comparisons were completed by six months after birth. Factors Associated with Breastfeeding Initiation and Duration Four categories proposed by Duckett, Henly, and Garvis (1993) were used to organize  34  variables that may assist in making predictions about breastfeeding duration. The four categories include maternal antecedents, internal resources, external resources, and interferences. Maternal Antecedents Maternal antecedents, or demographic variables associated with breastfeeding duration include age, education, marital status, smoking status, parity, and attendance at antenatal classes (Avery, Duckett, Dodgson, Savik, & Henly, 1998; Boettcher, Chezem, Roepke, & Adams, 1999; Janke, 1993; Scott & Binns, 1999). Women most likely to initiate and continue breastfeeding have been older, married, well educated, Caucasian, non-smoking, and not working outside the home (James, Jackson, & Probart, 1999; Ko & Schulken, 1998; Novotny, Hla, Kieffler, Park, Mor, & Thiele, 2000; Tarkka, Paunonen, & Laippala, 1999; Zinn, 2000). Smoking has been identified as a significant predictor of breastfeeding cessation (Amir & Donath, 2002). Women who attend antenatal classes, and have friends and family with breastfeeding experience to assist them are more likely to breastfeed longer (Health Canada, 2000). In contrast, women have reported return to paid employment outside the home as a reason for weaning by four to six months after birth (Health Canada). Women who bottle fed previous children are more likely to again bottle feed subsequent children, or to discontinue breastfeeding sooner. In addition, women who had difficulties with breastfeeding previous children are less likely to initiate breastfeeding, or to continue to breastfeed subsequent children (Hill, Humenick, Argubright, & Aldag, 1997). Internal Resources According to Duckett et al. (1993), internal resources include when the decision to  35  breastfeed was made, the reasons for breastfeeding, and perceived success of early breastfeeding opportunities. As stated earlier, the timing of the decision and the intention to breastfeed have been cited as predictors of breastfeeding duration (Coreil & Murphy, 1988; O'Campo et al., 1992; Papinczak & Turner, 2000). Using logistic regression procedures, Buxton et al. (1991) identified four variables as significant predictors of failure to breastfeed for more than seven days. The variables included lower confidence in ability to breastfeed, less certainty in the decision to breastfeed, delayed initiation of breastfeeding, and lack of rooming in with the baby. Breastfeeding failure, defined for purposes of their study as weaning prior to seven days after birth, was approximately five times more likely among less confident and less certain women. Using a phenomenological approach, Bottorff (1990) studied what breastfeeding, and continuing to breastfeed, is like for mothers. Factors were explored including the mother's relationship with the baby, her commitment to breastfeeding, breastfeeding problems she may have encountered, and decisions she made about when to wean. The study revealed that feeling committed to the idea of breastfeeding is an important part of continuing to breastfeed, as is being persistent even when challenged by breastfeeding related problems. The author further determined being persistent related to benefits for the baby, and an affirmation that the woman was meeting her responsibilities of motherhood despite adversity. Similarly, Hauck and Reinhold's (1996) descriptive study identified four categories to determine breastfeeding success that included giving, persisting, meeting expectations, and accomplishing personal goals. Mozingo, Davis, Droppleman, and Merideth (2000) interviewed women (n=9) who weaned their infants within two weeks of birth. Findings revealed women felt uninformed about  36  the realities of becoming a breastfeeding mother. Women who participated in the study not only described incongruence between their expectations and the reality of the first weeks of breastfeeding, but also reported that the incongruence led to breastfeeding cessation. The authors appealed to health care professionals to inform women during the antenatal period, as to how they can better prepare for the realities of the early postpartum experience. External Resources External resources include breastfeeding encouragement and sources of information (Duckett et al., 1993). Although, the type and amount of social support will vary for each mother (Arlotti, Cottrell, Lee, & Curtin, 1998; Bar-Yam & Darby, 1997; Kuan, Britto, Decolongon, Schoettker, Atherton, & Kotagal, 1999; McNatt & Freston, 1992; Raj & Plichta, 1998), partner support to breastfeed has been cited as a pivotal factor to extend breastfeeding (Freed et al., 1992; Freed & Fraley, 1993; Gamble & Morse, 1992; Jordan & Wall, 1990; Sears, 1992). When a woman's husband or partner supported her decision to breastfeed, and understood the health benefits of breastfeeding for the infant and mother, the woman was more likely to initiate and continue breastfeeding (Kessler, Gielen, Diener West, & Paige, 1995). Women who received support, and those who perceived others supported their decision to breastfeed, were four times more likely to breastfeed in a study conducted by Chezem, Friesen, and Clark (2001). The literature supports inclusion of those persons the woman deems to be significantly involved in the infant's care in research or practice initiatives intended to enhance continuance of breastfeeding. In a study conducted to review the sources of infant feeding information used by  37  pregnant women, women were four times more likely to initiate a conversation about infant feeding methods with a family member or friend, than with a health care provider (Chezem et al, 2001). This finding prompted the authors of that study to recommend the inclusion of key individuals in antenatal education classes and support programs as a powerful strategy. Furthermore, interventions need to include encouragement to clarify the actual attitudes and perspectives among those the woman deems most influential in her decisions about infant and child feeding and to eliminate perceived ambivalence or non-support of a woman's breastfeeding behaviors (Arora et al., 2000). Studies conducted to assess the effect of professional support examined an array of health providers including lactation consultants (Auerbach, 1985; Hamelin, 2000; Lynch, Koch, Hislop, & Coldman, 1986), pediatricians (Serwint et al., 1996), midwives (Jamieson, 1990), medical students (Chan-Yip & Kramer, 1983), and public health nurses (Field & Renfrew 1991; McKeever et al., 2002). Sikorski and Renfrew's (1998) Cochrane Review concluded professional input prolongs breastfeeding, but not necessarily exclusive breastfeeding. Humenick, Hill, and Spiegelberg (1998) reported health professionals are in a key position to enhance breastfeeding success by providing timely, consistent, and culturally sensitive forms of support, particularly among first-time mothers. Porteous, Kaufman, and Rush (2000) found that women who were assigned to individualized professional support that was commenced in hospital and then continued once home, breastfed significantly longer (p = .005). A study conducted to evaluate the effect of peer counselor or lay supports on breastfeeding initiation and duration concluded lay support providers were considered as  38  effective as health care professionals (Mongeon & Allard, 1995). Fairbank et al.'s review (2000) noted that women who had received peer support reported they believed all women should be encouraged to seek peer support if they planned to breastfeed. Interferences Interferences, the fourth category proposed by Duckett et al. (1993), entails discouragement by support persons, breastfeeding problems, or in-hospital practices such as supplementation with formula. Breastfeeding duration studies seeking to determine the effect of labor analgesia, timing and type of initial feedings, type of delivery, length of hospital stay, and birth setting have had inconsistent results (Auerbach, 1988; Barros, Victora, Semer, Tonioli Filho, Tomasi, & Weiderpass, 1995; Bernard-Bonnin, Stachtchenko, Girard, & Rousseau, 1989; Carty & Bradley, 1990; Chapman & Perez-Escamilla, 1999; Dungy, Losch, & Russell, 1994a; Halpern, Levine, Wilson, MacDonell, Katsiris, & Leighton, 1999; Heck, Schoendorf, Chavez & Braveman, 2003; Hill, Humenick, Brennan, & Woolley, 1997; Janke, 1988; Janson & Rydberg, 1998; Perez-Escamilla, Pollitt, Lonnerdal, & Dewey, 1994; Riordan, Gross, Angeron, Krumweide, & Melin, 2000; Wright, Rice, & Wells, 1996). A meta-analysis of nine studies conducted by Bernard-Bonnin et al. to review the effect of hospital practices on breastfeeding duration concluded supplementation had a negative effect on duration while early contact with the infant demonstrated a significantly positive effect. A meta-analysis of infant feeding policies in maternity wards and their effect on breast-feeding success conducted by Perez-Escamilla et al. concluded commercial discharge packs had an adverse effect on lactation, but the impact of early mother-infant contact was unclear. The review team suggested it is necessary to learn more  39  about the potentially synergistic effect of hospital practices in combination with prenatal and postnatal breastfeeding interventions in order to design cost-effective strategies aimed at increasing lactation success. In the five studies reviewed by Renfrew and Lang (1994), no clear benefit of early breastfeeding initiation practices on breastfeeding duration was demonstrated. Although it has been suggested that support for breastfeeding by the woman's partner or spouse can be key to continued breastfeeding, it also has been reported that a partner or significant support person may negatively influence the woman's intentions to continue to breastfeed. In a study conducted in the United States (n=245), the number one reason why mothers stopped breastfeeding their babies was because they believed their husbands opposed the idea of their continuing (Arora et al., 2000). Forty percent of the 245 new mothers reported worries about their husbands as their primary reason to wean. The second reason pertained to negative influence about breastfeeding from the women's mothers. Only 13% of the women in Arora et al.'s sample were still breastfeeding by six months after birth. A key observation was that the women discontinued breastfeeding because they perceived their husband did not approve of them breastfeeding in public, a perception that was not consistent when the researcher questioned the respective husbands. Arora et al. thus encouraged inclusion of expectant fathers and grandmothers (who may not have breastfed their babies) during future antenatal information sessions and urged the women (couples) to share, and clarify, their breastfeeding beliefs and concerns. More formal sources of support, such as the care received from health professionals, also may have a negative effect on breastfeeding duration. Inconsistent or inaccurate breastfeeding  40  information has been one of the most frustrating and problematic components reported by mothers, according to local breastfeeding hotline statistics (S. Steckler, personal communication, June, 2000). Similar concerns have been reported in the literature (Coreil, Bryant, Westover, & Bailey, 1995; Humenick et al., 1998). Summary Breastfeeding and breast milk confer significant health benefits. The mean duration rate for breastfeeding among Canadian women remains low. Mercer's (1981; 1995) theoretical framework and the literature review confirm the complexities of the breastfeeding process, yet emphasize strategies and factors that may be useful in helping women continue to breastfeed. Studies have noted common predictors of breastfeeding duration to include maternal age, smoking status, and an early return to paid employment. However, none of these factors may be easily amenable to change. A variety of perinatal intervention studies have been conducted to affect breastfeeding outcomes and although complex, multilevel interventions, offered before and after the birth of an infant, have been viewed as effective, they are logistically difficult to plan, and expensive to implement. Findings in the literature revealed the potential benefit of small, informal group health education classes to promote breastfeeding duration, and the need to conduct a randomized controlled trial using sufficient power and an intention-to-treat analysis to study such interventions. In response to these recommendations, this study tested the effect of an innovative antenatal intervention on breastfeeding duration. No studies were identified that had asked women (and their partners) to prepare for a group session by identifying strategies they thought  41  would be helpful for them, particularly during the initial postpartum period. Therefore, the intervention not only built on the apparent success of small interactive groups noted in previous research, but was expanded to include a pre-workshop guide intended to encourage women (and their partners) to discuss their breastfeeding expectations and concerns, prior to and during a group workshop attended by other expectant couples.  42  CHAPTER TWO METHOD This chapter presents the study design, research questions, sample selection criteria, recruitment procedures, and intervention summary. Information regarding study instrumentation, ethical considerations, study strengths and limitations, and the projected plan for data analysis also is included. Design A randomized controlled trial of an antenatal educational intervention was conducted in Winnipeg, Manitoba among primiparous women who were planning to breastfeed their infants. The intervention group (n=l 11) received usual care, plus a self-assessment pre-workshop guide and an interactive, educational, antenatal workshop. The control group (n=98) received usual care. The purpose of the investigation was to test how long the women assigned to the intervention group breastfed their infants compared with the women assigned to the control group. Primary Research Question 1. Do women assigned to the intervention group at least mainly breastfeed* longer than women assigned to the control group? (* mainly breastfeeding defined as ingestion of breast milk plus up to one alternate feeding, such as commercially prepared formula, per day).  43  Secondary Research Questions 1. Do women assigned to the antenatal intervention report more or less maternal competence than women in the control group as measured by the Parenting Sense of Competence (PSOC) Scale? 2. Do women assigned to the antenatal intervention report more or less breastfeeding confidence than women in the control group as measured by the Breastfeeding Self-Efficacy Scale (BSES)? 3. Do women assigned to the antenatal intervention report more or less satisfaction about their breastfeeding experience than women in the control group as measured by the Maternal Breastfeeding Evaluation Scale (MBFES)? 4. Do women assigned to the antenatal intervention report more or less fatigue than women in the control group as measured by the Multidimensional Assessment of Fatigue (MAF) Scale? 5. Do women assigned to the antenatal intervention report more or less postpartum depressive symptoms than women in the control group as measured by the Edinburgh Postnatal Depression Scale (EPDS)? 6. What are the predisposing factors associated with breastfeeding duration? Procedures A randomized controlled trial was selected to eliminate selection bias. Randomization procedures were completed with the assistance of the Manitoba Nursing Research Institute (MNRI). Prior to recruiting any participants, cards marked with either a red or black color code were placed in separate opaque and sealed envelopes. The envelopes were then placed upright in a row in a container so that each time an envelope was to be withdrawn, a MNRI research  44  assistant (blinded in terms of study purpose) could select any one of the envelopes. Each time a participant was recruited to the study, the research assistant randomly selected one sealed envelope to determine the group to which the participant was to be assigned. A number was then written on the envelope to coincide with the participant's identification code. The contents of the envelope (i.e. either a card with a black mark to indicate the control group or a card with a red mark to indicate the intervention group) also were recorded with the participant's identification code on a master sheet. The group assignment was recorded on the participant's consent form. All documentation related to the randomization procedure was stored in a locked filing cabinet. Following recruitment of all participants, the investigator checked the master sheet to confirm every participant assigned to the intervention or control group had been appropriately notified. The investigator also confirmed that only participants assigned to the intervention group attended the workshop by comparing the master sheet with a workshop sign-in sheet. Sample The study used a non-probability, convenience sample of primiparous women recruited from prenatal classes. Approximately 4000 primiparous women deliver infants in Winnipeg, Manitoba each year. Three adjacent community areas within the City of Winnipeg were selected for recruitment because they were demographically homogeneous and had relatively high breastfeeding initiation rates. According to an analysis conducted by Statistics Canada of the NLSCY (1996) data on breastfeeding initiation and duration rates across Canada, up to 85% of Manitoba women initiated breastfeeding (see Figure 1). However, according to the same source, only 36% of these women continued to breastfeed six months after birth.  45  Figure 1. Percent of all Manitoba Infants Breastfed. (NLSCY, 1996) 100%-  j  2  4  6  8  10  15% breastfed at 1 year  12  14  16  Age o f infant (months) The three geographic areas in Winnipeg also were selected because the content and approach used in the delivery of antenatal classes were deemed common by the investigator after evaluating the curriculum and observing the classes related to breastfeeding. All the classes were taught by a small group of instructors (n=5) who shared teaching sites. The teachers estimated at least 90% of women who attended their classes were primiparous, and had indicated they planned to breastfeed. Sample size was projected a priori based on a power calculation. The estimation of a medium effect (i.e. a 20% increase in the proportion of women still at least mainly breastfeeding by 24 weeks after birth) was deemed clinically significant and reasonable to attain by the researcher. The estimation also relied on previous research in that three trials proposed increases  46  in the proportion of women ranging from 10-25 % as the criterion to determine if a medium effect had been achieved (Redman, Watkins, Evans, & Lloyd, 1995; Serwint et al., 1996; Sheehan, 1999). Eighty percent (.80) power was selected to decrease the risk of a Type II error. Because it was unknown whether the women assigned to the intervention group would breastfeed for as long as women assigned to the control group, a two-tailed alpha value was used. A sample of 87 in each group, (or a total of 174 women) was required to achieve power of .80 using a two tailed alpha at the .05 testing level (Pocock, 1983). To account for a projected attrition rate of approximately 10%, a minimum of 95 women were recruited for each group.  Participant Inclusion Criteria Eligible women for the trial were: •  first-time mothers who planned to breastfeed,  •  registered to attend a series of antenatal education classes,  •  a minimum of 18 years of age,  •  in their third trimester of pregnancy with a singleton fetus,  •  able to read, write, speak, and understand English,  •  willing to be contacted by a research assistant three times, and  •  able to attend an additional antenatal education class.  47  Participant Exclusion Criteria Ineligible women for the trial were those who: •  were experiencing self identified major health concerns or a history of clinical depression, and  •  had a midwife as a primary caregiver because the scope of practice includes extensive breastfeeding support.  Recruitment Procedure Subjects were recruited during antenatal education classes held in Winnipeg, Manitoba from March 2001 until June 2002 (see Table 3). Notices were posted in the antenatal classroom areas and mailed to class registrants. The investigator attended the fourth prenatal class in every series over the study recruitment period to present the details of the study. Expectant mothers (and their support persons) were invited to participate in the study. Those who were interested in participating remained after class and, after eligibility was assessed, their consent was obtained. No information was obtainedfromthose women who were eligible to participate in the trial, but who chose not to participate. A schematic model of the study design is depicted in Figure 2. Approximately 450 women were invited to participate in the study (see Figure 3). About 200 women either stated they were not interested in participating, or did not remain after the class to hear more about the study following the initial description and invitation to do so.  48  Table 3. Study Timeline Year/ Quarter  2001 12  3 4  2002 12  3 4  Receipt of Ethical Approval  X  Receipt of Access Approval  X  Organization of Recruitment Sites  X  Recruitment  X X X  X X  Data Collection and Management  X X X  X X XX  Data Analysis Summary of Findings Sent to Participants  2003 12  3 4  X X X X  Thirty of the remaining women were ineligible to participate because they were either not yet sure if they planned to breastfeed (BF), had previously had children, were expecting twins, or had a history of depression. Two hundred and nine eligible women indicated their willingness to participate by completing the consent form. All women who signed a consent form were asked to complete and submit to the investigator a copy of the antenatal personal information form (Appendix III) designed by the investigator for use in this study. Participants were then given four packages of questionnaires. Each package contained a copy of the instruments (Appendices IV-VIII) to be completed at each of four times, and stamped envelopes addressed to the researcher. One package was to be returned within one week of recruitment to the study, and the remaining packages were to be mailed at 2, 12, and 24 weeks after birth.  Figure 2. Study Design Primiparous Women Planning to Breastfeed Recruited from Series of Antenatal Classes  Eligibility Criteria Reviewed & Written Consent Obtained at Antenatal Class Baseline Personal Data Collected at Antenatal Class. Mail-in Questionnaires sent to Principal Investigator within One Week of Recruitment Random Assignment of Each Subject Control Group  Intervention Group  Usual Antenatal Care in Hospital and Community  Usual Care in Hospital and Community PLUS receipt of Pre-Workshop Guide and Attendance at a Group Workshop  Usual Postnatal Care Telephone Questionnaire at 2, 12, and 24 weeks after delivery by a research assistant (blinded as to group)  Usual Postnatal Care Telephone Questionnaire at 2, 12, and 24 weeks after delivery by a research assistant (blinded as to group)  Mail-in Questionnaires sent to principal investigator at 2, 12, and 24 weeks after delivery  Mail-in Questionnaires sent to principal investigator at 2, 12, and 24 weeks after delivery  I Outcomes Analyzed.  I Outcomes Analyzed  50  Figure 3. Flow Diagram - Trial Recruitment Outcomes Invited to Participate 11=450  s  Approximately 200 women indicated they were not interested in participating or indicated they did not meet the inclusion criteria  r  E  S3  Excluded (n=30) not sure about BF decision multiparous expecting twins history of depression  Assessed for Eligibility n=239  o bl s W  Randomized (n=209)  a w e  o a o fa  Allocated to Intervention (n=l 11) Attended group workshop (n=89) Did not receive intervention (n=22) Reasons: Not available to.attend or delivered infant prior to session  Lost to follow-up (n=5) Reasons: Not able to contact Withdrewfromparticipation (n=10), Reasons: No longer willing (n=9), infant prematurity (n=l)  1  3  Number of participants continuing to participate at: 2 wks (n=107) (97%), 12 wks (n=101) (91%), and 24 wks (n=96) (86.5%)  Included in Survival and Multivariate Analysis (n=l 11)  Allocated to Usual Care (n=98)  Lost to follow-up (n=4) Reasons: Not able to contact Withdrewfromparticipation (n=12), Reasons: no longer willing (n=10), perinatal loss (n=l), postnatal complications (n=l)  1  Number of participants continuing to participate at: 2 wks (n=89) (91%), 12 wks (n=84) (85.8%), and 24 wks (n=82) (83.7%)  Included in Survival and Multivariate Analysis (n=98)  51  On the next working day after recruitment, a research assistant from the MNRI, not otherwise associated with the study, assigned the subject to a group using the predetermined randomization procedure. The principal investigator notified all participants of their group allocation by telephone. Those assigned to the intervention group were mailed information, including the pre-workshop guide, and information about when to attend the group workshop. Participants were subsequently contacted by telephone at 2, 12, and 24 weeks after birth by a second research assistant (blinded as to group allocation) to collect infant feeding data using questionnaires developed by the investigator for this study (Appendices IX-XII). To control for contamination, women (and their partners) who were assigned to the intervention group were asked, when telephoned about their group allocation, not to share information obtained in the preworkshop guide, or during the group workshop, with participants assigned to the control group. The reason the two groups were not equal in size at baseline, (111 women in the intervention group versus 98 women in the control group), was related to a decision made by the administration of the regional health authority to discontinue prenatal classes at the largest recruitment site used in this study. At that point, the minimum number of women for each group had been recruited. At the same time, 21 of the original 230 recruitment envelopes remained unopened. The result was that the number of women allocated to the control group was smaller from the onset of the study, albeit sufficient to satisfy the minimum number of women needed (n=87).  52  The Intervention The intervention was designed to complement and extend breastfeeding content already taught in a series of antenatal education classes (Appendix XIII). The investigator observed existing antenatal classes on breastfeeding and reviewed documentation outlining the scope of topics to be presented. Content pertaining to infant feeding comprised approximately 2 of 16 hours of total prenatal class instruction time, although there were some minor variations depending on class size and instructor preference. Using the existing content as a starting point, the investigator designed the intervention to focus on strategies intended to prevent or minimize the reasons women give for weaning their infants and to encourage those strategies that have been associated with longer breastfeeding duration. Women assigned to the intervention group attended the group workshop after they had received the (usual care) content in the prenatal class series pertaining to breastfeeding. The intervention involved a written pre-workshop guide in booklet form, and attendance at an antenatal group workshop. The pre-workshop guide, developed by the investigator and a colleague (Kluka & LeDrew, 2000) (Appendix I), was mailed to the women prior to the interactive group workshop. The guide contained information on important aspects of breastfeeding and questions for self-reflection about the woman's expectations and coping styles. Expectant mothers were encouraged to share their thoughts and feelings about breastfeeding, and complete the guide with their partner or support person. The guide was designed as a workbook in that space was allocated for women to write down individual strategies that they perceived as useful to prevent or minimize problems they may experience in the early postpartum period  53 while breastfeeding. For example, parents were asked to list how they could plan to relinquish household responsibilities to devote more time and energy to infant care. Space was also available for women to list alternatives they may want to consider such as taking a warm bath if this is what they felt would be relaxing and important to them. The ninety minute interactive group workshop (Appendix II), facilitated by a perinatal nurse who was also a lactation consultant, addressed problems parents might have experienced, explored alternatives about how to prevent or minimize the problems, and urged participants to share strategies or ideas from their pre-workshop guides. The workshops were held on Saturday mornings, a time the participants suggested would be most appealing. The location was a university campus classroom and lounge. The parking was free and convenient. Partners or support persons were invited and refreshments were available during a fifteen-minute break. A slide presentation was used to highlight content and present strategies in order to encourage participants to consider their responses and share ideas with others in the group. The nurse facilitator encouraged participants to explore strategies for managing initial challenges. The facilitator had been trained by the investigator to conduct the workshop in a similar manner each time. Time was offered during the class, the break, and after the class for more individualized dialogue among participants and the facilitator. Instrumentation To select the most appropriate measures, the researcher reviewed published breastfeeding assessment tools considered to be appropriate to measure the identified concepts. Guided by the literature review and theoretical framework, instruments that measured maternal competence,  54  breastfeeding confidence, satisfaction, fatigue, and depression were selected. The investigator constructed measures to determine breastfeeding duration, demographic variables, and delivery characteristics for study purposes (Appendices III, IX-XII). Duration ofBreastfeeding Each participant was contacted by telephone at 2, 12, and 24 weeks after birth by a research assistant (RA) to collect infant feeding data using instruments developed by the investigator (Appendices IX-XII). Five operational categories were used to measure breastfeeding duration. The mutually exclusive categories encompassed what was fed to the infant (type, volume, frequency, or combination thereof), method of feeding (breast, gavage, or bottle), and by whom the infant was fed (Ellis, Hewat, & Livingstone, 1991). Each of the following includes optional use of vitamin supplements or prescribed medication: 1) Exclusively Breastfeeding - Only Breast milk or Expressed Breast Milk (EBM). 2) Primarily Breastfeeding - Breast milk or EBM, plus up to one alternate feeding per week. 3) Mainly Breastfeeding - Breast milk or EBM, plus up to one alternate feeding per day. 4) Partially Breastfeeding - Breast milk or EBM, but mainly alternate feedings per day. 5) Weaned - No Breast milk or EBM for past week, and no plan to resume breastfeeding. Parenting Sense of Competence (PSOC) Scale Gibaud-Wallston and Wandersman (1978) developed the PSOC scale (Appendix TV). The scale assesses parents' perceptions of their abilities to handle the situational demands of parenting. Seventeen items form two subscales. The first subscale is entitled "Skill/Knowledge" and it assesses the parents' perceptions of the degree to which they have acquired the skills and  55  understanding to be a good parent. The second subscale entitled "Valuing/Comforting" assesses the degree to which the individual values parenthood and is comfortable in that role. Each of the 17 items is rated by parents on a 6-point scale from "strongly disagree" to "strongly agree." Possible scores can range from 17 to 102. The combined score provides an overall appraisal by a parent of his or her own functioning in the parenting role. Higher scores indicate a higher sense of competence. Psychometric properties of the scale were assessed using responses of first-time parents (n=132); the subgroup of mothers comprised 66 women (Gibaud-Wallston & Wandersman, 1978). Item analyses produced alpha coefficients of .82 for the Skill/Knowledge subscale and .70 for the Valuing/Comfort subscale. The Cronbach's alpha coefficient for the total scale was .83. Pearson's r-values were significant at the p< .01 level and rangedfrom.46 to .82 when responses among participants were compared when the infant was six weeks and six months of age. Convergent validity was assessed by anticipating a positive correlation between age of the baby and the parent's sense of competence, especially among the Skill/Knowledge subscale items. The mothers' responses at six weeks and six months after birth did correlate with their babies's age as expected for the Skill/Knowledge sub-score and total score, but not for the Valuing/Comfort sub-score. To further test convergent validity, scores of the PSOC were compared with scores on a Personal Feelings Scale and items of a General Well-Being Scales. Positive correlations achieved by the testing were significant (p < .05) (Gibaud-Wallston & Wandersman, 1978).  56  Mothers who perceived their baby as "difficult" also reported significantly lower PSOC scores than those who perceived their baby as "easy." Furthermore, mothers with low social support felt significantly less competent as parents than mothers with more support (GibaudWallston & Wandersman, 1978). Both sets of results lend further support to the scale's validity. The PSOC has been used to compare maternal-infant attachment among experienced and inexperienced mothers at one, four and eight months postpartum (Mercer & Ferketich, 1994a). A significantly positive relationship was observed between maternal competence and maternal attachment at each of the three times. However, when perceived maternal role competence was compared between high-risk (i.e. hospitalized during pregnancy for a risk condition) (n = 120) and low risk women (n=182), Mercer and Ferketich (1994b) reported no significant differences between the groups over time. Maternal role competence scores were higher for both groups at four and eight months than at one month after birth. Breastfeeding Self-Efficacy Scale (BSES) The instrument used to measure breastfeeding confidence was the BSES (Dennis & Faux, 1999) (Appendix V). Content for the 33 items was derived from literature pertaining to breastfeeding problems and factors related to breastfeeding duration. A Likert-type scale of one to five indicates degree of confidencefrom"not at all confident" to "always confident." Possible scores rangefrom33 to 165 and a higher score is indicative of more confidence. Using data collectedfromCanadian women, Dennis and Faux (1999) have completed assessments of content validity, internal consistency reliability, construct validity, and predictive validity. Testing revealed a Cronbach's alpha coefficient of .96 for the total scale (n=130).  57  Dennis and Faux (1999) reported "the BSES was assessed for construct validity using three methods: factor analysis, comparison of contrasted groups, and correlations with measures of theoretically related constructs" (p. 403). Factor analysis resulted in three subscales with items pertaining to breastfeeding technique, maternal attitudes and beliefs, and maternal activities to aid breastfeeding. However, Dennis and Faux later deleted the five items that comprised the third factor as they accounted for only 5.1% of the variance. Two additional examinations of the tool further supported construct validity according to Dennis and Faux. Using contrasted groups represented by women who had breastfeeding experience with those who did not, significant differences in breastfeeding self-efficacy were determined among primiparous (M = 159.0, SD = 25.31) and multiparous women with breastfeeding experience (M = 179.2, SD = 20.53) (t = .9, p < .001). As well, the BSES positively correlated with a tool measuring a similar construct, i.e. a measure of achievement tendency motivation (r = 35, p < .001), but correlated negatively with a measure of general self-efficacy (r = -2.0, p = .03). Dennis and Faux suggested support for predictive validity was achieved in subsequent testing (n=l 19) since the higher the BSES score, the more likely the mothers were still exclusively breastfeeding at six weeks (F = 9.89, p < .001). Preliminary psychometric testing, therefore, suggested that the BSES is a reliable and valid means of determining how confident a mother is feeling about her breastfeeding experience at a given time. Maternal Breastfeeding Evaluation Scale (MBFES) Maternal satisfaction with the breastfeeding experience was measured by the MBFES as developed by Leff, Jefferis and Gagne (1994) (Appendix VI). The research team categorized  58  qualitative data into five categories of successful breastfeeding: infant health, infant satisfaction, maternal enjoyment, desired maternal role attainment, and lifestyle compatibility (Leff, Gagne & Jefferis, 1994). Items on the scale, which were derived from the qualitative data, each identify indicators of breastfeeding success from the perspective of breastfeeding women. Pilot testing procedures and preliminary psychometric testing (n=442) resulted in a 30 item scale with three factors accounting for 38.5% of the variance: maternal enjoyment/role attainment, infant satisfaction/growth and lifestyle/maternal body image (Leff, Jefferis, & Gagne). The measure consists of a Likert-type scale of one to five to indicate the extent of agreement from "strongly disagree" to "strongly agree." The range of possible scores extends from 30-150 and a higher score indicates greater satisfaction. Initial testing (n=442) of the scale resulted in a Cronbach's alpha correlation coefficient of .93 for the overall scale. The alpha coefficients for the three subscales were .93 for maternal enjoyment/role attainment, .88 for the infant satisfaction/growth, and .80 for lifestyle/maternal body image (Leff, Jefferis, & Gagne, 1994). Riordan, Woodley, and Heaton's (1994) testing (n=73) of the MBFES further supported the scale's internal consistency in that Cronbach's alpha coefficients for the scale and subscales were .94, .91, .83, and .84, respectively. Similarly, testretest procedures that were completed by Leff, Jefferis, and Gagne to assess the scale's stability over time, revealed coefficients of .93 for the overall scale, and .93, .94, and .82 for the three respective subscales. To assess the construct validity of the scale, correlations were completed with theoretically related constructs including the mothers' perceptions of overall breastfeeding  59  success, intended duration of breastfeeding, and actual duration of breastfeeding. Leff, Jefferis, and Gagne (1994) reported the scale and subscales correlated positively with the mothers' perceptions of overall breastfeeding success (.83, .79, .70, .55 respectively). Riordan et al.'s (1994) examination of the tool resulted in the following correlations between the MBFES scale and intention to breastfeed (.30, .25, .28, .28) and between the MBFES and actual duration of breastfeeding (.39, .34, .35, .31 respectively). Mothers, therefore, may describe their breastfeeding experience as having been highly successful even if they did not breastfeed for as long as others who may describe their experience as less successful. Riordan et al. also tested the three dimensions of maternal satisfaction by comparing correlations between the MBFES subscales and reported a range of correlations from "r = .56 for lifestyle/maternal body image with infant satisfaction/growth to r = .75" (p.233). Leff et al. (1994) noted testing of construct validity among diverse populations had not been completed. Multidimensional Assessment ofFatigue (MAF) Scale The instrument selected to measure fatigue was the MAF Scale (Belza, 1990) (Appendix VII).  Belza derived items for this scale by modifying the Piper Fatigue Scale (Piper, Lindsey, &  Dodd, 1989), a 41-item measure designed for use with and tested among oncology patients. The MAF scale contains 16 items and measures four dimensions of fatigue: severity, distress, degree of interference in activities of daily living, and frequency. Fourteen items pertaining to severity, distress, and interference are scaled from 0 to 10 with the ends anchored respectively by "not at all" to "a great deal." The final two items, pertaining to frequency of fatigue use a categorical scale (0-4) with labels identified for each choice.  According to Belza's (1990) guide for scale use, a global fatigue score is calculated by summing the severity and distress items, adding the item mean for the interference items, and adding the product of the categorical score on frequency multiplied by a constant (2.5). If interference in an activity is attributable to something other than fatigue, that item is not included in the total score. Similarly, if the response to fatigue severity is 0, none of the other items are completed and the total score is zero. Strengths of the MAF scale include the self-administered format, length of time required to complete the scale (less than 5 minutes), and inclusion of multiple items to allow a more stable measurement of fatigue (Belza, 1990). Initial testing of the tool by Belza, Henke, Yelin, Epstein, and Gilliss (1993) with a sample of 133 subjects yielded a Cronbach's alpha coefficient for the total scale of .93. In 1995, Belza refined the MAF Scale and tested patients with rheumatoid arthritis (n = 51) to compare outcomes matched for age and gender. The scale was further examined for concurrent validity through comparisons with the Profile of Mood States (McNair, Lorr, & Droppleman, 1971). According to Belza, outcomes indicated the MAF demonstrated concurrent (r =.84, p <.01) and divergent (r = -.62, p <.01) validity. However, stability testing resulted in correlations ranging from .73 to .47 when assessments were completed at three different times. Meek et al. (2000) included psychometric testing of the MAF scale with patients undergoing cancer treatment (n=147) and reported the instrument demonstrated good internal consistency reliability estimates, as measured by a Cronbach's alpha correlation coefficient of .88. Factor analysis of the MAF revealed only a two factor solution accounting for 63% of the variance, a finding that suggests the scale's items did not perform as well in examining the intended four dimensions of fatigue  61  when used with patients experiencing cancer. Wambach (1998) assessed maternal fatigue levels using the MAF scale among breastfeeding primparae (n=41) during the first nine weeks after birth. Testing demonstrated support of the scale's internal consistency in that Cronbach's alpha coefficients for the scale at each of four test times ranged from .89 - .93. After comparing outcomes using a second fatigue scale, the Lee Fatigue Scale (Lee, Hicks, & Nino-Murcia, 1991), Wambach concluded the two measures seemed consistent in producing similar correlations, and therefore only one measure would be needed in future investigations. Although no recommendation was made in terms of which of the two scales should be used, the finding offers further support for the validity of the measures in that they are intended to measure similar constructs. Edinburgh Postnatal Depression Scale (EPDS) The instrument selected to detect postnatal depressive symptoms was the EPDS (Cox, Holden, & Sagovsky, 1987) (Appendix VIII). The 10-item scale can be completed within five minutes and the authors suggest the tool should be completed when other family members are not present. Response categories are scored from 0-3 according to increased severity of the symptom. A score of nine is suggested for use as a threshold criterion to determine depression by primary care workers. The EPDS has been used in numerous studies worldwide and has been translated into several languages (Clifford, Day, Cox, & Werrett, 1999). Cox et al. (1987) regard the scale as one component of a broader assessment of a woman's emotional state. They also view the scale as appropriate for use among women in community settings. A validation study was conducted with mothers (n = 84) by comparing  62  responses with two concurrent measures, the Goldberg Standardized Psychiatric Interview (Goldberg, Cooper, Eastwood, Kedward, & Shepherd, 1970), and the Research Diagnostic Criteria for depressive illness (Spitzer, Endicott, & Robins, 1975). Findings suggested the tool was a promising means of identifying women with depressive symptoms. Split-half reliability testing of the EPDS was .88 and the alpha coefficient was .87. The authors reported the scale demonstrated satisfactory sensitivity (86%), specificity (78%) and a positive predictive value of 73%. Murray and Carothers (1990) replicated the testing approach using a larger sample (n=702) to determine if the scale was a reliable screening tool. Outcomes of their study revealed further support for use of the scale as a principal screening instrument. Murray and Cox (1990) tested the scale for use during pregnancy. The EPDS scores oflOO women were compared using the same two concurrent measures as used in the 1987 study. Sensitivity was 100%, false positive rate 4% and positive predictive value 60%. Personal Information - Data Collection Forms  Tools to record demographic variables and delivery characteristics that might impact breastfeeding duration were developed by the researcher based on the literature review and theoretical framework (Appendices III and IX-XII). The following demographic variables were included: age, education, whether the participant had been breastfed, how committed they felt to breastfeeding, the relationship of the person they perceived as having the most influence on their breastfeeding decisions, smoking status, how much physical help, and how much emotional support they received. Delivery characteristics included delivery mode,, length of labor, use of a Labor-Delivery-Recovery-Postpartum (LDRP) unit, length of hospital stay, analgesia received  63  during labor, and whether the infant received supplemental feedings while in hospital.  Qualitative Data Collection All participants were asked two open ended questions. The first question inquired about what antenatal concerns pertaining to breastfeeding they were experiencing (Appendix III). The second question was asked during the telephone interviews at 2, 12, and 24 weeks after birth (Appendices I-K), in the event that a participant informed the research assistant that she had weaned her infant. Participants who had weaned their infants were asked how they knew it was time for them to stop breastfeeding. Ethical Considerations The study received ethical approval from the University of British Columbia, Behavioral Sciences Screening Committee, and the University of Manitoba Ethical Review Committee for studies pertaining to the use of human subjects. The study also received agency approval for access to participants from the Winnipeg Regional Health Authority. All potential participants who had registered for a series of antenatal classes, offered in the community as part of the usual care, received a letter of information about the study (Appendix XIV). During the initial class of the series, the class instructor advised the registrants that a nurse researcher would be attending the fourth class to provide further explanation about the study outlined in the letter and invite them to participate. Anyone not interested in hearing about the study would be welcome to leave as the nurse researcher had been asked to provide the information about the study opportunity at the end of the class. The nurse researcher advised all potential participants that their participation was  64  voluntary, and that they could withdrawfromthe study at any time simply by notifying either the investigator or research assistant. All women were assured their care would not be jeopardized if they declined to participate or later withdrewfromthe study. Participants were given an opportunity to ask questions before consenting to join the study. Informed consent was obtained from each participant (Appendix XV). Participants also were assured that any identifying information would be kept confidential. Lists identifying actual participants with the numerical codes for identification were kept in a locked cabinet. Instruments were numerically coded and no names were noted except for initial coding purposes and follow up phoning. Information needed by the research assistant to phone the participants was stored in a locked cabinet. Research assistants and the statistician were required to sign a statement of confidentiality (Appendix XVI). Only coded data was used during data entry and analysis procedures. At the conclusion of the study, the principal investigator will shred all personal identification information. To comply with the conditions granted by the ethical approval process, coded raw data will be stored in a locked data storage cabinet within the MNRI for 10 years, and then shredded. Research assistants hired to contact participants were trained by the investigator to increase the consistency of their approach to collecting data and to limit the telephone interview to include only the required information. Participants were referred to their respective community health nurses when they expressed questions or concerns unrelated to the study. Telephone contact with the subjects was scripted to minimize any negative verbal reactions to actual breastfeeding outcomes. For example, participants were asked, "How did you know it  65  was time to stop?" rather than "Why did you quit?" This distinction was made so that women who participated in the study were less likely to feel they should have breastfed longer. Subjects were considered to be at limited risk from study participation. The principal investigator agreed to contact any participant who indicated she was experiencing thoughts of harming herself, i.e. item 10 of the Edinburgh Postnatal Depression Scale, to encourage the participant to seek professional assistance. Women in the intervention group may have benefited from receiving additional breastfeeding information as provided in the pre-workshop guide or group session. A summary of the study outcomes will be mailed to all participants who had indicated they would like to receive the information. Strengths and Limitations of the Study The study was considered to have strong internal validity because the design was a prospective two group randomized controlled trial (RCT) (Chalmers, 2000; Jadad, 1998). According to Armitage and Colton (1998), a RCT should be viewed as the essential means for testing the efficacy of therapeutic innovations. Selection bias was reduced by the study protocol in that participants were allocated to either the intervention or control group through use of sealed, opaque envelopes. Moreover, the person selecting the envelope was not involved in recruitment or conduct of the study. This trial constitutes essential research aimed at testing an antenatal nursing intervention, which may assist women to breastfeed longer. It builds on previous work that suggested antenatal interventions contribute to longer duration rates. No systematic differences in the care provided to the participants in the control group were identified thereby limiting performance bias. The sample size was sufficient to detect the effect of the  66  intervention on the dependent variable. The projected effect size was based on prior research (Redman et al., 1995; Serwint et al., 1996; Sheehan 1999), and the researcher's clinical experience. The intervention was derived from empirical evidence about the reasons women report for weaning, as well as strategies considered helpful to continue with breastfeeding over time. Research assistants, who collected data at 2, 12, and 24 weeks after birth by telephone interview, were blinded as to which group the participant had been assigned, thus limiting detection bias. Potential bias due to maternal recall has been a criticism of breastfeeding research. However, Launer et al. (1992) conducted a retrospective infant feeding study, as part of a larger prospective study, regarding child and adult health patterns. They concluded "retrospective infant feeding data based on maternal recall of events up to eighteen months in the past could be used with confidence in epidemiological studies" (Launer, et al., p. 203). The research assistant collected feeding data until six months after the infant's birth by asking the participant to provide infant feeding information based on the previous week. Breastfeeding research also has been criticized for not using consistent definitions when reporting the quantity, quality, and duration of the experience (Labbok & Krasovec, 1990). In response, the current study used specific criteria, as suggested in the literature, to interpret quantity of breast milk received by the infant, and to determine if other food substances were being offered (Ellis, et al., 1991). Therefore, in addition to the feeding categories, data was collected to reflect what was fed to the infant, how often, and by whom (Armstrong, 1991). Data was analyzed using an intention-to-treat approach, a preferred and conservative  67 means of analysis because it provides a more valid assessment of treatment efficacy as it relates to actual clinical practice (Armitage & Colton, 1998; Moher, Schultz, & Airman, 2001; Pocock, 1983; Pocock, Clayton, & Airman, 2002). Hollis and Campbell (1999) declare an intention-totreat analysis is essential for a pragmatic trial. A limitation of this study is that baseline demographic data did not include measures of maternal work and smoking status, and body mass index. Return to employment has been reported by women as a reason to wean their infants (Health Canada, 2000). This trial was conducted after a one year parental leave had been approved by the Canadian government. Concerns about return to work were therefore considered to be potentially less problematic among this group of breastfeeding women especially early in the postnatal period. However, since eligibility for a parental employment leave is subject to individual circumstances, not all women in this study may have qualified. Women had the option of reporting work-related concerns on the MAF or during the telephone interviews when asked how they knew it was time for them to stop breastfeeding. Collection of baseline data to ascertain smoking status should have been completed to facilitate comparisons regarding the proportion of women who resumed smoking after birth. Participants also should have been asked to indicate their height and pregravid weight in order to further test how predictive obesity may be in terms of predicting breastfeeding duration in view of Donath and Amir's (2000) study which detected a relationship between obesity and cessation of breastfeeding. Further limitations of the study include the sampling frame and a possible Hawthorne effect. The sampling frame targeted women who were most likely to continue to breastfeed (i.e.  68  women who were educated, were at least 18 years of age, had already decided to breastfeed, and had opted to attend a series of antenatal educational classes). Therefore, the results of this trial cannot be generalized beyond the characteristics of the convenience sample. Demonstrating that a significantly higher proportion of women would continue to breastfeed because of the intervention may have been more difficult because of the homogeneity of the groups which minimized variability for a number of variables. Finally, women may have been more likely to report they breastfed longer because they were aware that breastfeeding is a desirable behavior by health care professionals in general, and of particular interest to the researcher. The potential bias toward breastfeeding could have occurred, however, with any of the participants regardless of the group to which they had been assigned. The possibility exists that the actual breastfeeding outcomes were different because the participants knew they were being studied. This potential effect constitutes a threat to the external validity of the study, a common criticism of a randomized trial. Plan for Data Analysis Two research assistants and the investigator were involved in data entry and validation procedures. Questions about coding or inclusion of variables were discussed immediately with the investigator; changes in coding were logged arid documented. Entries were compared with previous data collection points to assess for discrepancies. For example, the date of weaning obtained during the two week follow up telephone interview, was compared with the data obtained during the 12 and 24 week telephone interviews. All data forms were checked for consistency of identifiers and completeness of information. Missing values were examined to  69  determine if an error had been made. Each response was checked to determine if the values were consistent with the range of response options. The database was backed up on to a zip drive and a hard drive owned by the Manitoba Nursing Research Institute (MNRI). The complete database also was transferred onto a working spreadsheet to ensure the integrity of the original database. Instrument items were reverse-coded and summed as directed by the scale developers. Recoded variables were created to maintain the integrity of the original data set. Using the Statistical Program for Social Scientists (Version 11), data was analyzed on the basis of intention-to-treat, meaning that the data provided from all women who participated in the study was analyzed according to their group allocation. The level of statistical significance was set at p < .05 (two-tailed) for all analyses. Analysis procedures proposed for each of the research questions were as follows: Primary Question 1. Do women assigned to the intervention group at least mainly breastfeed* longer than women assigned to the control group? (*mainly breastfeed defined as ingestion of breast milk plus up to one alternate feeding, such as commercially prepared formula, per day). Analytic Approach to the Primary Question The proportion of women who continued to at least mainly breastfeed was first compared using chi-square analyses. Risk estimates were calculated to determine if the intervention affected at least mainly breastfeeding by 2, 12, and 24 weeks after birth. A McNemar's Test was conducted for mainly breastfeeding between every pair of adjacent time points to check for changes over time, and compare for homogeneity between groups. Duration of any  70  breastfeeding was then calculated as a continuous variable, i.e. the number of weeks since birth. Kaplan-Meier survival analyses were used to assess the effects of individual variables (Parmar & Machin, 1995). The log rank test was used to test for equality. The Cox proportional hazards model was applied to investigate the simultaneous effect of multiple explanatory variables of the overall risk of cessation of breastfeeding. Secondary Questions  1. Do women assigned to an antenatal intervention report more or less maternal competence than women in the control group as measured by the PSOC? 2. Do women assigned to an antenatal intervention report more or less breastfeeding confidence than women in the control group as measured by the BFSES? 3. Do women assigned to an antenatal intervention report more or less satisfaction about their breastfeeding experience than women in the control group as measured by the MBFES? 4. Do women assigned to an antenatal intervention report more or less fatigue than women in the control group as measured by the MAF? 5. Do women assigned to an antenatal intervention report more or less postpartum depressive symptoms than women in the control group as measured by the EPDS? 6. What are the predisposing factors associated with breastfeeding duration? Analytic Approach to the Secondary Questions  For questions one to five, t tests were first completed to compare the mean scores of the major scales and subscales and identify variables that were significantly related to still at least mainly breastfeeding by 2, 12, and 24 weeks after birth. Repeated measures ANOVA testing  71  was then undertaken and a Hotelling's trace test statistic was calculated for each question to compare the above qualities between the two groups over time (Bray & Maxwell, 1985). Variables that were significantly different were then further examined using multivariate analyses to determine if these variables continued to be significantly related to breastfeeding duration. For question six, given the degree of homogeneity between the groups, data collected from all women who had participated in the study was examined together. Nominal and ordinal data was analyzed by nonparametric procedures, and interval data by t tests, to determine which variables suggested a significant association with the primary research question. Cross tabulations were used to look for associations with the categorical outcome of at least "mainly" or "not mainly" breastfeeding (i.e. feeding data indicated the woman was either partially breastfeeding or had weaned the infant). Survival analyses also were conducted to examine nominal and ordinal variables to determine if they were significantly associated with at least mainly breastfeeding. Cox's proportional hazard modeling and multivariate logistic regression procedures were used to further analyze multiple variables that had demonstrated a statistically significant relationship to the primary dependent research variable.  72  CHAPTER THREE FINDINGS This chapter presents the findings of the study. The chapter begins by describing and comparing the sample characteristics. The results associated with the primary and secondary research questions and qualitative outcomes are then reported. Sample Characteristics Participants' Demographic Information Data collected during the initial antenatal contact revealed participants ranged in age from 18-42 years (M = 28, SD = 4.13) (see Table 4). Ninety-two percent of the participants reported either their husband or partner attended the antenatal classes with them. The majority of women indicated they had completed postsecondary education, and about one half reported they had been breastfed. Most felt "very committed" to breastfeeding and stated the person who most influenced their breastfeeding decisions (i.e. their husband or partner) also was "very committed" to breastfeeding (see Table 5). Participants identified their mother, friend, sister, or other person also as being important in their decisions about breastfeeding. Most women reported they received at least minimal physical help in the home, and that they had emotional support available to themfromfamily and friends. Comparison of Baseline Variables by Group Chi-square and t test analyses did not reveal significant differences between the two groups as illustrated in Tables 4 and 5.  73  Table 4. Comparison of Groups by Baseline Demographic Variables Control n=98  Intervention n= 111  Variables  t  (%)  18-22  6.1  9.9  23-27  32.7  35.1  28-32  42.9  41.4  33-37  14.3  11.7  38-42  4.1  1.8  Education <High School  4.1  2.7  High School  15.5  14.4  College  22.7  22.5  University  55.7  60.4  2.1  x  2  Sig.  (%)  Age  Other  Statistics  2.246  .691  2.834  .586  2.536  .281  —  Participant was BF Yes  42.7  51.9  No  39.6  37.0  Not sure  17.7  11.1  74  Table 5 . Comparison of Groups by Personal Factors Control n = 98  Intervention n= 111  Statistics  t  Variables (%)  x  2  Sig.  (%)  Participant's Commitment to BF  2.258  Neutral (or not sure)  3.1  2.8  Minimally committed  1.0  2.8  Moderately committed  3.1  4.6  Very committed  7.3  8.3  Greatly committed  19.8  13.9  Totally committed  65.6  67.6  .812  •  Relationship of BF Decision Support  2.211  Husband/Partner  26.2  24.7  Mother  27.4  27.8  Friend  8.3  9.3  Sister  8.3  5.2  Other  28.6  33.0  Partner's (or other's) Commitment  .579  Neutral  1.3  Minimally committed  1.3  2.3  Moderately committed  5.3  7.0  Very committed  6.6  7.0  Greatly committed  13.2  11.6  Totally committed  71.1  72.1  .819  .837  75  Table 5. Comparison o f Groups by Personal Factors (Continued) Control n = 98  Intervention n= 111  Variables (%)  Statistics  t  1.1  Moderately supportive  1.1  2.9  Very supportive  5.4  6.7  Greatly supportive  5.4  4.8  Totally supportive  87.1  84.8  Who Provides Physical Help 95.2  94.6  Friend  2.4  2.2  Mother  2.2  1.2  Other  1.1  1.1  How Much Physical Help  .867  No help  1.1  1.9  Very little help  2.1  3.8  Slightly helpful Low level help  5.7 2.1  2.8  16.8  7.5  Minimally helpful  8.4  7.5  Moderately helpful  17.9  10.4  Neutral  .699  .701  .251  .969  1.0  Minimally supportive  Husband/Partner  Sig.  2  (%)  Support of Partner (or Other) Neutral  X  .120  76  Table 5. Comparison of Groups by Personal Factors (Continued) Control n = 98  Intervention n=lll  Variables  t  (%)  (%)  20.0  23.6  Greatly helpful  6.3  12.3  Totally helpful  25.3  24.5  Very helpful  Statistics  x  2  Sig.  However, t testing of baseline interval variables revealed a significant difference between the two groups with respect to the antenatal depression scores in that women in the control group reported slightly higher scores indicating they felt more depressed (p=.039). The between group differences did not continue to be significant at later test times. To control for this difference among baseline data between the two groups, the antenatal depression score was subsequently entered into multivariate logistic and Cox's proportional hazard regression models. The relationship was non-significant (p = .643) Description of Variables for the Total Sample Following Delivery Sixty-three percent of the participants gave birth vaginally and 58% delivered in LaborDelivery-Recovery-Postpartum (LDRP) units. Eighty-seven percent of all women reported use of analgesia, and of those, 34% received an epidural anesthetic and 36% received a combination of pharmacologic measures. The average length of labor was estimated to be less than 12 hours by 43% of women, between 13 and 24 hours by 34% of women, and more than 24 hours by 14%  77  of women. Nine percent of women were scheduled for elective cesarean sections. Thirty percent of the women reported they were discharged prior to 48 hours after birth. Only 3 % of the women recruited to the study experienced postpartum or newborn complications; one woman reported she experienced a stillbirth at seven months gestation. No other infant or maternal admission to hospital post birth was reported. Forty-two percent of the infants were supplemented with additional fluids while in hospital; nurses were the most common health care provider to offer the infant a supplemental feeding: Formula was used to supplement breastfeeding 3 5 % of the time and water 4 % of the time (3% of mothers were not sure what had been given to the baby but reported a supplemental feeding had been given). Twenty-nine percent of the women offered their infant expressed breast milk in a bottle by two weeks after birth. Seventy-six percent of the mothers had introduced solid foods (less than 2 teaspoons per day) to their infants by three months after birth. Group Comparisons of Variables Following Delivery By two weeks after birth, there were two significant differences between the groups; how often breastfeeding help had been requested, and from whom. Compared with the women in the control group, women in the intervention group reported they had sought help significantly more often (x = 5.339, p = .021), and that the help they requested was a visit from a Community 2  Health Nurse (CHN) (% = 4.170, p = .041). No significant differences were detected between the 2  two groups among any of the other variables compared at 2 or 12 weeks after birth. Significant between group differences were detected by 24 weeks after birth in terms of  78  who had helped the women continue to breastfeed. Women in the control group more frequently obtained help from the L a Leche League (x = 4.757, p = .029), the Breastfeeding Hotline (x = 2  2  5.780, p = .016), a friend ( = 4.004, p = .045), or their mother ( = 5.007, p = .025) when 2  x  2  x  compared to help sought by women in the intervention group. Primary Research Question - Duration of Breastfeeding 1. Did women assigned to the intervention group continue to at least mainly* breastfeed longer than women assigned to the control group? (*mainly breastfeed defined as ingestion of breast milk plus up to one alternate feeding, such as commercially prepared formula, per day). To compare feeding data between the groups, a chi-square analysis was completed. The number of women continuing to breastfeed within each category and group are noted in Table 6. At 24 weeks after birth, 57% (55/96) of the women assigned to the intervention group and 50% (41/82) of the women assigned to the control group, were still at least mainly breastfeeding. At 2, 12, and 24 weeks, the chi-square statistic was 1.311 (p = .86), 1.460 (p = .83), and 1.411 (p = .84). Therefore there were no significant differences between the groups regarding breastfeeding duration. At the time of randomization, 111 women were assigned to the intervention group and 98 women to the control group. Subsequent comparisons between the two groups were completed at each time point to determine how many were still at least mainly breastfeeding using a denominator that represented only the number of women continuing to participate in the trial. Hollis and Campbell (1999) report no consensus exists regarding how missing responses should be handled in intention-to-treat analyses. Furthermore, Dennis, Hodnett, Gallop, and Chalmers  79  (2002) and Redman et al. (1995) referred to an adjusted denominator in their breastfeeding trials. Therefore, comparisons were completed by reporting women who had weaned separately from those who had withdrawn from participation.  Table 6. Number of Women in each Group by Feeding Category or Study Withdrawal  Time  I - 2 wks  C - 2 wks  I - 12 wks  C - 12 wks  I - 24 wks  C - 24 wks  Revised n due to withdrawals  n == 107  n = 89  n = 101  n = 84  n=96  n = 82  4 opt outs  9 opt outs  10 opt outs  14 opt outs  15 opt outs  16 opt outs  Exclusively  81 (76%)  66 (74%)  61 (60%)  47 (56%)  29 (30%)  22 (27%)  Primarily  3  (2%)  2  (2%)  4  2  6  6  Mainly  5  (5%)  7  (8%)  13 (13%)  10 (12%)  20 (21%)  13 (16%)  Partially  11 (10%)  7  (8%)  4  5  13 (14%)  12 (15%)  Weaned  7  7  (8%)  19 (19%)  28 (29%)  29 (35%)  Category:  (7%)  ( 4 % )  (4%)  (2%)  (6%)  20 (24%)  (6%)  (7%)  ^Intervention group, C=Control group  More specifically, the comparison between the groups revealed that by: 1. 2 weeks after birth, the same proportion of the women continued to at least mainly breastfeed in each group, 89/107 (83%) of the women in the intervention group, and 75/89 (84%) of the women in the control group [0.923, (0.430,1.980) (NS)]. Seven women had opted to wean their infants in each group. More of the women in the control group 9/98 (9%) opted out of the study  80  compared to the number of women 4/111 (3%) in the intervention group. 2. 12 weeks afterbirth. 78/101 (77.2%) of the women in the intervention group were still at least mainly breastfeeding versus 59/84 (70.2%) of women in the control group [1.437, (0.743, 2.778) (NS)]. The proportion of women who had weaned their infants was higher in the control group 20/84 (23.8%) than in the intervention group 19/101 (18%). A higher proportion of women in the control group 14/98 (14.2%) had withdrawn from the study versus the women in the intervention group 10/111 (9%). 3. 24 weeks after birth, 55/96 (57%) of women in the intervention group were still at least mainly breastfeeding versus 41/82 (50%) of the women in the control group [1.342, (0.741, 2.427) (NS)]. The proportion of women who weaned was higher in the control group 29/82 (35%) compared with that in the intervention group 28/96 (29%). By 24 weeks after birth, a similar proportion of women had withdrawn from the study in that 15/111 (13.5%) of the women in the intervention group had withdrawn versus 16/98 (16.3%) in the control group. Overall, there were 196/209 (94%) women continuing to participate in the study by 2 weeks after birth, 185/209 (89%) by 12 weeks after birth, and 178/209 (85%) by 24 weeks after birth. Attrition pertains to the number of participants who are either lost to follow up or who withdrew from participation in the study. The total rate of attrition for the study was therefore 15% by 24 weeks after birth. Comparable numbers of participants either withdrew from the study or were lost to follow up in both groups, thereby, limiting attrition or exclusion bias. Repeating the cross-tab calculations to compare the groups in terms of "exclusive" breastfeeding revealed the following odds ratios and 95% confidence intervals at 2, 12, and 24  81  weeks afterbirth [1.086 (0.568,2.075)(NS), 1.200 (0.668, 2.160) (NS), and 1.181 (0.613, 2.273) (NS)], respectively. A second sub-group comparison was then conducted to determine the proportion of women who continued with "any" (or at least "partially") breastfeeding versus those who had weaned. Using the total number of women originally randomized to each group as the denominator to compare the two groups since either women were still breastfeeding or not, 68/111 (61.2%) of women in the intervention group and 53/98 (54%) of women in the control group were still continuing to breastfeed. Odds ratios and 95% confidence intervals at the three test points revealed that these differences were not significant [1.773 (0.780, 4.032) (NS), 1.501 (0.830, 2.717) (NS), and 1.342 (0.774, 2.331) (NS)]. The proportion of women continuing to breastfeed was further compared between the two groups using survival analyses (Figure 4). Kaplan-Meier curves were plotted to describe the weaning patterns for each group using the continuous variable, number of weeks still breastfeeding on the horizontal axis, and the proportion of women weaning (cumulative survival) in each group on the vertical axis. No significant difference was obtained using survival analyses between the two groups with respect to breastfeeding duration. Cases in the survival analyses were censored when a woman either withdrew from the study (n = 31), never did initiate breastfeeding (n = 5), or if she was still breastfeeding beyond 24 weeks after birth (n=l 19). The mean duration of any breastfeeding among women in the intervention group was 20.09 weeks compared to 19.49 weeks for women in the control group.  82  Figure 4. Proportion of Women Continuing to Mainly Breastfeed by Group  Randomized Group ° control ° intervention  Breastfeeding Duration in Weeks The log rank (.5003) and Breslow (.4591) statistics comparing the two groups in terms of breastfeeding duration were non-significant, as well as a comparison of point estimates for each group (see Table 7) (Armitage & Coltman, 1998; Pocock, Clayton, & Airman, 2002). Table 7. Proportion of Women Mainly Breastfeeding at 2, 12, and 24 Weeks Group/Time  Intervention Group (95% CI)  Control Group (95% CI)  2 weeks after birth  .9440 (.9005 - .9875)  .9021 (.8411 - .9630)  12 weeks after birth  .7883 (.7097 - .8669)  .7680 (.6782 - .8578)  24 weeks after birth  .6771 (.5862 - .7680)  .6339 (.5306 - .7372)  Point estimates (proportions) and 9 5 % confidence intervals obtained from survival analysis  83  Sub-group analyses were completed using survival analysis procedures to compare the groups in terms of the proportion of women who were "exclusively" breastfeeding. No significant differences were obtained. However, a McNemar's Test, conducted between every pair of adjacent time points to check for changes over time, and a Cochran's Q statistic revealed a significant difference in the pattern of changes between the two groups. Between 2 weeks and 12 weeks, women assigned to the intervention group did not systematically change in their breastfeeding status (p = .146). However, more women assigned to the control group did change in that women who were at least mainly breastfeeding at 2 weeks after birth were no longer at least mainly breastfeeding by 12 weeks after birth (p< .001). The finding suggested variables that were significantly associated with mainly breastfeeding at 2 weeks after birth should be subsequently entered into regression modeling procedures to examine what variables might be predictive of continuing to mainly breastfeed by 24 weeks after birth. Secondary Research Questions 1. Did women assigned to the antenatal intervention report more or less maternal competence than women in the control group as measured by the Parenting Sense of Competence (PSOC) Scale? 2. Did women assigned to the antenatal intervention report more or less breastfeeding confidence than women in the control group as measured by the Breastfeeding Self-Efficacy Scale (BSES)? 3. Did women assigned to the antenatal intervention report more or less satisfaction about their breastfeeding experience than women in the control group as measured by the Maternal Breastfeeding Evaluation Scale (MBFES)?  84  4. Did women assigned to the antenatal intervention report more or less fatigue than women in the control group as measured by the Multidimensional Assessment of Fatigue (MAF) Scale? 5. Did women assigned to the antenatal intervention report more or less postpartum depressive symptoms than women in the control group as measured by the Edinburgh Postnatal Depression Scale (EPDS)? To answer the secondary research questions 2 to 5, repeated measures ANOVA statistical procedures and a Hotelling's test statistic were then used to compare the two groups over time (see Table 8). There were no significant differences between the mean scores of the women in the intervention and control groups when testing maternal competence or breastfeeding confidence. A significant difference was detected when comparing the three mean subscale scores of the Maternal Breastfeeding Evaluation Scale (MBFES). No significant differences existed between the groups with respect to maternal enjoyment/role attainment or infant satisfaction/growth. However, the subscale scores pertaining to aspects of maternal lifestyle resumption/body image was significantly different between groups (p = .049). Women assigned to the control group reported slightly higher scores at baseline whereas women in the intervention group reported higher lifestyle subscale scores by 24 weeks after birth. The total mean scores of the MBFES were not significantly different and no significant differences were observed after comparing the mean fatigue or depression scores or subscale scores.  85  Table 8. Repeated Measures ANOVA Results of Interval Variables  Variable  Group  Mean at each test time Tl T2 T3 T4  Maternal Competence Intervention  74  76  84  87  Control  73  76  84  86  Knowledge Subscale  H .902  .935 Intervention  32  34  39  40  Control  31  33  39  40  Values Subscale  .786 Intervention  43  43  46  47  Control  43  43  46  46  Breastfeeding Confidence  .286  Intervention  120  134  147  147  Control  125  138  147  145  T l = antenatal data collection, T2 = 2 weeks after birth, T3 = 12 weeks after birth, T4 = 24 weeks after birth, H = Hotelling's Trace test statistic  86  Table 8. Repeated Measures ANOVA Results of Interval Variables (Continued)  Variable  Group  Mean at each test time Tl  T2  T3  T4  Breastfeeding Satisfaction  H .530  Intervention  107  119  127  130  Control  112  124  132  132  Maternal Enjoyment/Role Attainment Subscale  .844  Intervention  44  46  48  62  Control  46  47  50  64  Infant Satisfaction/Growth Subscale  .806  Intervention  28  33  36  35  Control  29  35  37  36  Maternal Lifestyle/Body Image Subscale  .049*  Intervention  27  29  31  33  Control  29  29  31  31  Multidimensional Fatigue Assessment  .689  Intervention  3.4  4.6  2.4  2.4  Control  3.5  4.8  2.8  2.5  T l = antenatal data collection, T2 = 2 weeks after birth, T3 = 12 weeks after birth, T4 = 24 weeks after birth, H = Hotelling's Trace test statistic  87  Table 8. Repeated Measures ANOVA Results of Interval Variables (Continued) Variable  Group  Mean at each test time Tl T2 T3 T4  Severity Subscale  H  .316 Intervention  12  13  Control  11  14  8 10  9 .687  Distress Subscale Intervention  3.9  . 4.9  3.0  2.9  Control  3.7  5.0  3.3  13.2  Interference with ADL Subscale  .706  Intervention  36  47  25  25  Control  38  50  30  26 .661  Frequency of Fatigue Subscale Intervention  5.3  5.0  4.1  4.2  Control  5.2  5.3  4.3  4.5 .395  Depression Intervention  1.7  7.8  5.3  4.1  Control  2.0  7.9  5.5  5.4  T1 = antenatal data collection, T2 = 2 weeks after birth, T3 = 12 weeks after birth, T4 = 24 weeks after birth, H • Hotelling's Trace test statistic  88  Psychometric Properties ofMajor Study Instruments Five instruments were selected for use to measure the variables identified in the secondary research questions (2 to 5). Cronbach's alpha correlation coefficients for each of the scales and subscales are summarized in Table 9. According to Nunnally (1978), a minimum criterion of .70 is necessary to demonstrate adequate internal consistency reliability. Two of the subscales, when tested in the current sample, did not meet this criterion; i.e., the maternal lifestyle/body image subscale (.67) of the Maternal Breastfeeding Evaluation Scale (MBFES) and the frequency subscale (.58) of the Multidimensional Assessment of Fatigue (MAF). Table 9. Internal Consistency Values of Major Study Instruments Measure  Cronbach's Alpha  Total Parental Sense of Competence Scale Knowledge Subscale Values Subscale  .89 .85 .81  Breastfeeding Self-Efficacy Scale Technique Subscale Maternal Attitudes and Beliefs Subscale  .94 .92 .90  Maternal Breastfeeding Evaluation Scale Maternal Enjoyment/Role Attainment Subscale Infant Satisfaction/Growth Subscale Lifestyle/Body Image Subscale  .70 .89 .75 .67  Multidimensional Fatigue Assessment Severity Subscale Distress Subscale Degree of Interference in ADL Subscale Frequency  .85 .93  Edinburgh Postnatal Depression Scale  .84  .84 .58  Distress subscale alpha coefficient not available as scale includes only one item to measure this dimension.  89  It is unclear why the reliability coefficient for items of the maternal lifestyle/body image subscale of the MBFES did not meet the minimum of .70 during baseline data collection. However, further reliability testing did reveal sufficient correlations of the subscale at 2, 12, and 24 weeks after birth in that the coefficients were .71, .79, and .82, respectively. In contrast, the frequency subscale of the MAF, which has not been tested with similar cohorts and consists of only two items to measure this dimension, did not result in any further support regarding stability over time in that alpha coefficients in subsequent testing at 2, 12, and 24 weeks after birth were .57, .51, and .41, respectively. It is possible the response options for these items require modification and further testing prior to future use with expectant and new first-time mothers. Principal components factor extraction with varimax rotation procedures were then completed on the multidimensional study instruments to determine how the items loaded. Results revealed the items of the Parenting Sense Of Competence (PSOC) Scale loaded in a two factor solution with items 1-8 loading together within the knowledge subscale and items 9-17 loading together within the values subscale as described by Gibaud-Wallston and Wandersman (1978) (see Table 10). Items of the Breastfeeding Self Efficacy Scale (BSES) loaded on two factors as shown in Table 11. Described by Dennis and Faux (1999) as pertaining to technique, the first factor includes 20 items. Item 29 did not load well on any factor. Items 4 and 10 loaded on both factors. Dennis and Faux's second factor, described as pertaining to intrapersonal thoughts contains 13 items. In this testing, all of the items loaded well. Items of the MBFES loaded on three factors (see Table 12). The first factor that includes 13 items is described by Leff et al. (1994) as pertaining to the theme of maternal enjoyment/role attainment. Item 25  90  loaded on this factor but seemed to relate more to infant satisfaction. The second factor that contains 7 items pertains to the Leff et al.'s theme of infant satisfaction. Items 15, 19, and 28 did not load on this factor although the items pertained most to that subscale. The third factor with 10 items, related to aspects of maternal lifestyle and body image. Item 26 did not load on this factor, as expected. Items of the MAF resulted in a four factor solution but not as expected in view of Belza'a (1990) findings, in that the items intended to measure severity and distress clustered together as one dimension (see Table 13). The items intended to pertain to interference with aspects of daily living split between three factors. The final two items of the MAF intended to measure frequency of fatigue comprised the fourth factor. Two-tailed Pearson correlation coefficients were calculated at each of the four times data was collected. Findings support the construct validity of the five major instruments selected for use. At baseline, the scores indicating maternal competence and confidence correlated positively (.489) (p = .01), as did competence and satisfaction (.494) (p = .01), and confidence and satisfaction (.578) (p =.01). In comparison, fatigue correlated negatively with confidence (.2.64) (p = .01) and satisfaction (-.208) (p= .01) but positively with depression (.345) (p =.01). Depression scores correlated negatively with competence, confidence, and satisfaction, but not significantly. At two weeks after birth, all correlations resembled the direction of the relationships observed at baseline with the exception of the coefficients calculated when testing depression. The two week depression scores were negatively associated with maternal competence (p =.01), confidence (p =.01), and satisfaction (p =.01). Similar findings were observed when the scores  91  were examined at both 12 and 24 weeks after birth in that the direction of the relationship between the variables was the same and all coefficients were significant (p= .01). Construct validity of the instruments selected for use in this investigation was thus supported.  92  Table 10. PSOC Items with Principal Components Varimax Factor Loadings Item  Loading Factor 1 Factor 2 I. The problems of taking care of the baby will be easy to solve once .39 .30 you know how your actions affect your baby, an understanding I have acquired. 2.1 will meet my own personal expectations for expertise in caring for my baby.  .54  .36  3.1 will make a fine model of a new mother to follow in order to learn what she would need to know in order to be a good parent.  .52  .39  4. Being a parent will be manageable, and any problems will be easily solved.  .69  .21  5. If anyone can find the answer to what is troubling my baby, I will be the one.  .57  .27  6. It will be easy to know when you are doing a good job or a bad job being a parent.  .65  .24  7. Considering I am expecting my first child, I feel thoroughly familiar with the role.  .70  .13  8.1 honestly believe I will have all the skills necessary to be a good mother to my baby.  .66  .27  9.1 know being a parent will be rewarding and am not feeling frustrated.  .28  .48  10.1 feel like I will be in control and not being manipulated.  .44  .53  II. 1 feel as prepared to be a good mother as my mother would have been when she had me.  .23  .37  12.1 feel like I will be able to complete what I need to do.  .36  .44  13. When I wake up in the morning, I will like I have accomplished a lot.  .39  .51  Table 10. PSOC Items with Principal Components Varimax Factor Loadings (Continued) Item  Loading Factor 1 Factor 2  14. My talents and interests will include being a parent.  .34  .42  15. Being a mother of an infant will be interesting and I am motivated to do a good job.  .16  .58  16.1 will be relaxed and calm as a parent.  .40  .57  17. Being a good mother will be a reward in itself.  —  .64  Note: error detected on last item under Factor 1  Table 11. BSES Items with Principal Components Varimax Factor Loadings Item 1. Ensure that my baby is properly latched on for the whole feeding.  Loading Factor 1 Factor 2 .82 .09  2. Recognize when my baby is finished breastfeeding.  .14  .62  3. Manage the breastfeeding situation to my satisfaction.  .47  .67  4. Determine that my baby is getting enough milk.  .49  .42  5. Recognize the signs of a good latch.  .01  .73  6. Feel if my baby is sucking properly at my breast.  .11  .78  7. Maintain my milk supply by using the 'supply and demand' rule.  .33  .45  8. Finish feeding my baby on one breast before switching to the other breast.  .03  .46  9. Position my baby correctly at my breast.  .18  .75  Table 11. BSES Items with Principal Components Varimax Factor Loadings (Continued) Item  Loading Factor 1 Factor 2  10. Manage to keep up with my baby's breastfeeding demands.  .52  .54  11. Take my baby off the breast without pain to myself.  .10  .57  12. Hold my baby comfortably during breastfeeding.  .12  .72  13. Manage to breastfeed even if my baby is crying.  .31  .58  14. Monitor how much breast milk my baby is getting by keeping track of my baby's urine output and bowel movements.  .26  .19  15. Keep my baby awake at my breast during a feeding.  .31  .53  16. Feed my baby every 2-3 hours.  .41  .18  17. Deal with the fact that breastfeeding can be time consuming.  .64  .12  18. Accept the fact that breastfeeding may temporarily limit my freedom.  .55  .05  19. Comfortably breastfeed with my family members present.  .13  .28  20. Comfortably breastfeed in public places.  .22  .31  21. Count on my friends to support my decision to breastfeed  .21  .04  22. Keep feeling that I really want to continue to breastfeed my baby  .73  .16  23. Keep wanting to breastfeed my baby.  .78  .08  Table 11. BSES Items with Principal Components Varimax Factor Loadings (Continued) Item  Loading Factor 1 Factor 2  25. Breastfeed my baby without using formula as a supplement.  .76  .08  26. Motivate myself to breastfeed successfully.  .87  .06  27. Feed my baby with breast milk only.  .80  .06  28. Continue to breastfeed my baby for every feeding.  .83  .06  29. Depend on my family to support my decision to breastfeed.  .18  .12  30. Refrain from bottle feeding for the first few weeks.  .59  .15  31. Successfully cope with breastfeeding like I have with other challenging tasks.  .75  .20  32. Be satisfied with breastfeeding experience.  .67  .42  33. Focus on getting through one feed at a time.  .64  .29  Table 12. MBFES Items with Principal Components Varimax Factor Loadings Item 1. With breastfeeding I will feel a sense of inner contentment.  Factor 1 .74  Loading Factor 2 .17  Factor 3 .18  2. Breastfeeding will be a special time with my baby.  .65  .03  .27  3. My baby won't be interested in breastfeeding.  -.12  .40  .35  4. My baby will love to nurse.  .20  .65  .24  5. It will be a burden being my baby's main source of food.  .52  -.11  .56  6.1 will feel extremely close to my baby when I breastfed.  .74  .09  .20  7. My baby will be an eager breastfeeder.  .15  .77  .14  8. Breastfeeding will be physically draining.  .02  .27  .49  9. It will be important to me to be able to nurse.  .47  .13  .17  10. While breastfeeding, my baby's growth will be excellent.  .31  .64  .03  11. My baby and I will work together to make breastfeeding go smoothly.  .46  .41  -.16  12. Breastfeeding will be a very nurturing, maternal experience.  .75  .16  .09  13. While breastfeeding, I will feel self-conscious about my body.  .15  -.07  .60  14. With breastfeeding, I will feel too tied down all the time.  .17  -.09  .77  Table 12. MBFES Items with Principal Components Varimax Factor Loadings (Continued) Item 15. While breastfeeding, I will worry about my baby gaining enough weight.  Factor 1 -.08  Loading Factor 2 .18  Factor 3 .51  16. Breastfeeding will be soothing when my baby is upset or crying.  .23  .61  -.04  17. Breastfeeding will be like a high of sorts.  .62  .26  -.04  18. The fact that I could produce the food to feed my own baby will be very satisfying.  .70  .10  .12  19. In the beginning, my baby will have trouble breastfeeding.  -.02  .30  .40  20. Breastfeeding will make me feel like a good mother.  .49  .30  .05  21.1 will really enjoy nursing.  .57  .36  .36  22. While breastfeeding, I will be anxious to have my body back.  .13  -.08  .50  23. Breastfeeding will make me feel more confident as a mother.  .45  .39  -.03  24. My baby will gain weight really well with breast milk.  .26  .70  .05  25. Breastfeeding will make my baby feel more secure.  .52  .49  -.09  26.1 will easily fit my baby's breastfeeding with my other activities.  .17  .44  .33  27. Breastfeeding will make me feel like a cow.  .33  -.07  .59  28. My baby will not relax while nursing.  .11  .30  .54  98  Table 12. MBFES Items with Principal Components Varimax Factor Loadings (Continued) Item  Loading Factor 1 Factor 2  Factor 3  29. Breastfeeding will be emotionally draining.  .18  .25  .68  30. Breastfeeding will feel wonderful to me.  .53  .48  .34  Table 13. MAF Items with Principal Components Varimax Factor Loadings Item Factor 1  Loading Factor 2 Factor 3 Factor 4  1. To what degree have you experienced fatigue?  .24  .80  .27  -.16  2. How severe is the fatigue which you have been experiencing?  .46  .74  .21  -.05  3. To what degree has fatigue caused you distress?  .21  .82  .29  -.13  4. Do household chores?  .58  .45  .29  -.12  5. Cook?  .64  .29  .43  -.09  6. Bathe or wash?  .24  .17  .82  -.08  7. Dress?  .31  .15  .86  .01  8. Work?  .81  .10  .26  -.03  9. Visit or socialize with friends or family?  .47  .32  .56  -.11  In the past week to what degree has fatigue interfered with your ability to:  99  Table 13. MAF Items with Principal Components Varimax Factor Loadings (Continued) Item 10. Engage in sexual activity?  Loading Factor 1 Factor 2 Factor 3 Factor 4 -.07 .76 .15 .09 .74  .32  .21  -.15  -.05  .45  -.14  .30  13. Walk?  .68  .05  .45  -.10  14. Exercise other than walking?  .85  .12  .17  -.11  15. Over the past week, how often have you been fatigued?  -.05  -.02  -.02  .88  16. To what degree has your fatigue changed during the past week?  -.16  -.20  -.06  .81  11. Engage in leisure and recreational activities? 12. Shop and do errands?  6. What are the predisposing factors associated with breastfeeding duration? To answer research question six, both univariate and multivariate analyses were completed using the datafromthe total sample. Variables entered into multivariate regression models either were statistically significant when tested using univariate procedures or were selected because they represented the key variables identified in the primary and secondary research questions. Predisposing Factors Associated with Breastfeeding Duration Cross tabulations were completed at 2, 12, and 24 weeks after birth (see Table 14). Mothers who did not smoke, or who had completed a higher level of education, were more likely to be at least mainly breastfeeding at each time. Additional factors positively associated with at  100  least mainly breastfeeding were: at 2 weeks after birth, infants who had not received supplemental feedings; at 12 weeks after birth, vaginal delivery and a visit from a community health nurse within the first two postpartum weeks; and at 24 weeks after birth, whether or not the infant's mother had been breastfed.  Table 14. Cross-tab Results of Categorical Variables and Mainly Breastfeeding. Time / Variable  Sig.  OR  Lower  Upper  Resumed Smoking  .038  0.266  0.070  1.003  Postsecondary  .001  4.057  1.734  9.490  Supplemental Feedings  .003  0.309  0.014  0.695  Vaginal Delivery  .041  2.137  1.022  4.464  Postsecondary  .005  3.237  1.393  7.522  Resumed Smoking  .038  0.259  0.066  1.007  CHN Visit  .032  1.858  0.952  3.626  Postsecondary  .011  2.962  1.257  6.982  Mother Breastfed  .006  2.329  1.266  4.287  Resumed Smoking  .008  .0934  .0119  .7961  2 weeks after birth:  12 weeks after birth:  24 weeks after birth:  Note: * Upper and lower confidence intervals include value of 1.0  101  Factors Associated with Increased Breastfeeding Factors associated with increased breastfeeding were further substantiated using survival analysis procedures. Variables that demonstrated a significant relationship included nonsmoking status, whether the mother had been breastfed, and whether the mother had been visited by a community health nurse (CHN) (see Figures 5 - 8). At two weeks after birth, 10/196 (5%) women reported they had resumed smoking. Log rank and Breslow statistics were significant in terms of the relationship between smoking resumption and breastfeeding duration where p = .0153 and p = .0186, respectively (see Figure 5). By 12 weeks after birth, 12/185 (6%) of women reported they had resumed smoking. Log rank and Breslow statistics were again significant where p = .0054 and p = .0067, respectively (Figure 6). Data by 24 weeks after birth revealed only 5/178 (2%) participants reported they were smoking cigarettes. However four of the women who had previously reported smoking had since withdrawn from study participation. Of the twelve women who had reported smoking at any of the three times they were contacted, only two had continued to breastfeed by 24 weeks after birth. Smoking status at 24 weeks after birth was not significantly associated with breastfeeding duration. The log rank (p = .0144) and Breslow (p = .0204) statistics testing the variable pertaining to whether the mother had been breastfed showed a significant relationship to breastfeeding duration (Figure 7). Of the 204 women who completed this question, 97 (41%) reported they had been breastfed, whereas 107 stated either they had not been breastfed or did not know. Figure 8 depicts the significant relationship between breastfeeding duration and whether the  102 mother had been visited by a community health nurse (CHN) within two weeks after birth. Fifty-eight percent of the participants reported they had been visited by a CHN. Log rank (p = .0318) and Breslow (p = .0188) statistics further support the significant effect of the CHN visit by two weeks after birth.  Figure 5. Proportion of Women Continuing to Mainly Breastfeed by Smoking Status at 2 weeks  Do you smoke?  Breastfeeding Duration in Weeks  103  Figure 6. Proportion ofWomen Continuing to Mainly Breastfeed by Smoking Status (12 weeks)  c u  £ jo  D o you s m o k e ?  o c o w c o o  o  no  yes  OH  X~  10  20  Breastfeeding Duration in Weeks  Figure 7. Proportion of Women Continuing to Mainly Breastfeed by Whether Mother Breastfed  Mother was Breastfed yes no 0"  10  20  Breastfeeding Duration in Weeks  104  Figure 8. Proportion of Women Continuing to Mainly Breastfeed by Whether Mother Received CHN Visit by 2 weeks  CHN a yes  Breastfeeding Duration in Weeks Additional Variables Associated with Breastfeeding Duration Antenatal commitment to breastfeeding was a significant predictor of continuing to at least mainly breastfeed at 2 weeks (p = .029), 12 weeks (p = .000), and 24 weeks (p = .002) after birth. Mann-Whitney U tests were repeated to assess "exclusively" breastfeeding at each test time. Results revealed a woman was more likely to still be "exclusively" breastfeeding by two weeks when her partner was the most influential person in her decision to breastfeed (p=.047), and when her partner provided the physical help (p=.045). T tests were then completed to compare the mean scores of the major scales and subscales. Women who were still at least mainly breastfeeding by 2, 12, and 24 weeks after birth reported significantly more breastfeeding confidence (p = .000), and more satisfaction,  105  particularly with respect to infant satisfaction (p = .000), and maternal enjoyment/role attainment (p =.046), when they responded to questionnaires at 2, 12, and 24 weeks after the birth of their infant. Women who reported more antenatal depression were significantly less likely to still be at least mainly breastfeeding by 2 weeks after birth (p =.041). Secondary variables that may help to further understand breastfeeding duration were then tested using multivariate analysis procedures. Categorical (group allocation, completion of postsecondary education, delivery mode, smoking status, if the baby had received supplemental feedings while in hospital, whether the mother had been breastfed, or received a CHN visit by two weeks after birth), ordinal (antenatal commitment to breastfeed), and interval variables [knowledge (subscale of the maternal competence scale), breastfeeding confidence, maternal satisfaction, fatigue, and depression scores] were selected for entry into the multivariate logistic backward, stepwise, conditional, regression procedures. Each of these variables were either statistically significant during univariate testing procedures or represented a central variable as outlined in the primary or secondary research questions. The skill/knowledge subscale scores were selected for entry, rather than the total competence scale, because the univariate analysis had revealed that by two weeks after birth, women with more knowledge were less likely to still be at least mainly breastfeeding. Since this finding did not seem clinically likely, further analysis was warranted. Also, the subscale scores of the competence scale pertaining to values/comforting were not selected for entry into the multivariate procedures as there were no differences between the two groups and because these scores are less likely to vary over time according to Gibaud-Wallston and Wandersman (1978). Outcomes completed at each of the  106  three postnatal test times are presented in Tables 15-20. The full models outline which variables were entered into the regression equation whereas the final model for each test time includes only those variables that were significant by the final (eighth) step. One degree of freedom has been used in each calculation.  Table 15. Logistic Regression to Model Mainly Breastfeeding at 2 weeks - Full Model Variable  Adjusted OR  Allocated to Intervention .466 Postsecondary  15.061  95% CI  .068 -  Sig.  3.183  .436  1.374- 165.115  .026  Vaginal Delivery  2.827  .178-  44.794  .461  Resumed Smoking  3.109  .087- 110.824  .534  Supplemental Feeds  2.601  .476 - 14.211  .270  Mother Breastfed  2.347  .307-  17.923  .411  .046 -  1.146  .073  2.173  .360-  13.103  .397  More Knowledge  .769  .604-  .981  .769  More Confidence  1.222  1.083 -  1.378  .001  More Satisfaction  .959  .885 -  1.039  .302  1.015  .921 -  1.118  .766  .386  .183-  .831  .012  More Commitment CHN Visited  More Fatigue More Depression  .230  107  Table 16. Logistic Regression to Model Mainly Breastfeeding at 2 weeks - Final Model Variable  Adjusted OR  95% CI  Sig.  More Confidence  1.155  More Knowledge  .824  .697-  .975  .024  More Depression  .506  .298 - .859  .012  8.242  1.178- 57.662  .034  Postsecondary  1.082- 1.232  .000  The logistic regression further confirmed that by two weeks after birth, women who continued to at least mainly breastfeed had more education, and reported more breastfeeding confidence. Women who had reported more knowledge on the competence subscale were slightly less likely to still be at least mainly breastfeeding by 24 weeks after birth, as were women who reported more depression. By 12 weeks after birth, women who continued to at least mainly breastfeed had reported more confidence by two weeks after birth, and been assigned to the intervention group. Women who had resumed smoking were less likely to still be at least mainly breastfeeding. By 24 weeks after birth, women assigned to the intervention group, and those who reported more satisfaction with breastfeeding within two weeks of the birth, were more likely to still be at least mainly breastfeeding. If the woman was smoking, it was more likely she had weaned by 24 weeks after birth.  Table 17. Logistic Regression to Model Mainly Breastfeeding at 12 weeks - Full Model Variable  Adjusted OR  Allocated to Intervention4.009  95% C I  Sig.  1.223 - 13.147  .022  More Commitment  1.325  .852- 2.060  .212  Vaginal Delivery  2.798  .746- 10.499  .127  Postsecondary  1.881  .431 - 8.215  .401  Resumed Smoking  .137  .019-  .971  .047  Supplemental Feeds  .575  .171 - 1.932  .371  CHN Visited  .391  .130- 1.176  .095  Mother Breastfed  .846  .274- 2.613  .772  More Knowledge  .906  .825 - .996  .041  More Confidence  1.035  .987- 1.085  .155  More Satisfaction  1.065  1.005 - 1.129  .032  .962  .901 - 1.027  .244  1.030  .695 - 1.527  .882  More Fatigue More Depression  Table 18. Logistic Regression to Model Mainly Breastfeeding at 12 weeks - Final Model Variable  Adjusted OR  95% CI  Sig.  Allocated to Intervention  3.885  1.355 - 11.136  .012  More Confidence  1.046  1.003- 1.092  .037  Resumed Smoking  .152  .025 - .913  .039  Table 19. Logistic Regression to Model Mainly Breastfeeding at 24 weeks - Full Model Variable  Adjusted OR  Allocated to Intervention  95%, C I  Sig.  2.279  .955 -5.439  .064  .978  .645 - 1.485  .918  2.426  .699- 8.421  .163  .103  .011 - 1.001  .050  Vaginal Delivery  1.819  .608 -5.436  .284  Supplemental Feeds  .354  .140- .892  .028  Mother Breastfed  1.193  .522- 2.724  .676  CHN Visited  .510  .220- 1.180  .116  More Knowledge  .900  .833 -.972  .007  More Confidence  1.013  .876- 1.052  .495  More Satisfaction  1.056  1.010 • 1.104  .017  More Fatigue  .978  .931 - 1.028  .382  More Depression  .936  .707- 1.238  .642  More Commitment Postsecondary Resumed Smoking  110  Table 20. Logistic Regression to Model Mainly Breastfeeding at 24 weeks - Final Model Variable  Adjusted OR  Allocated to Intervention 2.583  95% CI  Sig.  1.127-  5.921  .025  More Knowledge  .912  .849-  .979  .011  More Satisfaction  1.070  1.038 -  1.104  .000  .068  .007-  .640  .019  Resumed Smoking  Finally, Cox's proportional hazard regression analyses were used to determine which variables remained significant considering all variables at all points in time. Variables were selected for entry into the first step of a backward regression model either because they were statistically significant in the univariate analyses or because they represented the variables as outlined in the primary or secondary research questions (see Table 21). Results of the multivariate analyses after a total of three steps revealed three predictive variables of breastfeeding to include higher breastfeeding confidence (p = .001), non-smoking status (p = .017), and whether the participant had received a CHN visit within two weeks after birth of the infant (p = .023) (see Table 22).  Table 21. Cox's Proportional Hazard Regression - Full Model Variable  Adjusted OR  95% CI  Sig.  Allocated to Intervention 1.800  .882-  3.672  .106  More Commitment  .213  .025 -  1.823  .158  Mother Breastfed  .754  .372-  1.527  .432  Resumed Smoking  .175  .050-  .610  .006  CHN Visited  .375  .182-  .773  .008  Postsecondary  .896  .350-  2.297  .820  Vaginal Delivery  .938  .367-  2.399  .893  Supplemental Feeds  1.811  .832-  3.943  .135  More Knowledge  1.057  .995 -  1.124  .074  More Confidence  .976  .946-  1.006  .120  More Satisfaction  .981  .948-  1.015  .267  1.040  1.001 -  1.080  .044  .914  .781 -  1.163  .463  More Fatigue More Depression  112  Table 22. Cox's Proportional Hazard Regression - Final Model Variable  More Confidence Resumed Smoking CHN Visited  Adjusted OR  95% CI  Sig!  1.030  1.012-  1.049  .001  .283  .100-  .797  .017  1.105  1.106-  4.000  .023  Qualitative Outcomes Participants were asked two open-ended questions and outcomes were analyzed as one group. The first question asked all of the participants to indicate any antenatal concerns pertaining to breastfeeding. The second question, raised by the research assistant during the telephone interviews at 2, 12, or 24 weeks but only if a participant indicated she had weaned her infant, determined how the participant knew when it was time for her to stop breastfeeding. Using content analysis, a numbering system was developed for themes that emerged from the written comments pertaining to antenatal breastfeeding concerns, and comments about how a woman had known it was time for her to stop breastfeeding. Themes were developed independently; first by the principal investigator, and then by a colleague with expertise. For example, the theme "insufficient milk" was ascribed for written commentsfromthe women that included "the well was dry," "I just didn't have enough milk," or "my milk was drying up." The naming of the theme related to the word mostfrequentlyused in the comment and/or the overall impression as to the intent of the participant. Themes were grouped according to whether they applied most to the mother, the baby, or the dyad. When a code did not fit into a major theme,  113  such as 'we never did start' it was coded as 'other.' When asked antenatally to indicate whether they had concerns about breastfeeding and if so to indicate what concerns they were experiencing, 130/209 (62%) of participants indicated at least one concern. The concern reported the most frequently pertained to breastfeeding technique or "not knowing how to do it." Other examples of how this concern was reported included: "my friend had trouble learning how so I'm wondering if I'll have trouble," "I feel like I need to learn more about how to breastfeed," and "it sounds harder than I thought." The second most frequently reported concern pertained to the fear of pain while breastfeeding. Examples of this theme included: "handling the pain," "blisters and bleeding," and "chapping, soreness or cracking." The remainder of the themes that emerged from the participant data in decreasing frequency included concerns about an insufficient milk supply, the baby not gaining enough weight or worrying about the infant's health overall, how to manage breastfeeding in public or other concerns. Examples of participant responses that were coded as "other" included "inverted nipples," "return to work," "whether the baby would switch later," "knowing how long to continue," "my breast is more a sexual part of my body," and ""my baby will be born early so I won't get a chance to breastfeed." At 2, 12 and 24 weeks after birth, participants reported the following responses when asked how they knew it was time for them to stop breastfeeding: 1. by two weeks after birth a total of 14 women had weaned: reasons included sore breasts and nipple pain (n=3), insufficient supply (n=2), frustration (n=2), infant weight loss (n=2), nipple inversion or surgery (n=2), baby seemed unsatisfied (n=l), not working for them (n=l), and had  114  planned to wean by then (n=l). 2. by twelve weeks after birth a total of 25 women had weaned since last contact: reasons included insufficient supply (n=9), infant weight concerns (n=4), baby refusing breast (n=4), maternal fatigue (n=2), mastitis (n=2), allergy to milk (n=l), wanted more freedom (n=l), vasospasms (n=l), and mom no longer interested (n=l). 3. by twenty-four weeks after birth a total of 18 women had weaned since last contact: reasons included insufficient supply (n=6), infant teething (n=2), return to work (n=2), just had enough (n=2), maternal health (n=3), convenience once weaned (n=l), too tired of pumping (n=l), and infant preferred expressed breast milk (n=l). When all the themes were compared, the most commonly reported reason for weaning up to the time the infant was six months old related to the mother's perception she had an insufficient supply of breast milk followed by the infant's perceived or actual response to breastfeeding, and interference with other aspects of activities of daily living. Summary of Findings Findings of the trial which tested the effect of an antenatal intervention on breastfeeding duration revealed no difference between the two groups with respect to the proportion of women who continued to at least mainly breastfeed by 24 weeks after birth. However, women in the intervention group were significantly more likely to seek assistance by two weeks after birth, and the help was most often forthcoming from a community health nurse. Women assigned to the control group sought more help later than 12 weeks after birth and the help most often was from the La Leche League, the Breastfeeding Hotline, friends, or their mother.  115  Secondary research questions were analyzed to determine whether significant differences existed between the two groups in terms of the five variables of the theoretical framework. Using repeated measures ANOVA, no differences were obtained when comparing the two groups in terms of their self-reported levels of parenting competence, breastfeeding confidence, fatigue, or depression scores. Significant differences were observed between the two groups in terms of maternal satisfaction (lifestyle/body image) subscale scores by 24 weeks after birth. Women in the control group reported significantly more depression antenatally, but the depression scores did not continue to be significantly different (p = .649) in the postnatal period. The secondary research questions were intended to contribute to further understanding regarding breastfeeding duration outcomes. Interestingly, although the maternal satisfaction (lifestyle/body image subscale) score was significantly greater among women assigned to the intervention group at 24 weeks after birth, there were not significantly more women assigned to the intervention group still at least mainly breastfeeding when compared to the control group. Univariate analyses revealed women were more likely to continue to at least mainly breastfeed if they had more educational preparation, been breastfed themselves, reported they felt committed to breastfeeding antenatally, delivered their infant vaginally, and/or been seen by a CHN within 2 weeks after birth. Women who resumed smoking or whose infants had received supplemental feedings while in hospital were more likely to wean earlier. Multivariate analysis revealed three predictive variables of breastfeeding duration to include greater breastfeeding confidence, non-smoking status, and if the participant had received a CHN visit within 2 weeks after birth.  116  Qualitative findings offered further insight about the breastfeeding problems expectant women anticipated, as well as the reasons why they knew it was time for them to wean. When expectant mothers were asked what concerns they were experiencing about breastfeeding, 38% did not report any concerns. For women who reported concerns, the most frequent responses were not knowing the technique, followed by fear of pain, and/or not being able to achieve an effective latch. When the women were later asked how they knew it was time for them to stop breastfeeding, the participants most frequently reported they did not feel they had an adequate supply of milk.  117  CHAPTER FOUR  DISCUSSION  In Chapter Four, primary and secondary study outcomes are discussed within the context of the literature. Components of conducting the trial are re-examined according to Moher et al.'s (2001) recommendations to strengthen clinical trials. Implications of the study for nursing practice, research, education, and health policy are outlined. Discussion of Findings Primary Outcome The trial was conducted to determine the effect of an antenatal educational intervention on breastfeeding duration. No statistically significant difference in breastfeeding duration was identified when the two groups were compared to determine the proportion of women who were still at least mainly breastfeeding, (i.e. women who were breastfeeding and offering up to one alternate feeding per day) by 24 weeks after birth. No other studies have been identified since the onset of this trial that aimed to test an antenatal educational intervention to prolong breastfeeding. However, two major intervention studies to extend breastfeeding duration have been published (Dennis et al., 2002; Kramer, Chalmers, Hodnett, Sevkovskaya, Dzikovich, & Shapiro, et al., 2001). Thefindingsof the PRomotion Of Breastfeeding Intervention Trial (PROBIT) study, provided clear evidence as to the efficacy of complex multi-site interventions  118  aimed at changing practices in hospitals and clinics as opposed to focusing on an intervention in an educational format in the antenatal period (Kramer, et al.). The intervention consisted of staff receiving extensive training emphasizing "methods to maintain lactation, promote exclusive and prolonged breastfeeding and resolve common problems" (Kramer et al., p. 414). The study demonstrated that infants allocated to the intervention group were significantly more likely to be exclusively breastfed at 6 months after birth (p< .001). Dennis et al.'s randomized controlled trial tested a telephone based postnatal peer support intervention, rather than an antenatal educational intervention, in a Canadian location (n=256). Significantly more mothers in the peer support group than in the control group continued to breast-feed at 3 months post partum (p = 0.01). The findings of this study testing an antenatal education intervention were consistent with previous randomized controlled trials testing the effect of an antenatal intervention where no statistically significant differences were observed (Hill, 1987; Kaplowitz & Olson, 1983; Loh et al., 1997; Serwint et al., 1996; Sheehan, 1999) (see Table 1). Outcomes of the study conducted by Sheehan were most comparable to the current investigation in terms of sample size, intent, and findings. Sheehan tested two strategies of providing an antenatal educational intervention. One participant group received informationfroma peer resource group, and the second group received comparable informationfroma nurse midwife childbirth educator. The current study tested the effect of a pre-workshop guide and interactive group session. Although there was no significant difference between the two groups, Sheehan noted women in both groups breastfed longer than national norms.  119  A similar result was obtained in this study in that 57% of women in the intervention group and 50% of women in the control group were still at least mainly breastfeeding by 24 weeks after birth. Each of these proportions exceeds the Manitoban provincial average of 36% as reported in the NLSYC (1996) (see Figure 1.). In addition, proportionately more women in this study breastfed longer when compared to a recently published provincial average of 42% in Manitoba (Martens, Derksen, Mayer, & Walld, 2002). The intent of the study was to affect a 20% increase in the proportion of women still breastfeeding beyond the provincial baseline statistic of 36% (see Figure 1). Fifty-seven percent of the women assigned to the intervention group continued to at least mainly breastfeed by 24 weeks after birth. Guise, Palda, Westhoff, Chan, Helfand, and Lieu (2003) would support this aim in that their systematic review and meta-analysis demonstrated, "there appears to be greater effectiveness of educational sessions in populations where the pre-intervention breastfeeding rate is less than 50%" (p. 73). However, 50% of the women assigned to the control group also continued to at least mainly breastfeed by 24 weeks after birth. Observing a statistically significant difference between the two groups may have been less likely, given the overall tendency for more of the women in each group to breastfeed longer than average. The fact that the proportion of women who continued to breastfeed is 7% higher in the intervention group suggests the educational intervention was helpful, yet insufficient, to create a statistically significant difference between the two groups. Some might argue that a 7% increase between the two groups is clinically important. But in order to determine if a 7% increase is statistically significant, a larger trial would be required to detect a smaller effect (Young,  120  BresnitZ, & Strom, 1983). A sample of 1063 women in each group, or a total of 2126 women, would be required to achieve power of .80 using a two tailed alpha at the .05 testing level to detect a 7% increasefroma baseline breastfeeding duration rate of 36% at 24 weeks after birth (Cohen, 1977). A sample of 838 women in each group, or a total of 1676 women, would be required to achieve power of .80 using a two tailed alpha at the .05 testing level to detect a 10% increase from a baseline breastfeeding duration rate of 36% at 24 weeks after birth (Cohen).  There are two possible explanations as to why women in this study continued to breastfeed longer than national or provincial norms. Women may have breastfed longer because they were participants in a research study that asked them to report on various aspects of their breastfeeding and first-time parenting experiences. It seems likely that women who volunteer to participate in breastfeeding research may have been more likely to seek knowledge about the benefits of breastfeeding, and therefore feel more determined to continue to breastfeed their infants. A second, and more likely reason why women in this study may have breastfed longer, is because their demographic profile and personal characteristics were similar to those identified by Scott and Binns (1999) as variables associated with prolonged breastfeeding duration. For example, participants had already made the decision to breastfeed, indicated they felt committed to breastfeeding, and reported they had supportive family and friends. In addition, the sample represented a middle-class, well-educated, non-smoking group. Collectively, these variables and the homogeneity of the sample probably contributed to the likelihood that these participants would breastfeed longer than average (Maclean, 1998; Scott & Binns, 1999).  121  The homogeneity of the groups in terms of the characteristics discussed is therefore a likely reason for the failure of the intervention to demonstrate a statistically significant effect on the primary outcome variable. Furthermore, the most recent Manitoban statistics report 42% of infants ever breastfed were still breastfeeding by six months after birth (Martens, et al., 2002). It is possible that the six month breastfeeding duration rate for this group of select women, even without an intervention, would have been higher than the 36% that was published at the time this study was designed. This possibility is supported by the proportion of women (50%) assigned to the control group who continued to at least mainly breastfeed by 24 weeks after birth. To propose a further increase of 20%, as projected in previous trials (Redman et al., 1995; Serwint et al., 1996; Sheehan, 1999) may not have been realistic for this sample, given the many demographic and personal characteristics associated with extended breastfeeding. Future investigations aimed to test a 20% proportional increase should therefore recruit from less homogenous samples that represent the actual breastfeeding population, or consider conducting a larger trial aimed to demonstrate a smaller effect. The findings of this study are unlike those reported in other related trials (Duffy et al., 1997; Kistin et al., 1990; Rossiter, 1994; Sciacca et al., 1995; Pugin et al., 1996; Wiles, 1984) that reported a significant increase in breastfeeding duration among women assigned to an antenatal educational intervention (see Table 1). It is therefore crucial to consider why this intervention did not demonstrate a significant effect on breastfeeding duration. The study design and intervention will be re-examined given the reason(sO) may be due to weaknesses in the design, the intervention, or both. Recommendations for future practice and research intended to  122  prolong breastfeeding duration are discussed in view of the study's findings. The Study Design The study design carefully considered the criteria used to appraise prior research as outlined in Table 2, and the recommendations of Moher et al. (2001) to strengthen clinical trials. A randomized, controlled design was used. Guided by the literature, factors associated with breastfeeding duration were used to establish participant inclusion and exclusion criteria to control for confounding variables. Responding to criticisms in the literature which indicated previous research was characterized by inconsistent breastfeeding definitions (Armstrong, 1991) standardized categories (Ellis et al., 1991) were used in this study to assess the amount of breast milk the infant consumed, and to facilitate more precise comparisons between the two groups. The number of women who withdrew from the study, and their reasons for declining to participate, were recorded. In addition, data entry specified if a woman had opted out of the study rather than assuming she had stopped breastfeeding. An attrition rate of 15% was both projected and observed. The number of women opting out of the study was comparable between the two groups by 24 weeks after birth. Only 82 women assigned to the control group were still participating in the study by 24 weeks after birth. Ideally, recruitment should have continued until the mode of group assignment was exhausted so that the groups were equal, and the possibility of retaining the minimum number of women needed per group (n=87) was increased. Although blinding of the participants as to group allocation was not possible, the women were not aware that the study's primary outcome variable pertained to breastfeeding duration. The study introduction and consent form noted the researcher was interested in learning how  123  nurses can better assist first-time mothers. Participants in the intervention group were asked not to share the information outlined in the pre-workshop guide or group workshop to minimize contamination across groups. To determine how well the major instruments performed in this sample, testing was completed. Acceptable indications of the scales' internal consistency, reliability, and validity were demonstrated with the exception of one fatigue subscale measurement pertaining to frequency of fatigue. Data analysis procedures were completed as proposed using the intentionto-treat approach. Further analyses of the data was conducted to determine if there was a clinically significant difference using only the data reported by the 89 women who attended the workshop, out of the total possible 111 women assigned to the intervention group. No significant differences in breastfeeding duration were demonstrated. In summary, the study design has been re-examined in relation to the primary outcome and the criteria used to appraise previous breastfeeding trials. The design should have been sufficient to detect a statistically significant difference between the two groups with respect to the proportion of women still at least mainly breastfeeding by 24 weeks after birth. However, observing a 20% proportional increase between the two groups would be unlikely if the baseline breastfeeding duration rates were higher than originally identified. In addition to the reexamination of study design elements, a review of the intervention components will now be discussed in view of the fact that no statistically significant difference was detected between the groups.  124  The Intervention  A review of the literature revealed a limited number of randomized, controlled trials (n =11) that had tested the effect of an antenatal intervention on breastfeeding duration. Outcomes of these trials were inconsistent. Although antenatal interventions appeared promising, there was a paucity of evidence to declare whether an antenatal educational intervention alone could prolong breastfeeding duration. No studies had been conducted to test an antenatal educational intervention to affect breastfeeding duration in Canada. Administration of an educational intervention during the third trimester of pregnancy was considered appropriate in view of Mercer's (1981; 1995) theoreticalframeworkand the recommendations of previous researchers (Hauck & Dimmock, 1994; McLeod, Pullon, & Cookson, 2002; Susin, Giugliani, Kummer, Maciel, Simon, & da Silveira, 1999). Inclusion of a workshop as part of this intervention was further supported by the findings of two systematic literature reviews published since the onset of this trial. Ines Couto de Oliveira, Comacho, and Tedstone (2001) concluded that only antenatal interventions that incorporated a group session were associated with extended breastfeeding duration. Guise et al.'s review and meta-analysis • (2003) concluded educational programs to be the most effective single intervention to positively affect breastfeeding initiation and short term duration. Antenatal intervention studies relying solely on the use of written materials to extend breastfeeding duration were ineffective (Kaplowitz & Olson, 1983; Loh et al., 1997). Two intervention studies that relied on distribution of teaching booklets after the birth of the infant were unable to affect how long the women continued to breastfeed, leading researchers to urge  125  expansion of the intervention in future investigations (Curro, Lanni, Scipione, Grimaldi, & Mastroiacovo, 1997; Hauck & Dimmock, 1994). In view of these findings, this researcher designed an intervention including a pre-workshop guide and an interactive group workshop. Contents for the guide and workshop reflected the scope of information recommended by previous researchers as being pertinent in assisting first-time breastfeeding mothers (Curro et al.; Hauck & Dimmock). Contents for the guide and group workshop also were given careful consideration so as to address the reasons Canadian women have reported for weaning, as well as strategies women have reported as being helpful in continuing to breastfeed. The focus of the guide and workshop relied on variables from the theoretical framework, and evidence from the literature regarding factors associated with early weaning or prolonged lactation. It was considered crucial to plan how best to highlight the most salient information while encouraging the participants to contribute their own ideas and experiences. It also was imperative to plan an intervention that was complementary to information already received during usual care practices. In the interests of providing salient information, participants were informed, for example, that first-time parents have generally reported that the amount of fatigue they had experienced after birth far exceeded what they had expected. Participants in this study were asked how they might plan to minimize the effect of fatigue. What would help them to feel less tired? Who could help them with other household responsibilities especially during the first few weeks after the birth of their infant? In the opinion of the researcher, it was this type of individualized exploration and/or approach that made the content of the intervention workshop different from  126  what they may already have been told in the existing prenatal class series (usual care). Time was not allocated during the usual prenatal classes to explore new ways of preventing, or minimizing, the reasons women have reported as to why they have weaned their infants. The pre-workshop guide and interactive workshop were, therefore, designed to provide additional ideas about how to manage when challenged by common problems and to assist women to breastfeed longer. However, despite the fact that issues such as concern about technique, or wondering if the baby was getting enough milk (Hill, 1991; Hill, Humenick, & West., 1994; Stainton, Murphy, Higgins, Neff, Nyberg, & Ritchie, 1999) were discussed in the group session, the intervention did not prevent mothersfromweaning for those reasons. It is possible the scope of the current intervention was too broad to adequately address the most salient reasons women choose to wean. Future interventions may be more effective if the focus was reduced to only these aspects. It is also possible participants or support persons may have been less involved in the preworkshop activities, and at times participants were less interactive during the group workshops, than originally intended in the study design. It is unknown whether the participants in the intervention group used the guide as a resource after the birth of their infant. Although expectant fathers or other support persons were welcomed to participate in both components of the intervention, the intervention may have needed more content that was intended especially for them to increase their interest and participation. Careful consideration of these elements of design is thus warranted prior to future use. Perhaps a series of group sessions would facilitate even more rapport, peer support, and time to adequately discuss relevant material.  127  In summary, an antenatal educational intervention comprised of a pre-workshop guide and interactive group session was designed to prevent or minimize the reasons Canadian women have reported as to why they have weaned and to share strategies women have reported as being helpful to continue to breastfeed. However, the intervention did not result in significantly more women in the intervention group at least mainly breastfeeding 24 weeks after birth as compared to women in the control group. Secondary Outcomes Secondary outcomes were analyzed to better understand breastfeeding duration. Even though at baseline, the control group had reported significantly higher depression scores, they did not continue to be significantly different when tested at 2, 12, or 24 weeks after birth. The only significant difference between groups revealed by repeated measures ANOVA testing pertained to the amount of satisfaction women reported by 24 weeks after birth (p = .049). Although this is difficult to interpret, it is possible that the content of the educational workshop reinforced the positive aspects of breastfeeding for women, their babies and their family. It is also possible that early contact with the CHN, enhanced their breastfeeding confidence leading to increased satisfaction and prolonged duration. On the other hand, women in the control group may have tried to deal with their feeding problems alone and later sought assistance when they were feeling frustrated or unsatisfied and lifestyle issues were more entrenched. Given there was limited variability between and within the groups, subsequent analysis of the secondary outcomes were analyzed by using the reported scores of all women in the study. Thefindingsrevealed that by two weeks after birth, women who reported more competence on  128  the knowledge subscale were slightly less likely to still be at least mainly breastfeeding than women who reported less knowledge. This finding seemed clinically unlikely. Responses were re-examined to confirm data entry had been completed accurately. It seems logical the finding may be due to an error arising from the number of comparisons that were undertaken. Subsequent comparisons between groups of the knowledge subscale scores at 12 and 24 weeks after birth were not significant. Women in this study who reported more breastfeeding confidence and satisfaction at 2, 12, and 24 weeks after birth were significantly more likely to still be at least mainly breastfeeding at all three of the same test times (p=.000). This finding is consistent with the literature reviewed to design the intervention, and with more recent publications. For example, Chezem, Friesen, and Boettcher (2003) recently reported longer duration of breastfeeding by women who reported higher breastfeeding confidence scores. In addition, their investigation found a significant association between breastfeeding confidence and maternal competence. In a recent publication, Creedy, Dennis, Blyth, Moyle, Pratt, and Vries (2003) reported outcomes of testing the Breastfeeding Self-Efficacy Scale (BSES) antenatally, and at one week and four months postpartum. In a sample of Australian women (n=276), mothers with higher antenatal BSES scores were significantly more likely to be breastfeeding at one week and four months after birth (Creedy et al.). Consistent with self-efficacy theory, which suggests that repeatedly accomplishing successful breastfeeding increased a mother's perception of her ability to breastfeed (Dennis, 1999), the one week BSES score in the current study was significantly higher than the antenatal mean score. Cooke, Sheehan, and Schmied (2003) have reported low total  129  satisfaction scores using the Maternal Breastfeeding Evaluation Scale (MBFES) were significantly predictive of premature weaning suggesting high satisfaction scores would be more likely among women continuing to breastfeed as found in this study. Women who reported more fatigue at 2 weeks after birth were significantly less likely to be at least mainly breastfeeding by 24 weeks after birth. Women reported the highest level of fatigue by two weeks after birth, a finding that is consistent with the observations of Milligan et al. (1997) and Gardner (1991). Women assigned to the control group in this study reported slightly more fatigue and more depression at each test time, but the differences were not statistically significant. Bozoky and Corwin (2002) have reported fatigue as early as seven days postpartum to be predictive of depression by one month after birth (p < .05). In this study, women who reported more depression at 2 weeks after birth were significantly less likely to be at least mainly breastfeeding by 2 weeks after birth (p= .041). A subsequent analysis of the depression scores revealed 34% of the women in this study scored a minimum of nine on the Edinburgh Postnatal Depression Scale, suggesting they may be mildly depressed (Cox et al., 1987). This subgroup of women were significantly more likely to have weaned their infants by 24 weeks after birth (p = .041). Thefindingis consistent with the literature cited earlier linking the onset of depressive symptoms with premature cessation of breastfeeding (Cooper, et al., 1993; Fleming et al., 1988; Galler et al., 1999; Henderson et al., 2003; Misri, et al., 1997). In a more recent publication by Dennis (2004), 30% of the Canadian mothers she studied scored nine or more on the EPDS within two weeks after the birth of their infants. Dennis'sfindings,and the results of the current study, indicate substantially more  130  women may be experiencing postpartum depression symptoms than previously documented (Beck, 1998; O'Hara & Swain, 1996). Further research to assist women who are identified as depressed is thus essential as one of the benefits of reducing depression includes the potential of longer breastfeeding duration. In addition to the analysis of the central variables, all demographic and delivery characteristics were examined in relation to whether the mother was continuing to at least mainly breastfeed (see Table 14). Mothers who did not smoke, or who had completed postsecondary education were more likely to be at least mainly breastfeeding at 2,12, and 24 weeks after birth. Infants who had not received supplemental feedings were more likely to still be at least mainly breastfeeding by two weeks after birth. Women who delivered their infants vaginally, or had received a visit by a community health nurse (CHN) were more likely to still be at least mainly breastfeeding by 12 weeks after birth. Women who had been breastfed as infants were more likely to still be at least mainly breastfeeding by 24 weeks after birth. Relatively few mothers (12/185 or 6%) in this study reported they had resumed smoking, yet this variable continued to be a significant predictor of breastfeeding duration, a finding that is consistent with related studies (Hill & Aldag, 1996). A consistent negative association has been reported between maternal smoking habits and breastfeeding duration according to Scott and Binns's (1999) literature review. A trial conducted by Johnson, Ratner, Bottorff, Hall, and Dahinten (2000) tested a postpartum intervention to prevent smoking relapse and demonstrated women in the control group were twice as likely to resume daily smoking (p = .03) by six months after birth. Secondary analysis of the data further revealed that 65% of  131  women who relapsed to daily smoking weaned prior to 26 weeks (Ratner, Johnson, & Bottorff, 1999). Therefore, the study by Ratner et al. supported the association between smoking resumption and early weaning found in this study. Women with postsecondary education, in this study, were more likely to at least mainly breastfeed longer. The positive association between higher education and continued breastfeeding has been well documented in the literature (Avery et al., 1998; Boettcher et al., 1999; Dennis, 2002a; Scott & Binns, 1999). The finding that supplemental feedings negatively affected duration is of concern. Despite the growing body of evidence and apparent institutional support to promote breastfeeding, infants continue to receive supplemental feedings at rates that far exceed recommended levels. In a study conducted in Quebec by Loiselle, Semenic, Core, Lapointe, and Gendron (2001), 68% of first-time (n=108) mothers reported their infants received supplements of water or formula. Redman et al., (1995) reported continued breastfeeding by four months was more likely amongst women whose baby did not receive a bottle feed while in hospital (p = .023). Forty-two percent of the infants in this study received supplemental feedings while in hospital according to the data reported by the mothers. In contrast, according to a report of the Breastfeeding Committee for Canada of guidelines for the WHO/UNICEF Baby-Friendly Initiative in Canada (2003), only infants otherwise well enough to be with their mothers while in hospital diagnosed with very specific medical conditions or those infants whose mothers are ill, are justified to receive any nutritional supplements. The mothers and babies in this study were rarely challenged by any medical condition and so it is questionable why so many babies  132  received the additional nutritional sources. In addition to the finding regarding supplementation of feedings during hospitalization, 76% of women in this study had introduced solid foods by three months after birth, whereas, the Canadian Pediatric Society, Dieticians of Canada, and Health Canada (1998) have recommended delaying the introduction of solid foods until the infant is six months of age. If breastfeeding duration rates are to increase in Canada, then more attention must to be given to determine how parents can be better educated to provide the best nutritional practices for their infants and small children (Bernard-Bonnin et al., 1989; Perez-Escamilla et al., 1994). Only 63% of the women in this study delivered their infants vaginally; 37% of the infants were delivered by cesarean section. The rate far exceeds the national average of 22.5% as recently reported by the Canadian Institute for Health Information (2004). The report links the increase in birthing by cesarean section to a decline in the number of family doctors providing full maternity care. Health Canada's (2003) Canadian Perinatal Health Report links the increase in the rate of cesarean deliveries to the trend that more primiparous women are now older and to fact the majority of births have been conducted in large urban centers. These variables match the characteristics of the women in this study and may explain the particularly high rate of cesarean deliveries observed. During the interactive group session, women were encouraged to contact their CHN early once home from hospital after the birth even if they were not having any breastfeeding problems. The women were told that other women had reported it was helpful to have someone confirm  133  things were going well. By two weeks after birth, the frequency of breastfeeding help requested by women in the intervention group was significantly higher (p= .021) than the amount of help requested by the women in the control group, and the help that was requested was that of a community health nurse (CHN) (p— .041). Also by two weeks after birth, the women in the intervention group reported slightly lower mean scores on the lifestyle subscale as compared to the women assigned to the control group. Differences in terms of the amount and type of help requested are viewed as outcomes of the intervention effect since women in the intervention group were encouraged to seek early help, and to consult their CHN about support for breastfeeding especially if they were experiencing any difficulties. In contrast, women in the control group sought help significantly more often from the La Leche League (x = 4.757, p = .029), the Breastfeeding Hotline (x = 5.780, p = .016), a friend ( 2  2  2 x  = 4.004, p = .045), or their mother (x = 5.007, p = .025) between four and six months after birth 2  when compared to women in the intervention group. It is unclear why the women in the control group would seek assistance significantly more often at a later time. Perhaps women in the control group felt more determined to continue to breastfeed despite not having received the potential benefit of the study intervention, or their persistence was related to a Hawthorne effect, i.e. knowing they were participating in a research study promoted their determination to continue the behavior. Also, their mean scores on the lifestyle subscale were lower than the scores of the women assigned to the intervention group by 24 weeks after birth. Therefore, it is also possible women in the control group sought more help later related to lifestyle issues in terms of continuing to breastfeed. Perhaps they were less satisfied while trying to manage breastfeeding  134  as they resumed aspects of their lifestyle because they did not have the opportunity to better anticipate strategies to consider as discussed in either the pre-workshop guide and/or interactive group session. The final variable that was significantly related to continued breastfeeding at 24 weeks after birth pertained to whether the mother had been breastfed. The literature supports a positive association between women continuing to breastfeed when those close to them (i.e. friends, family, or other) have had positive breastfeeding experiences (Chezem et al., 2001). Predictive Variables The study yielded three predictive variables of breastfeeding duration; maternal breastfeeding confidence, non-smoking status, and whether the mother had been visited by a community health nurse within two weeks of the birth. Maternal breastfeeding confidence, now considered to be a significant and modifiable predictor of breastfeeding duration (Blyth, Creedy, Dennis, Moyle, Pratt, & Devries, 2002; Creedy et al., 2003; Ertem, Votto, & Leventhal, 2001), was the most significant variable to predict breastfeeding duration in this study. The preworkshop guide and workshop were designed to enhance breastfeeding competence, confidence, and satisfaction. Women were told, for example, that other women had reported it had helped them to identify a person who had a positive breastfeeding experience, especially if it was someone who had been successful despite experiencing breastfeeding challenges. Study participants also were asked to think about other events in their lives, and to consider how successful they thought they had been when faced with learning new behaviors or finding new  135  ways to manage. However, despite the use of these strategies within the intervention, there were no significant differences in terms of the amount of breastfeeding confidence reported between the two groups. Perhaps the women in this study were more confident about breastfeeding because they were more likely to have completed postsecondary education, reported supportive family and friends, and indicated high levels of commitment to breastfeeding. The intervention referred to the importance of not smoking while breastfeeding, the risk of the baby inhaling second-hand smoke, and to the outcomes of previous research linking smoking resumption with premature weaning. No significant differences were detected regarding the number of women who resumed smoking who had been assigned to the intervention group as compared to the number of women who resumed smoking who had been assigned to the control group. It is likely women in this study received information about the importance of not smoking especially while pregnant or breastfeeding through public media announcements and/or have been influenced by recent municipal decisions to ban smoking in public places as only 6% of women reported smoking resumption. Results of this study also demonstrated that women and their infants benefited from visits by a community health nurse (CHN) within two weeks of the birth. Participants in the intervention group were encouraged to seek help early and to request CHN visits for help and reassurance that they were doing well breastfeeding their infants. Usual care practices mandate telephone contact with each new mother (or a home visit if no telephone is available) within the first 24 hours of hospital discharge. The nurse offers to make a home visit within seven days of hospital discharge if no particular problem is identified, or sooner if the mother indicates a  136  problem or concern. Women i n the intervention group were encouraged to not only seek early assistance once home, but to benefit from the expertise and encouragement of their C H N even i f no problems were apparent. A s Hauck, Langton, and Coyle (2002, p. 12) suggest, " i f health professionals wish to promote breastfeeding, they must not underestimate the importance of providing reassurance, as this feedback is crucial for breastfeeding mothers struggling along their 'path of determination' to overcome difficulties and achieve breastfeeding targets."  Qualitative Outcomes In addition to the quantitative data, qualitative data was obtained to seek further information about the antenatal breastfeeding concerns women experienced, as well as how women knew it was time for them to wean. Previous trials, as shown in Table 1, did not include qualitative data components. W h e n asked antenatally about 'what breastfeeding concerns they were experiencing,' women reported they were most concerned about technique, pain, or not having enough m i l k to meet their infants' needs. When asked 'how they knew it was time for them to stop breastfeeding,' the women most frequently reported they perceived they did not have enough breast m i l k to sustain their infant. However, it is possible some women may have felt more comfortable in offering this "more socially acceptable" answer, but actually believed they had stopped for other reasons such as the personal inconvenience or the amount time that is required to breastfeed ( H i l l , 1991). The findings i n this study resemble the findings o f Redman et al.'s (1995) trial where women were asked about the reasons they had stopped breastfeeding. The main reason for discontinuing breastfeeding by four months also pertained to concerns about "not enough milk."  137  The finding also is consistent with outcomes reported by Binns and Scott (2002) who reported "anxiety over the sufficiency of breast milk supply was the most serious problem in that it often resulted in the cessation of breastfeeding" (p. 13), and Hill's (1991) analysis of perceived insufficiency as a major reason for breastfeeding failure. Antenatally, women also reported either "not knowing how to do it" or concerns as to "whether she (and the baby) would achieve an effective latch." Feeling unfamiliar with how to breastfeed, or wondering if one will be successful in learning the technique, are two challenges which continue to require attention if breastfeeding rates are to increase. Perhaps women born and raised in Canada may be more concerned prior to the birth about being able to breastfeed, because they have had few opportunities to observe other women breastfeeding while growing up. In this study, "whether the mother had been breastfed" was significantly associated with continued breastfeeding. Efforts to support women who initiate breastfeeding now, may have even more long-term beneficial effects, in that more future mothers may have been breastfed, and therefore may be more likely to breastfeed their children. The intervention included content pertaining to decreasing pain or discomfort if experienced. Strategies also were shared as to how women (and their partners or families) could assess whether the baby was receiving an adequate amount of milk, such as the number of voids or stools per day, or feeling the difference in the breast after a breastfeeding session. A l l participants were urged to seek early professional assistance if they felt pain, or were concerned about their supply of breast milk. The fact that other mothers had reported weaning for these reasons also was shared with the participants. However, the intervention was not effective in  138  extending duration outcomes as women indicated these same reasons for weaning. Recommendations for Future Research The findings o f this study lead to suggestions for future intervention studies intended to prolong breastfeeding duration. Methodological aspects o f future study designs and interventions require careful consideration.  Methodological Considerations A future study o f the effect o f an antenatal educational intervention should extend the comparison o f breastfeeding duration between groups until one year after birth. Trends i n the data collected i n this study suggested significant differences may have been obtained i f the participants were contacted until 12 months after birth. Specifically, more than one-half o f the data (119/209 or 57%) from the sample was censored i n the survival analysis because the women were continuing to breastfeed beyond 24 weeks. It would be important to determine how the Kaplan Meier curves compared after 24 weeks by projecting a longer time-to-event outcome (Pocock et al., 2002). It would also be worthwhile to consider plotting the curves i n an upward direction (Pocock et al.). Fatigue levels o f women i n the control group continued to be higher than the levels reported b y women assigned to the intervention group at all four test times, although the differences were not statistically significant. Again, it may have been useful to continue data collection until one year after birth i n order to examine and compare later fatigue scores. It may also be advisable to consider use o f an alternate instrument to measure fatigue that has been tested among childbearing populations as items o f the Multidimensional Assessment o f Fatigue  139  (MAF) Scale did not load as anticipated in this study. The MAF Scale has been used primarily to measure fatigue among those diagnosed with either rheumatoid arthritis or cancer. As stated earlier, it is recommended future trials also target less homogenous populations to represent actual baseline breastfeeding rates, or plan a large trial to examine a smaller, yet clinically meaningful, effect size. This study was completed within a small geographic area and had a homogenous sample so generalizability to other areas, and other populations, was not possible. With ethical approval, data obtained in this study of Canadian women could be used to further test the psychometric properties of some of the major instruments. For example, since the beginning of the study, Dennis (2003) has since published a shorter version of the Breastfeeding Self-Efficacy Scale (BSES) containing only 1 4 items. Further analysis could be conducted on the data obtained for this study using only those items retained by Dennis. Of particular interest, scores on the Edinburgh Postnatal Depression Scale (EPDS) were used in Dennis's study as one measure of comparing outcomes of the revised BSES with theoretically related concepts, so further examinations of construct validity may be possible using the data collected in this study. Other recommendations for future research include the expansion of provincial and national health databases to track the number of women who initiate, and continue to breastfeed, while in hospital, and once home, to facilitate an examination of overall trends. Chart audits and retrospective studies could further assist in determining the rate of supplemental feedings that are  140  being offered to infants, compared to the number of medically indicated instances. Similar methods can be used to document the introduction of solid foods, and associated trends in care such as increased contact with new mothers by health care professionals to emphasize the health benefits for the child by continuing to offer the infant only breast milk until he/she is six months of age. Research could be completed using feeding data recorded electronically during each contact with a mother or infant, such as during follow-up immunization appointments in the doctor's office or community health unit. Recommendations for Intervention Design The intervention in this study was designed for women able to read and understand English, attend an additional prenatal class on a university campus, and women who may be willing to participate with others in an interactive, facilitated session. With modification, the antenatal educational intervention proposed for use in this study may have potential for use with alternate populations of women, such as those who are less educated or reported less support. Modification may include altering the location of a group session to facilitate attendance. Content of the pre-workshop guide might be better shared in a more direct, and/or personal way during a small group session, rather than in a written format as provided in this study. Research should be conducted on an intervention that is more interactive and more focused on the individual concerns of women in the group than was the intervention in this study. For example, a recent matched cohort study (n = 458) done to compare the effect of a group approach reported... "group prenatal care provides a structural innovation, permitting more time for provider-patient interaction, and therefore the opportunity to address clinical as  141  well as psychological, social, and behavioral factors to promote healthy pregnancy" (Ickovics et al., 2003, p. 1051). Ickovics et al. determined birth weight was greater for infants of women in group versus individual prenatal care leading them to plan an upcoming trial to test the effect of a facilitated group session. Although the aim of that study did not pertain to breastfeeding duration, the findings lend further support to the importance of a strongly facilitative approach to group learning to effect positive perinatal outcomes. More interactive sessions might include a group session whereby a facilitator is asked to lead a discussion on a range of topics such as perceived insufficient milk supply, how to seek help early if pain is being experienced while breastfeeding, or how to breastfeed when outside the home. More research is necessary to better understand how to assist women to determine they have an adequate milk supply, how to assure women when their babies are responding well to breastfeeding, and how to incorporate breastfeeding while proceeding with aspects of their lifestyle. Qualitative inquiries should be conducted to gain further insight into strategies deemed useful from the perspectives of breastfeeding women. One of the key sources of support involves inclusion of those deemed most important to the expectant mother (Arora et al., 2000; Chezem et al., 2001). Fathers could be encouraged to attend group sessions planned to discuss their pivotal role in terms of continued breastfeeding. Although partners, or other support persons, were welcome to review the pre-workshop guide with the participant and/or attend the group session, only portions of the content were designed specifically for them (see Appendices I and II). Future interventions should include more emphasis targeted for the woman's partner or support person. Information could include how  142  fathers (or others) can further support their partners and infants as they are learning breastfeeding techniques, or how to support them to continue with breastfeeding even when faced with common problems or challenges (Bar-Yam & Darby, 1997; Freed et al., 1992). Tiedje and Darling-Fisher (2003) contend that a major challenge for health care providers pertains to how to better include fathers. In response they have published an article outlining steps that should be considered when planning future interventions to extend breastfeeding duration in order to heighten interactions between providers and fathers. Therefore a study of a father-focused antenatal intervention is warranted. Future intervention studies should consider inclusion of both an antenatal education component, and a peer support component (Guise, et al., 2003; Vari, Camburn, & Henly, 2000). In this study, a visit by a community health nurse (CHN) was predictive of continued breastfeeding at 24 weeks after birth. A study of a C H N intervention aimed at increasing the duration of breastfeeding should be designed. Community health nurses could be specifically educated to provide detailed and individualized support, as recommended by Hauck et al. (2002) following an assessment of the many factors in a woman's home and community life that may be having an impact on her ability to cope with the complex demands of breastfeeding (Scott & Binns, 1999). Pugh, Milligan, and Brown's (2001) research supports the potential benefit of a multifaceted strategy on breastfeeding duration. A pilot study (n=20) explored whether a focused intervention offered by a breastfeeding support team (community health nurse and peer counselor) providing hospital visits, home visits, and telephone support yielded improved  143  breastfeeding outcomes. Women who received the intervention were more likely to continue to breastfeed by five months after birth than those who received usual care. Of particular interest, is their preliminary finding that their intervention might have reduced fatigue, depression, and anxiety. In more recent research, using the same intervention strategy, Pugh, Milligan, Frick, Spatz, and Bronner (2002) have reported a randomized trial of 41 women examining comparative costs of continued breastfeeding with the breastfeeding support team. Their findings, and the findings of this study, suggest a larger three group trial should now be conducted to compare the costs and consequences of a breastfeeding support team, a CHN visiting all new mothers within two weeks of delivery, and usual care. Such an intervention would need to be adapted to mesh with the philosophy of the Canadian Health Care System, and individual community health units. Studies of a community health intervention, or one of a CHN combined with peer support would also increase understanding of the relationship of this kind of support to breastfeeding confidence in women. Because of the increasing cesarean section rates in Manitoba and in Canada, further study to identify the impact of this trend on breastfeeding initiation and duration rates should be conducted. Rowe-Murray and Fisher (2002) found that initiation was delayed after cesarean delivery, but that there was no difference in breastfeeding duration rates. Since of the time of her work, however, cesarean delivery rates have escalated. Implications for Practice The study constitutes a significant contribution to health care planning in that it has tested the effect of a nursing intervention. Breastfeeding confidence was a significant predictor of  144 duration in this study; new ways of reinforcing breastfeeding confidence are needed i f women are to continue to breastfeed longer. The study supports the value of early C H N visits for all primiparous women who are breastfeeding, even i f the woman is not experiencing breastfeeding related problems. Contact with a health professional should include careful assessment of fatigue and onset of depressive symptoms. Expectant women could be further encouraged to think about ways they can conserve their energy. Partners and family members should be urged to seek assistance if they feel concerned about the woman's emotional status. The release of Guise et al.'s (2003) systematic review and meta-analysis constitutes "the first analysis of all breastfeeding interventions singly and in combination in developed countries" (p. 77). In response to the review, The Canadian Task Force on Preventive Health Care has released six major recommendations on interventions to promote breastfeeding (Palda, Guise, & Wathen, 2004). The first recommendation supports provision of structured antepartum educational programs and postpartum professional/peer support to promote breastfeeding initiation and duration. That recommendation fits with the findings of this study because the antepartum intervention resulted in a proportional increase of 7% of women continuing to at least mainly breastfeed by 24 weeks after birth. Subsequent recommendations support provision of rooming-in, and early maternal contact to promote breastfeeding. The recommendations of the Canadian Task Force on Preventive Health Care do not support sole use of written materials, or commercial discharge packs to new mothers. Finally, the report acknowledges there is insufficient evidence to make a recommendation regarding advice by primary health care providers (physician or midwife). However, this study demonstrated that first-time mothers  145  breastfed longer when they had received a home visit from a community health nurse. Nurses, in their practice, must adhere to the evidence-based recommendations such as those outlined above, and consider research outcomes such as the findings of this study, when revising practice protocols, to better promote prolonged breastfeeding in Canada. Nurses, and other health professionals, also must identify further means of integrating self-efficacy enhancing strategies to assist women to achieve their desired breastfeeding outcomes. The primiparous women in this study, who were well educated, committed to breastfeeding antenatally, and had extensive breastfeeding support from family and friends, still benefited from the CHN visit even if the reason for the visit was not because a specific problem had been identified. This professional support may be even more important among women experiencing any breastfeeding related difficulties. Women who are not able to identify family orfriendswho have had positive breastfeeding experiences are among those most likely to benefit from what health professionals and peer groups have to offer. Knowledgeable, consistent, and culture-sensitive support is needed throughout the experience. Both professional and peer support should be geared to introducing and maintaining confidence-building strategies. Curbing practices of unnecessary supplemental feedings in hospital for example, is likely to limit women receiving inaccurate messages that their milk supply may be inadequate. Not only did the findings of this study show nurses giving breastfed infants supplemental feedings in hospital adversely affected breastfeeding duration, but this occurred among a sample of educated mothers who then reported they weaned because they perceived they had an insufficient supply of breast milk. Use of supplemental feedings when not required for medical reasons is therefore likely to  146  deter other mothers who may not have had comparable opportunities to question if nutritive substances other than breast milk were necessary for their infants. Future interventions also should target women most at risk to wean early such as those who are younger with less education, those who have quit smoking during pregnancy, and those who have few social supports. A check-list of risk items based on empiric evidence could be used to identify women at most risk for early breastfeeding cessation. Beck (1998) recommends women should be screened for symptoms of depression during each contact with a health care professional until one year after birth. The findings of this study suggest even more women may be suffering from mild depression than previously realized. Screening for maternal depression antenatally, and then during each postnatal contact, such as during clinic visits for infant immunizations, is strongly substantiated by the findings of this study. Although as Cox and Holden (1994) warn, if screening is to be completed to identify women at most risk, suitable follow-up resources must also be in place. Checklists to identify women at risk for weaning also could be used by health care providers postnatally. For example, Martens (2000) identified two survey questions that could be used to identify women at high risk for weaning. The first question pertained to whether the mother felt satisfied with breastfeeding, and the second question identified the number of breastfeeding problems the mother was experiencing. Mothers who were not satisfied were 12.4 times as likely to wean compared with mothers who were satisfied (p =.0005). Mothers who reported more than one breastfeeding problem were 6.2 times as likely to wean compared to  147  mothers with one or no breastfeeding problems (p =.009). More evidence, such as the identification of specific questions to identify the most pivotal reasons for weaning will assist nurses by streamlining practice efforts. The findings of this study suggest an additional postnatal question for breastfeeding mothers such as, "To what extent are you satisfied with your supply of breast milk?" may help to identify women who require further assistance or reassurance to prevent premature weaning. Implications for Education In addition to the implications for research and practice already outlined, there also are implications for education. The literature review focused attention on antenatal educational interventions that have been tested to affect breastfeeding duration. The most recent systematic review of primary care-based interventions to promote breastfeeding supports educational programs to be the most effective single intervention (Guise et al., 2003). Furthermore, the meta-analysis conducted by Guise et al. revealed "a combination of education and support may be more effective than support alone, but not more effective than education alone" (p.70). Clearly, nurses involved in caring for expectant women and new mothers, including pre-natal class instructors, are obligated to learn more about how to better assist breastfeeding women and their partners. Prenatal class content should be evaluated regularly to confirm the content is current and salient. Group sessions could be planned to further encourage sharing of ideas and more active participation among expectant parents. Inclusion of new parents willing to share their experiences during antenatal sessions may facilitate peer support perinatally. Many people have  148  internet access either in their homes or in public libraries. Perhaps expectant women would appreciate the flexibility of having more resources available via electronic sources, rather than having to attend a series of prenatal classes in person. Montgomery (1999) reports that generalist health care providers are often inadequately prepared to assist women and their infants as they learn to breastfeed, especially if the new mothers are experiencing breastfeeding difficulties. Nurses, who are well informed about current practices in breastfeeding, could contribute further to the provision of accurate and consistent information about breastfeeding by presenting information at professional continuing education sessions and lay opportunities such as neighborhood health fairs. Community health nurses could allot more time to increasing public awareness about the many long term health and cost benefits associated with breastfeeding. In-hospital practices such as offering supplemental feedings to breastfed infants who are not medically compromised demands serious attention including continuing education opportunities for staff. Thefindingsof Taddei, Westphal, Venancio, Bogus, and Souza's (2000) randomized trial in Brazil supported a growing body of evidence that training hospital health professionals in breastfeeding promotion resulted in increased breastfeeding duration. Continuing education topics should be geared to increase adherence to the recommendations outlined by Palda et al. (2004) on behalf of the Canadian Task Force on Preventive Health Care. Nursing curricula have been criticized for not adequately preparing nurses about breastfeeding (Freed, Clark, Harris, & Lowdermilk, 1996). Undergraduate and advanced practice programs for nursing education should emphasize content that better prepares future  149  practitioners to assist breastfeeding women and infants. New teaching strategies could be incorporated into undergraduate nursing curricula, such as the use of interactive web sites that demonstrate how to achieve an effective latch when breastfeeding. Clinical instruction and observation during practice opportunities is essential for student nurses learning how to help new mothers with breastfeeding techniques. In addition to basic information about the process of breastfeeding, undergraduate nursing students should be aware of the most common reasons women report for weaning, as well as strategies to prevent early cessation such as how to determine if the baby is receiving enough breast milk, or how to help the mother try alternate breastfeeding positions if she is experiencing pain when breastfeeding. Nurses also need to learn more about how to better educate parents regarding the introduction of solid foods and how to help women to maintain smoking abstinence. Nurses also should find ways to more often include expectant fathers in antenatal learning opportunities. Women in this study were more likely to exclusively breastfeed longer when they felt their partners supported their decision to breastfeed. Tiedje and Darling-Fisher (2003) have suggested specific strategies to help health care providers involve fathers more; the suggestions could be considered for inclusion during practice opportunities. Sessions also should be designed that are aimed to more specifically address the needs of expectant and new fathers. Implications for Health Policy Helping women to breastfeed longer has profound health benefits. The empirical  150  evidence should be used to guide, and substantiate, the most appropriate allocation of health funding. For example, the first recommendation made by Palda et al. (2004) supports structured antenatal educational resources to promote breastfeeding. Existing pre-natal classes, however, may require evaluation to determine if the content related to breastfeeding duration emphasizes the most salient reasons why women wean. Communities that have reduced antenatal education opportunities such as prenatal classes must reconsider the potential negative affect the decision may have in terms of breastfeeding duration. In lieu of the more traditional series of prenatal classes, strategies such as web-based resources, or small group interactive sessions, could be offered focusing on what to expect in terms of breastfeeding, especially for first-time parents. Pairing women with peers antenatally with the expectation that the peer support would continue after the birth of the infant is one of the recommendations made by the Canadian Task Force on Preventive Health Care (Palda et al., 2004). Organization and training of a peer support program requires the expertise and assistance of a coordinator, a resource that may not yet be considered a priority by administrators who determine funding allocations. Evidence from research outcomes must be shared by nurses with those making policy decisions in order to establish the type of programs that are necessary to establish, and maintain, a peer support network to prolong breastfeeding. Redman et al. (1995) recommended programs designed to increase breastfeeding will need to address underlying factors such as hospital policy in view of thefindingsof their randomized trial. Their in-depth perinatal intervention that included an antenatal educational session, in hospital contact, and extra professional support once home failed to extend  151  breastfeeding duration. As stated, in-hospital practices such as routine offerings of supplemental feedings to breastfed infants, and increasing cesarean delivery rates adversely impacted the breastfeeding duration outcomes of the women in this study. Testing of antenatal interventions to determine efficacy must consider the potential effect of entrenched institutional practices. Policy makers must be made aware breastfeeding duration can be extended by antenatal interventions and home visit support of community health nurses (CHN). More resources are needed to assure enough community health nurses are available to conduct antenatal group sessions and postnatal visits to all first-time breastfeeding mothers. It is time funding priorities recognized more practice initiatives intended to promote healthy behaviors and confidencebuilding strategies, rather than triaging women to identify problems that justify additional professional assistance. At a societal level, the practice and benefits of breastfeeding require further understanding and appreciation. Women who continue to breastfeed could be better recognized and commended through national publicity campaigns that include opportunities to emphasize the many health benefits of breastfeeding. Policies could be expanded to further limit smoking or provide incentives for non-smokers. Nurses should be stronger advocates of breastfeeding and inform policy makers better of the potential health and cost benefits. More CHN support is necessary to educate parents regarding when to introduce solid foods given the findings of this study which demonstrated even well educated mothers commenced this practice much earlier than recommended. Dissemination of Findings Dissemination of study outcomes assists in guiding future health practice and research.  152  Participants were sent a one-page summary of the study's outcomes. Presentations will be held for interested staff of the Winnipeg Regional Health Authority (WRHA), university faculty and students, and a project summary will be sent to the WRHA senior administrative team. Initiative will be taken to present the findings at international, national, or provincial professional conferences. For example, abstracts submitted for presentation at the Canadian Association of Women's Health, Obstetric and Neonatal Nurses annual conference, the Mount Sinai Women's and Infants' Health Program Breastfeeding Conference, and the Manitoba Breastfeeding Research and Clinical Rounds have been accepted for presentation. The study will be included among topics available for presentation in the Sigma Theta Tau Speakers resource list for use among members, and the media. The study also will be noted among the Canadian Nurses' Foundation/Canadian Health Services Research Foundation (CNF/CHSRF) websites, and the Cochrane Date Base. Various formats, such as a lay abstract of the dissertation, will be forwarded to the CNF/CHSRF and Health Sciences Research Foundation, as agreed. Any reports will include an acknowledgement of funding received to support completion of the study. Study results will also be made available in appropriate written form for distribution in the form of a journal, newsletter, or care guideline format. It is recognized the study may be of interest among lay groups such as Y Neighbors, or Breastfeeding Drop-In Centers. Findings will be forwarded for publication consideration in research, education, perinatal nursing, and other disciplinary journals, as deemed appropriate. The completed dissertation will be made available for distribution through academic library systems. A summary of the study will be sent to the WRHA and Victoria General Hospital, as agreed in writing.  153  Summary and Conclusions Guided by Mercer's (1981; 1995) theory of maternal role attainment, the study was conducted to test whether an antenatal intervention was effective in extending breastfeeding duration. Women were randomly assigned to either an intervention or control group, and feeding data was obtained until the infant was six months old. The intervention consisted of a preworkshop guide and attendance at a workshop. An intention-to-treat method of analysis was used. There was no significant difference between the groups in terms of the proportion of women still at least mainly breastfeeding by 24 weeks after birth. Secondary outcomes pertaining to maternal competence, confidence, satisfaction, fatigue, and depression were analyzed to better understand factors influencing the duration of breastfeeding. Women in the intervention group reported more satisfaction 24 weeks after birth related to resuming aspects of their lifestyle while continuing to breastfeed. Multivariate analyses yielded three predictive variables of breastfeeding duration. The variables were higher breastfeeding confidence, nonsmoking status, and whether the mother had been visited by a community health nurse within two weeks of the birth. With some modification in focus and approach, the intervention used in this study should be further tested using a less homogenous sample. Targeting women who are most likely to wean early is a crucial next step. It also is recommended a trial be conducted to test the effect of a revised antenatal education intervention that includes further strategies to increase the involvement of partners, and a perinatal CHN/peer support component.  Co £  ii S3?;  5:  I  I it 1  I  O ^1  I 5>'  fb  I f  ! ! 5 I 1^! I 1 I If! I r I I !I i  o  s5  tI  I!  & rb or  t ft  — •> fb  ^  vQ  3"  Cu  I Cu i3. I  r6  fb  -3"  I  1  I  cn  a  O  3  •I* •!• CD  „  SL co  3" 3  era 3 &5  H  Oi  3*  cr  CD o  5' «= 00  P  8 g* § & J*  CD  CD CD  8.  *< 8 Cu  H 5'  CD  X X)  i-t c/i  co O  CD O  co CD CD  Cu I—»•  5  0  co' ^  3*  e  era  era  CD  x  CD  CD  CD  X)  cr  g  p co  CD  3.  CD  1  CD  CD  o  Cu  3  CD  Cu  •8 8 Cu  co  Q. 03  &  CfQ CD  X  CD  2.  1.1  O  3  «rt-  o t/3  CD P  3  1 ^ ^ co CD fD O  era cr 1-1  CD  Cu  2 ^ CD  xT  c cr CD  to Co  CD  g  M  o  |  K  3  C "  ^  3  s  8 S  C^ TO  O c  .rt.  Er  co  V  P  9  o  Cu  CTQ CD rt- co  TO a TO TO  o  (=1 CD CD CD  P  cr CD  P  OQ CD CD  CD  ^  *=* c r c r 3'  ^ °  cr n v: co 5p i-t (t m 0 c °2  1 £ ICu cuI Icr o  Cu  o  CD • < < co" O CO C3 cr g O 2 o CD • .ps C D t3 CD r*- *T3 ^ S i-t <CD gCDcs 3 CD 1-1  Cu  P,  CT  g rn 2  o  N  B  §  CO  2 cr o y  2 cs  eC cu CD S C' H  a p co CS  co CD  CD  co  EL- CD  00  & cr  P  D CD ^ ?r C Cu rt  s* 2  & o o cs o  CD  Cu  o  d  OCDQ rt-  cr  CD  i-t & g: ^ ° 3"  p.  CD CO  1 CD  X  o_  cr  CD  CD  lr o  & O  CD  P  CO rt-  CD CD  Cu  I—»•  cs  cro  o  o c  JC  C T CD  Cu  d  cr  ^4- d ; ?B  CU co CD  ^3  P  CO <rt  CD CD  Cu  5'  0Q  •>3  tr* &!• CD  o  B  D CD  g . co  3  o  s. cr  3 co co cr  3  P P  %  CO  3  co  e U  3cr  O  3  •3  co  5  co  3  a 3 I &  Cu  3 co  :  N> cr XS co P xi cr l-l I - I 3 c o CO 3 S tre Cu 3 p « CO o • O co i-b  co  r-*  CO  co  83  CO  CO  o  I-t  fS  1  cr o 3,  I-t  co 3  CD  Pi  CO  p  3 cT p Z 3 co d: cr e"re cr p 5 ? P.  CO  co  ~ s-o  co  ^.  CD  I^I 5 0  < >B o cr a CO ^co cT 3 co q B. CO 3 & cr cr NJ E o §• cT 3  ^  CD P ~  £ CO  CO  3  3  CO  o" co  CTQ CO  i-t  o o o  3  CO  X} CO  ft CD P  I  <  cr  pj cr 3  cu  P  5  o Cu &] co CO  P  v. co 3 Cu p & ere' cr CD r* ° ft, ^ o ccoo £L CO S" £ o Oi—" Cu cT^ i-+) ' co pu OO !"< CO ^ o cr. CD KJ H 3 a cr. <r 3  CD  p.  CO  co ^  CO C5  o cr i-t CO  P  ere co 8. P  co  CO  3 . co  CD CO  & 3  3 cu '-i  cr  3  "3  CO  cn co  tre cp  3  Cu O  1-1  Cu v< co" O  3 •  3-  3  o  o  3  o  co  p co |  -H  CP  CO  Cu  cr cr  cr. *o §  1^ E g*  Cu  C ft  c r ^ «> •• c (t  g w  Cu  s  CD  co  cr  IH CD co  g.  X  cr  o  d  r-+cr CD  cr CD  i-t c:  P  i-t-a %  O CT CT O CO O  |-i  «  co  § I-era 2; <r  2* o  CD  B.  a. CS  ere  g  p  g  cs  ere  cv c  o CO  I CS  CD  ^ M  O  •S  O  5f  § °  o | ^  2  S^  O  o  rt-  p.  °  ^2  Cu 3 p S3 CD  a  * CD  p  CP p  <r  8p  g  Cu  CD  CD < 2t 3> Cb S  3  CD  O  cr  CD  o  «-*O  CD  CS  o  3  cr 3.  £ * CD  i-t c r c r  CD  P &  3  cr o o  o cs co rt.  o*  d cr 2 <r CD CD  o  co  Cu  cr o  ^  i-t cr CD  CO  CO  < Cu'  CD  Sga  CO  cr &  »i  O  Cu CT rt.  V-  CO  CO  o P> 8 2  e: 3 ^ g o d v; P. cr. e § 3 Cu  i-t  d  c CT  p  o v: ^  rt  CD*  o  CO  8-  ^  cr o  V-  ere  CP  cr  vT  3  o  CD  o o  cr 5 2  r-t*  o  a  o o  ffi a ts i 2 Fr ^  CU CD  1-*-  '-t t-4  3 ^ 3 :  <g* cr g*  rt.  ES" cu  Cu  o  cr  ?r ere  o ^  cr  c+> cr  2  CO  CD O  5  cr  rt-  cr ha P  Cb  1b I T  co  CD  p^  &P  CD rt5  P  ^3  C D P  >-*» CV  S  CD  cr  ([  cr  2* p i  6  C  V>  cr P -< a. «cr  CD  —'  r-> CO  P  CB  co  S  CD n co !-5  Cu  cr P _* CS O gere **  cr  v:  © a  I  I! P  co  3 c r CD  L i  E  ere  cr cr  3 3  ^ c  5" 2- P  CD  CO  ftf O  s  cT  CD  cr  » a* r-t-  3  CD  5  CT> re s  CD ^  *  cl  ere  «  ft  CD  o w cs cs  2*  cs tr rt- o ?r co  co 5  ^ s>  If  &:  CD Cu CD Cu  w  &l  s ^ <' cr §  CD  5 o «3  "3 ^ - a  P  B.  P V"  § a CD cr  3 * CD  0  CD  a^o 2,  d  rt  P  cs g  cr  o. e^l  CD  O  n  o  P  ,Cu  —* era ft— •5^^ CD  cr  vO  CD  O  I— *<  ^ rt CD  C"  p" ^  e rt  01 CP  CS i-b cs t-u ere CT  3 3  TO  o  to O  cn  I  Cf p  Co  cr" CD  TO  £"f  o cr  cr rt cr  3 CD CD  g  5"  g  rt  ^°  CD  Cu  R  a  sc  i-t g . Cu  cv Cu FJ  I  I I CD S3  3 ft  i-t  CD P  CO  CD  3t  H<  2 . cr-  rt c  cr •  3  OQ 2  O  cf. 3  3  rt  3 P  H  P.  cr  CD rt-  3' £  ^  ft  O  v  3  (t  S  rt  R  rt' CD Cu  s erg; oo  £ ^  CD"  2  2  < p  CD  CD  _  <- o •••• rt,^-3 O rt co Cf) o S t h CD 3 3 2 P CO O Cu rt.^ S3 o E? CD  cf 9  o 3 3  CD  CD  &  8 g  <—>• S OQ 3 3 CD  OQ 3 Cu  CD  <  3  co CO 3 » S  Cu  S3 CD  STCD  a c r O  p r t-. ^ 3 •a- ss P  CD  PJ. CD  CS " O ^3 <!  CD  3." 2* o  CD SB P  Cu  ts cr o  P •<:  CD  co CO CO  i  CO 03  CD H  S  p  o o co cr co o CD  33 3" do 13  ?f  CD  CD rtB C O v< a 1 OQ  a  cf^  S  R  Rf  CD ST  t g * 3 B  CD  K.  p^  2^  p  p'  co  *|  P  p T3  3  Cu  cr % CD  ' a cr  53  rt CO a sr  £  •-»• 3  OQ  I  1  g.3  O Q CD rt  s CD  CO  I T . 3. a" o v; O co rt  CD  w  X  3 rt Sf  CD P  a a- rtco co cu o> CD  p R  ;S  O  3  ^ cr  ,  CD  "O C rt  TO  P  Hi  rt  ^< co"  •-»•  1O 3>  cf cr  O  ^ Cu  cr  3  I. %  ft  % >B'3 ^ g  &  r-t3  S , cits OQ  Iff s*r ci  CO p  OQ O O Cu  CD  VO o  rt  pf  H  °> «< a O -  e» <-*>  rrp  o « " F> cr ri " 3 2 a"fi>5 ^ w _ s' d r t• ^ irt03 £5 p  »> -  *  . §  rt o  g-  * M  gI  co CD p  rt ^ ^ 2 2 2 fi B p  "O  ~ p co  las B^l Cr 3  d  P  cs CTQ O  o o o CD CO co CD  8  P CD ^ CO P a P 5. rt-".  «<  CD co O CD  3 CD  s  s  3  _ sr  sa»  1/1  2 -»• 2- <o - o o CD _ 3  P CD  4, 00  2 *  3 B> CD  o P  5 cr  »• 2.  co p » 2-»• oHp P ' S> Pa C•3D CcDoP D cd CD C < 3 C D O P CO  6  8» 3  i  o  P  o o- ^ 3 CrQ  O  i  i-t CD co C D co _- §f  p  Cu  t.  - c  CO  3 <  O  era p O H i-t —. 3 e8 re 5* P rt ^ 3 CD  8  3 CD* CD •->• o 3 CO p D 5; |-t CD C CD X P P- pco O- 3 CO 3 P ^ .ft < G O-»•CD  <**-  O-  3  be*  rt  P-  S  C D co  o p  co ^ co cr a rt-. CD 45 S co CD O p CD  I* p  5  Cu g rt- rt-  p* o  cr  ^  -  CD ^  ¥^ c  _ 2.  CD . CD  5" rt  ® If  5  a g  -  ^  cr cr 2 Cu  O co  *& CD CD co CD  .  •  2 2 cr 3 § °cr < cs. sf p CD Q t3 a* 3 < _. P CO CO o5  B ^g cr  i-t C D P ^ g* co i-f M- B^-B* CD D 5 C  a. p B  *< O  o c  ^ 3.  .  B* g:  C 5Dcre P . CD  CD - cr ^ P . a. a. .3 CD 3 » ft O P ^ CD p 3 ^ tre £ •p FT <  < §• " 5 o* i-t  S* ^  a*  "*. .<g « 3 0  " .  a«_  _5  O  -o  3 3 P s  rt-  3 v< P 3/po o 3 _ cr cr < CD CD CD p. <-+ ^ Cu a O  O  S-6 o g cr CD p  OQ  2 o P  o R* 3 CD P cr _5 o P cr 2 o  o p  I  ft cr v: p p o o 2 < c S. o- 2 CD l-t  cr  .  to y p fjs co' C3^ CO C C C CD • ff *<*  s*S 8|  s-  s*  l-t 3 ft gP o •-t ^o 2 &w P 5  CO CO  O  P Q  3 to rt  •rt • CO SC BD P- 3 CD g rt CO  g 5 cr D ^ . rt CD C P P  cr p  CD P  o S- ^ ^ o CD P p a* 2 •-»• OQ S ? o g CD p' cr CD P  rt, P  cn cn O o  3 a-  TO a o  cn O  Cu  1  CT 3  c-  ?f  H  <r o  o g* o CT  rt CT cn O 3 cn CT O tr o  1 B- oa  a  cr lr O v; CT  o  a- <: a  3  cr^<  o  3»2 CT . . cn_ o "3 3  3-  a e* cOn o  3 . p 2 ^ cr  Cu ^ cn p  ct 23  o  o  CT  CT 3  a- a  35  cr CT  CT cn  a  os  S p  P  i  2 o - o a*  3  °-& sosa GO  v;  O w  <_?_ •  Et  '  cr  § rt- 3 °  § a  O  a g a P—. a co rt  &  CT CT  o 8 & 3  £ * ~ 3 CT rt CT3 rt P  CfQ  P  co  O 3 O CfQ  13 co rt 3 cr CT CTQ CT CT % CT O 8 3 CO rt rt o 3 O 3 s* rt P Ol fQ 3 cr e. CCT o Cu rt  K- CT  o O  o 3  O  CT 3  w.  o co rt ^ rt cr  ta  cr CT o  •TS  p. a&? O CT <! CO  Io  <  a'cS  3 co TO j T  a^ »J8 g  CT p 3*  CT Si i"*J< co CO CO • O  H3.  rt-  O  CTit* CT §  I  3.  Cu  g*  -  5*  co Cu CT CT CT P 3 rt.3CT O rt- Cu  rt [-j rt05  rt  p  T  o ^ ^o CT « <r y  3 3 3.  P  rt-  a* SS P 8 9 3 CO  a o  Is Cu  P  5f> a  o CT  rt w o  co  cT ^ ffi 3 3 ~  P.  I  I Cl  a aTO (^>-  3  O  3> rt  TO O Ct  P CT  3^  a co  rt rt o  3  3f  a •>3  2. 3  —  I  co  i • o Z.  CT CT to o  3  o  Cu  CT  fTO  a  '  Cu cu  o  g  o  rt rt 3 CT g CT O CT rt co TO CT CT CO  O  P CO  uT  H S- S ai ?* R 2. a.  cr CP u  CT  C u ^  CT FT  at a CT 2 co 55.^ o rt a rt  o  O  rt  p  CT  CT rt-  P  rt O  aCT  8 CT  3 HJ  rt CT  *a S. a,Et 2a " *a rt  CO  VO  I  if  s*  I TO  3 a a TO'  TO O 3  aTO .3  I  vo  I o a  H  >3  9  I  I&  R  a  TO  5 TO  2  S c a  co co  154  References American Academy of Pediatrics, Working Group in Breastfeeding. (1997). Breastfeeding and the use of human milk. Pediatrics, 100(6), 1035-1039. Amir, L., & Donath, S. (2002). Does maternal smoking have a negative physiological effect on breastfeeding? The epidemiological evidence. Birth, 29(2), 112-123. Arlotti, J., Cottrell, B., Lee, S., & Curtin, J. (1998). Breastfeeding among low-income women with and without peer support. Journal of Community Health Nursing, 15, 163-178. Armitage, P., & Colton, T. (Eds.). (1998). Encyclopedia ofBiostatistics (Volumes 16). Chichester: John Wiley & Sons. Armstrong, H. (1991). International recommendations for consistent breastfeeding definitions. Journal ofHuman Lactation, 7(2), 51-54. Arora, S., McJunkin, C , Wehref, J., & Kuhn, P. (2000). Major factors influencing breastfeeding rates: Mother's perception of father's attitude and milk supply. Pediatrics, 106(5), E67. Auerbach, K. (2000). Evidenced-based care and the breastfeeding couple. Journal of Midwifery & Women's Health, 45(3), 205-211. Auerbach, K. (1988). Supplementation and breastfeeding duration. Journal ofHuman Lactation, 4(4), 155-156. Auerbach, K. (1985). The influence of lactation consultant contact on breastfeeding duration in a low-income population. Nebraska Medical Journal, 70(9), 341-346. Avery, M.D., Duckett, L., Dodgson, J., Savik, K., & Henly, S. (1998). Factors associated with very early weaning among primiparas intending to breastfeed: Maternal and Child Health Journal, 2(3), 167-179. Bandura, A. (1977). Self-efficacy: Toward a unifying theory of behavioral change. Psychological Review, 84, 191-215. Barros, F., Victora, C , Semer, T., Tonioli Filho, S., Tomasi, E., & Weiderpass, E. (1995). Use of pacifiers is associated with decreased breastfeeding duration. Pediatrics, 95, 497-499. Bar-Yam, N., & Darby, L. (1997). Fathers and breastfeeding: a review of the literature. Journal of Human Lactation, 13(\), 45-50.  155  Barber, C , Abernathy, T., Steinmetz, B., & Gharlebois, J. (1997). Using a breastfeeding prevalence survey to identify a population for targeted programs. Canadian Journal of Public Health, 88(4), 242-245. Basire, K., Pullon, S., & Mcleod, D. (1997). Baby feeding: the thoughts behind the statistics. New Zealand Medical Journal, 110, 184-187. Belza, B. (1990). Multidimensional Assessment ofFatigue (MAF) Scale Users Guide. Unpublished manuscript. Belza, B. (1995). Comparison of self-reported fatigue in rheumatoid arthritis and controls. Journal ofRheumatology, 22, 639-643. Belza, B., Curtis, J.H., Yelin, E.H., Epstein, W.V., & Gilliss, C L . (1993). Correlates of fatigue in older adults with rheumatoid arthritis. Nursing Research, 42(2), 93-99. Beck, C. T. (1998). A checklist to identify women at risk for developing postpartum depression. Nursing Research, 45, 297-303. Beck, C. T. (1995). The effects of postpartum depression on maternal-infant interaction: A metaanalysis. Nursing Research, 4, 298-304. Beck, C T . (2001). Predictors of postpartum depression. Nursing Research, 50(5), 275-285. Bernard-Bonnin, A., Stachtchenko, S., Girard, G., & Rousseau, E. (1989). Hospital practices and breastfeeding duration: A meta-analysis of controlled trials. Birth, 16(2), 64-66. Bick, D. (1999). The benefits of breastfeeding for the infant. British Journal ofMidwifery, 7(5) 312-319. Bick, D., MacArthur, C , & Lancashire, R.J. (1998). What influences the uptake and early cessation of breastfeeding? Midwifery, 14,242-247. Binns, C , & Scott, J. (2002). Breastfeeding: Reasons for starting, reasons for stopping and problems along the way. Breastfeeding Review, 10(2), 13-19. Blyth, R., Creedy, D., Dennis, C. 1., Moyle, W., Pratt, J., & De Vries, S. (2002). Effect of maternal confidence on breastfeeding duration: An application of breastfeeding selfefficacy theory. Birth, 29(4), 278-284. Bocar, D., & Moore, K. (1987). Acquiring the parental role: A theoretical perspective. (Unit 16, Lactation Consultant Series). Garden City Park, NY: Avery Publishing Group, 2-11.  156  Boettcher, J.P., Chezem, J.C., Roepke, J., & Adams, T.A. (1999). Interaction of factors related to lactation duration. The Journal ofPerinatal Education, 8(2), 11-19. Bottorff, J. (1990). Persistence in breastfeeding: a phenomenological investigation. Journal of Advanced Nursing, 15, 201-209. Bozoky, I., & Corwin, E. (2002). Fatigue as a predictor of postpartum depression. Journal of Obstetric, Gynecologic, and Neonatal Nursing, 31(4), 436-443. Bray, J., & Maxwell, S. (1985). Multivariate Analysis of Variance. Series: Quantitative Applications in the Social Sciences. Sage: University Paper. Bullock, C , & Pridham, K. (1988). Sources of maternal confidence and uncertainty and perceptions of problem-solving competence. Journal ofAdvanced Nursing, 13,321 -329. Buxton, K., Gielen, A., Faden, R., Brown, C , Paige, D., & Chwalow, A. (1991). Women intending to breastfeed: Predictors of early infant feeding experience. American Journal of Preventative Medicine, 7(2), 101-107. Caltabiano, N., & Caltabiano, M. (1996). Relationship between exhaustion and postnatal depression. Psychological Reports, 79, 225-226. Campbell, S., Cohn, J., Flanagan, C , Popper, S., & Meyers, T . (1992). Course and correlates of postpartum depression during the transition to parenthood. Development and Psychopathology, 4, 29-47. Canadian Institute for Health Information. (2004). More than 1 in 5 Canadian babies now delivered by C-Section. Retrieved April 24, 2004 from http://cihi.ca/cihiweb/dispPage. Canadian Pediatric Society, Dieticians of Canada, & Health Canada. (1998). Nutrition for Healthy Term Infants. Ottawa: Minister of Public Works and Government Services. Carty, E., & Bradley, C F . (1990). A randomized, controlled evaluation of early postpartum hospital discharge. Birth, 17(4), 199-204. Carty, E., Bradley, C , & Winslow, W. (1996). Women's perceptions of fatigue during pregnancy and postpartum. Clinical Nursing Research, 5(1), 67-80. Chalmers, I. (2000). Current controlled trials: an opportunity to help improve the quality of clinical research. Cardiovascular Medicine, 1(1), Retrieved December 8, 2003, from http://cvm.controlled-trials.com  157  Chan-Yip, A., & Kramer, M. (1983). Promotion of breast-feeding in a Chinese community in Montreal. Canadian Medical Association Journal, 129, November, 955-958. Chapman, D., & Perez-Escamilla, P. (1999). Does delayed perception of the onset of lactation shorten breastfeeding duration? Journal ofHuman Lactation, 15(2), 107-111. Chezem, J., Friesen, C , & Boettcher, J. (2003). Breastfeeding knowledge, breastfeeding confidence and infant feeding plans: Effects on actual feeding practices. Journal of Obstetric, Gynecologic, and Neonatal Nursing, 32(1), 40-47. Chezem, J., Friesen, C , & Clark, H. (2001). Sources of infant feeding information used by pregnant women. Journal ofPerinatal Education, 10(3), 20-26. Clifford, C , Day, A., Cox, J., & Werrett, J. (1999). A cross-cultural analysis of the use of the Edinburgh Post-natal Depression Scale (EPDS) in health visiting practice. Journal of Advanced Nursing, 30(3), 655-664. Cohen, J. (1977). Statistical Power Analysis for the Behavioral Sciences. New York: Academic Press. Cooper, P., Murray, L., & Stein, A. (1993). Psychosocial factors associated with the early termination of breast-feeding. Journal ofPsychosomatic Research, 37(2), 171-176. Cooke, M., Sheehan, A., & Schmied, V. (2003). A description of the relationship between breastfeeding experiences, breastfeeding satisfaction, and weaning in the first 3 months afterbirth. Journal ofHuman Lactation, 19(2), 145-156. Coreil, J., Bryant, C.A., Westover, B., & Bailey, D. (1995). Health professionals and breastfeeding counseling: client and provider views. Journal ofHuman Lactation, 11(4), 265-271. Coreil, J., & Murphy, J. (1988). Maternal commitment, lactation practices and breastfeeding duration. Journal of Obstetric, Gynecologic, and Neonatal Nursing, 17(4), 273-278. Cox, J., & Holden, J. (Eds.). (1994). Perinatal Psychiatry. Use and misuse of the Edinburgh Postnatal Depression Scale. The Royal College of Psychiatrists. Gaskell: The American Psychiatric Press, Inc. (pp. 1-275). Cox, J., Holden, J., & Sagovsky, R. (1987). Detection of postnatal depression: Development of the 10 item Edinburgh postnatal depression scale. British Journal of Psychiatry, 150, 782-786.  158  Creedy, D., Dennis, C-L., Blyth, R., Moyle, W., Pratt, J., & Vries, S. (2003). Psychometric characteristics of the breastfeeding self-efficacy scale: Data from an Australian sample. Research in Nursing & Health, 26, 143-152.  Curro, V., Lanni, R., Scipione, F., Grimaldi, V., & Mastroiacovo, P. (1997). Randomized controlled trial assessing the effectiveness of a booklet on the duration of breastfeeding. Archives ofDisease in Childhood, 76, 500-504.  Curry, M. (1988). Variables related to adaptation to motherhood in "normal" primiparous women. Journal of Obstetric and Gynecologic Nursing, March-April, 115-121.  Dennis, C-L. (1999). Theoretical underpinnings of breastfeeding confidence: A self-efficacy framework. Journal of Human Lactation, 15(3), 195-201.  Dennis, C-L. (2002a) Breastfeeding initiation and duration: A 1990-2000 Literature Review. Journal of Obstetric, Gynecologic, and Neonatal Nursing, 3/(10), 2-32.  Dennis, C-L. (2002b). Breastfeeding peer support: Maternal and volunteer perceptions from a randomized controlled trial. Birth, 29(3), 169-176. Dennis, C-L. (2004). Can we identify mothers at risk for postpartum depression in the immediate postpartum period using the EPDS? Journal ofAffective Disorders, 78(2), 163-169. Dennis, C-L. (2003). The breastfeeding self-efficacy scale: Psychometric assessment of the short form. Journal of Obstetric, Gynecologic, and Neonatal Nursing, 32(6), 734-744.  Dennis, C-L, & Faux, S. (1999). Development and psychometric testing of the breastfeeding self-efficacy scale. Research in Nursing and Health, 22, 399-409. Dennis, C-L., Hodnett, E., Gallop, R., & Chalmers, B. (2002). The effect of peer support on breast-feeding duration among primiparous women: a randomized controlled trial. Canadian Medical Asssociation Journal, 166(1), 21-28.  Dodgson, J., Tarrant, M., Fong, D., Peng, X., & Choi Hui, W. (2003). Breastfeeding patterns of primiparous mothers in Hong Kong. Birth, 30(3), 195-202. Donath, S., & Amir, L. (2000). Does maternal obesity adversely affect breastfeeding initiation and duration? Breastfeeding Review, 8(3), 29-33. Donath, S., Amir, L., & The ALSPAC Study Team. (2003). Relationship between prenatal infant feeding intention and initiation and duration of breastfeeding: a cohort study. Acta de odontologia pediatrica, 92, 352-356.  159  Duckett, L., Henly, S., & Garvis, M. (1993). Predicting breastfeeding duration during the postpartum hospitalization. Western Journal of Nursing Research, 15(2), 177-193. Duffy, E., Percival, P., & Kershaw, E. (1997). Positive effects of an antenatal group teaching session on postnatal nipple pain, nipple trauma and breastfeeding rates. Midwifery, 189195. Dungy, C., Losch, M., & Russell, D. (1994a). Hospital infant formula packs: Do they affect breastfeeding duration? Archives of Pediatrics and Adolescent Medicine, 151(1), 124729. Dungy, C , Losch, M., & Russell, D. (1994b). Maternal attitudes as predictors of infant feeding decisions. Journal of the Association for Academic Minority Physicians, 5(4), 159-164. Dykes, F., & Griffiths, H. (1998). Societal influences upon initiation and continuation of breastfeeding. British Journal ofMidwifery, 6(2), 76-80. Dykes, F., & Williams, C. (1999). Falling by the wayside: A phenomenological exploration of perceived breast-milk inadequacy in lactating women. Midwifery, 15, 232-246. Edwards, N. & Sims-Jones, N. (1998). Smoking and smoking relapse during pregnancy and postpartum: Results of a qualitative study. Birth, 25(2), 94-100. Elek, S., Hudson, D., & Fleck, M. (2002). Couples' experiences with fatigue during the transition to parenthood. Journal of Family Nursing, 8(3), 221 -240. Ellis, D., Hewat, R., & Livingstone, V. (1991). Report of an evaluation of the efficacy of a breastfeeding clinic in prolonging the duration of breastfeeding. Vancouver, BC: University of British Columbia, School of Nursing and Department of Family Practice. Ertem, I., Votto, N., & Leventhal, J. (2001). The timing and predictors of the early termination of breastfeeding. Pediatrics, 107(3), 543-548. Fairbank, L., O'Meara, S., Renfrew, M., Woolridge, M., Sowden, A., & Lister-Sharp, D. (2000). A systematic review to evaluate the effectiveness of interventions to promote the initiation of breastfeeding. Health Technology Assessment, 4(25), 1-171. Farber, G. (1997). Breastfeeding education/satisfaction - A literature review. International Journal of Childbirth Education, 12(1), 12-13.  160  Field, P., & Renfrew, M. (1991). Teaching and support: nursing input in the postpartum period. InternationalJournal of Nursing Studies, 28(2), 131-144. Flagler Virden, S. (1988a). Maternal role competence. Western Journal of Nursing Research, 10(3), 274-290. Flagler Virden, S. (1988b). The relationship between infant feeding method and the maternal role adjustment. Journal of Nurse-Midwifery, 33(1), 31 -35. Fleming, A., Ruble, D., Flett, G., & Shaul, D. (1988). Postpartum adjustment in first-time mothers: Relations between mood, maternal attitudes and mother-infant interactions. Developmental Psychology, 24,11-81. Fowles, E. (1998). The relationship between maternal role attainment and postpartum depression. Health Care for Women International, 19(\), 83-94. Freed, G., Clark, S., Harris, B., & Lowdermilk, D. (1996). Methods and outcomes of breastfeeding instruction for nursing students. Journal of Human Lactation, 12(2), 105110. Freed, G., & Fraley, J. (1993). Effect of expectant mothers' feeding plan on prediction of fathers' attitudes regarding breast-feeding. American Journal of Perinatology, 10(4), 300303. Freed, G., Fraley, K., & Schanler, R. (1992). Attitudes of expectant fathers regarding breastfeeding. Pediatrics, 90(2), 224-227. Galler, J., Harrison, R., Biggs, M., Ramsey, F., & Forde, V. (1999). Maternal moods predict breastfeeding in Barbados. Developmental and Behavioral Pediatrics, 20(2), 80-87. Gamble, D., & Morse, J. (1992). Fathers of breastfed infants: Postponing and types of involvement. Journal of Obstetric, Gynecologic, and Neonatal Nursing, 22(4), 358-365. Gardner, D. (1991). Fatigue in postpartum women. Applied Nursing Research, 4(2), 57-62. Gibaud-Wallston, J., & Wandersman, L. (1978). Development and utility of the parenting sense of competence scale. Paper presented at American Psychological Association Meeting, Toronto. Goldberg, D., Cooper, B., Eastwood, M., Kedward, H., & Shephard, M. (1970). A standardized psychiatric interview for use in community surveys. British Journal of Preventive and Social Medicine, 24, 18-23.  161  Green, C.P. (1999). Improving breastfeeding behaviors: Evidencefromtwo decades of  intervention research. Washington D.C.: The Linkages Project. Guise, J-M., Palda, V., Westhoff, C., Chan, B., Helfand, M., & Lieu, T. (2003). The effectiveness of primary care-based interventions to promote breastfeeding: Systematic evidence review and meta-analysis for the US Preventive Services Task Force. Annals of Family Medicine, 7(70-78).  Halpern, S., Levine, T., Wilson, D., MacDonell, J., Katsiris, S., & Leighton, B. (1999). Effect of labor analgesia on breastfeeding success. Birth, 26(2), 83-88. Hamelin, K. (2000). A demonstration project of additional lactation support in the community.  Unpublished manuscript. Hauck, Y., & Dimmock, J. (1994). Evaluation of an information booklet on breastfeeding duration: a clinical trial. Journal ofAdvanced Nursing, 20, 836-843. Hauck, Y., Langton, D., & Coyle, K; (2002). The path of determination: Exploring the lived experience of breastfeeding difficulties. Breastfeeding Review, 10(2), 5-12. Hauck, Y., & Reinbold, J. (1996). Criteria for successful breastfeeding: Mothers' perceptions. Australian College ofMidwives Incorporated, March, 21-27.  Health Canada. (2003). Canadian Perinatal Health Report, Minister of Public Works and Government Services, Ottawa, Canada Health Canada. (2000). Family-Centred Maternity and Newborn Care: National Guidelines,  Minister of Public Works and Government Services, Ottawa, Canada. Heck, K., Schoendorf, K., Chavez, G., & Braveman, P. (2003). Does postpartum length of stay affect breastfeeding duration? A population-based study. Birth, 30,(3), 153-160. Heinig, M., & Dewey, K. (1996). Health effects of breastfeeding for infants: a critical review. Nutrition Research Reviews, 9, 89-110.  Heinig, M., & Dewey, K. (1997). Health effects of breastfeeding for mothers: a critical review. Nutrition Research Reviews, 10, 35-56.  Hellin, K., & Waller, G. (1994). Mothers' moods and infant feeding. Predictions of problems and practices. Journal of Reproductive Infant Psychology, 10, 39-51.  162  Henderson, J., Evans, S., Straton, J. Priest, S., & Hagan, R. (2003). Impact of postnatal depression on breastfeeding duration. Birth, 30(3), 175-180. Hewat, R., & Ellis, D. (1984). Breastfeeding as a maternal-child team effort: Women's perceptions. Health Care for Women International, 5, 437-452. Hill, P. (1987). Effects of education on breastfeeding success. Maternal Child Nursing Journal. 16(2): 145-56. Hill, P. (1991). The enigma of insufficient milk supply. Maternal Child Nursing Journal, 16 313-317. Hill, P., & Aldag, J. (1996). Smoking and breastfeeding status. Research in Nursing and Health, 19, 125-132. Hill, P., & Humenick, S. (1996). Development of the H & H Lactation Scale. Nursing Research, 45(3), 136-140. Hill, P., Humenick, S., Argubright, T., & Aldag, J. (1997). Effects of parity and weaning practices on breastfeeding duration. Public Health Nursing, 14, (A), 227-234. Hill, P., Humenick, S., Brennan, M., & Woolley, D. (1997). Does early supplementation affect long-term breastfeeding? Clinical Pediatrics, 36(6), 345-350. Hill, P., Humenick, S., & West, B. (1994). Concerns of breastfeeding mothers: The first six weeks postpartum. The Journal ofPerinatal Education, 3(4), 47-58. Hoddinott, P., & Pill, R. (1999). Nobody actually tells you: a study of infant feeding. British Journal ofMidwifery, 7(9) 558-565. Hollis, S., & Campbell, F. (1999). What is meant by intention to treat analysis? Survey of published randomized controlled trials. British Medical Journal, 19, 670-674. Humenick, S., Hill, P., & Spiegelberg, L. (1998). Breastfeeding and health professional encouragement. Journal ofHuman Lactation, 14(4), 305-310. Humenick, S., Hill, P., & Wilhelm, S. (1997). Postnatal factors encouraging sustained breastfeeding among primiparas and multiparas. The Journal ofPerinatal Education, 6(3), 33-45.  163  Ickovics, J., Kershaw, T., Westdahl, C , Schindler Rising, S., Klima, C , Reynolds, H. et al. (2003). Group prenatal care and preterm birth weight: Results from a matched cohort study at public clinics. Obstetrics and Gynecology, 102(5), 1051-1057. Ines Couto de Oliveira, M., Camacho, L., Tedstone, A. (2001). Extending breastfeeding duration through primary care: A systematic review of prenatal and postnatal interventions. Journal ofHuman Lactation, 17(4), 326-342. Jacobson, S., Jacobson, J., & Frye, K. (1991). Incidence and correlates of breast feeding in socioeconomically disadvantaged women. Pediatrics, 88, 728-736. Jadad, A. (1998). Randomised Controlled Trials: A user's guide. London: BMJ Books. James, D., Jackson, R., & Probart, C. (1999). Factors associated with breastfeeding prevalence and duration. Journal of the American Dieticians Association, 95, 194-196. Jamieson, L. (1990). New developments in breastfeeding initiatives. British Journal of Midwifery, 7(5), 308. Janke, J. (1988). Breastfeeding duration following cesarean and vaginal births. Journal ofNurseMidwifery, 33(4), 159-164. Janke, J. (1993). The incidence, benefits and variables associated with breastfeeding: Implications for practice. Nurse Practitioner, 18(6), 22-28. Janson, S., & Rydberg, B. (1998). Early postpartum discharge and subsequent breastfeeding. Birth, 25(4), 222-225. Jenner, S. (1988). The influence of additional information, advice and support on the success of breast feeding in working class primiparas. Child: care, health, and development, 14, 319-328. Johnson, J., Ratner, P., Bottorff, J., Hall, W., & Dahinten, S. (2000). Preventing smoking relapse in postpartum women. Nursing Research, 49(1), 44-52. Jordan, P., & Wall, V. (1990). Breastfeeding and fathers: Illuminating the darker side. Birth, 77(4), 210-213. Kaplowitz, D., & Olson, C. (1983). The effect of an education program on the decision to breastfeed. Journal ofNutrition Education, 15(2), 61-65.  164  Kelly, M. (1999). Breastfeeding: the advantages for baby and mother. World of Irish Nursing, 7(6), 19-21. Kessler, L., Gielen, A., Deiner West, M., & Paige, D. (1995). The effect of woman's significant other on her breastfeeding decision. Journal of Human Lactation, 11(2), 103-109. Kirgis, C. (1988). Maternal role competence. Western Journal of Nursing Research, 10(3), 289290. Kirkley, D. (2000). Is motherhood good for women? A feminist exploration. Journal of Obstetric, Gynecologic, and Neonatal Nursing, 29(5), 459- 464. Kistin, N., Benton, D., Rao, S., & Sullivan, M. (1990). Breast-feeding rates among black urban low-income women: Effect of prenatal educatioa Pediatrics, 86(5), 741-746. Kluka, S., & LeDrew, M. (2000). Breastfeeding... Preparing for a satisfying start. Unpublished manuscript. Kramer, M., Chalmers, B., Hodnett, E., Sevkovskaya, Z., Dzikovich, I., Shapiro, S., et al. for the PROBIT Study Group. (2001). Promotion of breastfeeding intervention trial (PROBIT): a randomized trial in the Republic of Belarus. The Journal of the American Medical Association, 285(4), 413-420. Kramer, M., & Kakuma, R. (2002). Optimal duration of exclusive breastfeeding (Cochrane Review). In The Cochrane Library, Issue 3. Oxford: Update Software. Kuan, L., Britto, M., Decolongon, J., Schoettker, P., Atherton, H., & Kotagal, U. (1999). Health system factors contributing to breastfeeding success. Pediatrics, 104, 1-7, e28. Kunz, C , Rodriguez-Palmero, M., Koletzko, B., & Jensen, R. (1999). Nutritional and biochemical properties of human milk, Part I: General aspects, proteins and carbohydrates. In C. Wagner & D. Purohit (Eds.), Clinical Aspects of Human Milk and Lactation. Clinics in Perinatology. (26, 2, pp.307-311), Philadephia: W.B. Saunders. Labbok, M. (1999). Health sequelae of breastfeeding for the mother. In C. Wagner & D. Purohit (Eds.), Clinical Aspects of Human Milk and Lactation. Clinics in Perinatology. (26, 2, pp. 491- 504), Philadephia: W.B. Saunders. Labbok, M., & Krasovec, K. (1990). Toward consistency in breastfeeding definitions. Studies in Family Planning, 21(4), 226-230.  165  Launer, L., Forman, M., Hundt, G., Sarov, B., Chang, D., Berendes, H., & Naggan, L. (1992). Maternal recall of infant feeding events is accurate. Journal ofEpidemiology and Community Health, 46, 203-206. Lawson, K., & Tulloch, M. (1995). Breastfeeding duration: prenatal intentions and postnatal practices. Journal ofAdvanced Nursing, 22, 841-849. Lee, K., Hicks, G., & Nino-Murcia, G. (1991). Validity and reliability of a scale to assess fatigue. Psychiatry Research, 36, 291-298. Leff, E., Gagne, M., & Jefferis, S. (1994). Maternal perceptions of successful breastfeeding. Journal ofHuman Lactation, 10(2), 99-104. Leff, E., Jefferis, S., & Gagne, M. (1994). The development of the maternal breastfeeding evaluation scale. Journal of Human Lactation, 10(2), 105-111. Loh, N., Kelleher, S., Long, B., & Loftus, B. (1997). Can we increase breastfeeding rates? Irish Medical Journal, 90, 100-101. Loiselle, C , Semenic, S., Core, B., Lapointe, M , & Gendron, R. (2001). Impressions of breastfeeding information and support among first-time mothers within a multiethnic community. Canadian Journal ofNursing Research, 33(3), 31-46.  Lynch, S., Koch, A., Hislop, G., & Coldman, A. (1986). Evaluating the effect of a breastfeeding consultant on the duration of breastfeeding. Canadian Journal ofPublic Health, 77, 190 195. Maclean, H. (1998). Breastfeeding in Canada: A demographic and experiential perspective. Journal of The Canadian Dietetic Association, 59(1), 15-23.  Martens, P. (2000). A risk indicator for early cessation of breastfeeding: Postpartume measures of satisfaction and reported breastfeeding problems. Current Issues in Clinical Lactation, 15-23. Martens, P., Derksen, S., Mayer, T., & Walld, R. (2002). Being born in Manitoba Canadian Journal ofPublic Health, 93(2), S33-S38. Matthews, M. K. (1991). Mothers' satisfaction with their neonates' breastfeeding behaviors. Journal of Obstetic, Gynecologic, and Neonatal Nursing, 20(\), 49-55. Mclnnes, R., Love, J., & Stone, D. (2000). Evaluation of a community-based intervention to increase breastfeeding prevalence. Journal ofPublic Health Medicine, 22(2), 138-145.  166  McKeever, P., Stevens, B., Miller, K., MacDonell, J., Gibbins, S., Guerriere, et al. (2002). Home versus hospital breastfeeding support for newborns: a randomized, controlled trial. Birth, 29(4), 258-265. McLeod, D., Pullon, S., & Cookson, T. (2002). Factors influencing continuation of breastfeeding in a cohort of women. Journal of Human Lactation, 18(4), 335-342. McNair, D., Lorr, M., & Droppleman, L. (1971). Profile of Mood States. San Diego: Educational and Industrial Testing Service. McNatt, M., & Freston, M. (1992). Social support and lactation outcomes in postpartum women. Journal of Human Lactation,8(2), 73-77. McVeigh, C. (1997). Motherhood experiences from the perspective of first-time mothers. Clinical Nursing Research, 6(4), 335-348. Meek, P., Nail, L., Barsevick, A., Schwartz, A., Stephen, S., Whitmer, K., et al. (2000). Psychometric testing of fatigue instruments for use with cancer patients. Nursing Research, 49(4), 181-190. Mercer, R. T. (1981). A theoretical framework for studying factors that impact on the maternal role. Nursing Research, 30(2), 73-77. Mercer, R.T. (1985). The process of maternal role attainment over the first year. Nursing Research, 34(4), 198-204. Mercer, R. T. (1986). Predictors of maternal role attainment at one year postbirth. Western Journal of Nursing Research, 8(1), 9-32. Mercer, R. T. (1995). Becoming A Mother. Research on maternal identity: from Rubin to the present. New York: Springer. Mercer, R. T., & Ferketich, S.L. (1988). Stress and social support as predictors of anxiety and depression during pregnancy. Advances in Nursing Science, 26-39. Mercer, R. T., & Ferketich, S.L. (1994a). Maternal-infant attachment of experiences and inexperienced mothers during infancy. Nursing Research, 43(6), 344-351. Mercer, R. T., & Ferketich, S. L. (1994b). Predictors of maternal role competence by risk status. Nursing Research, 43(1), 38-43.  167  Mercer, R.T., Ferketich, S.L., & DeJoseph, J. (1993). Predictors of partner relationships during pregnancy and infancy. Research in Nursing and Health, 16, 45-56. Mercer, R. T., Ferketich, S.L., May, K., DeJoseph, J., & Sollid, D. (1988). Further exploration of maternal and paternal fetal attachment. Research in Nursing and Health, 11, 83-95. Mercer, R. T., Nichols, E., & Doyle, G. (1988). Transitions over the life cycle: A comparison of mothers and nonmothers. Nursing Research, 37(3), 144-151. Milligan, R. (1989). Maternal fatigue during the first three months of the postpartum period. Dissertation Abstracts International, 50, 07B.  Milligan, R., Lenz, E., Parks, P., Pugh, L., & Kitzman, H. (1996). Postpartum fatigue: Clarifying a concept. Scholarly Inquiry for Nursing Practice: An International Journal, 10(3), 279 291. Milligan, R., Parks, P., Kitzman, H., & Lenz, E. (1997). Measuring women's fatigue during the postpartum period. Journal ofNursing Measurement, 5(1), 3-16. Milligan, R., Pugh, L., Bronner, Y., Spatz, D. and Brown, L. (2000). Breastfeeding duration among low income women. Journal ofMidwifery & Women's Health, 45(3), 246-252. Misri, S., Sinclair, D., & Kuan, S. (1997). Breastfeeding and postpartum depression: is there a relationship? Canadian Journal ofPsychiatry, 42(10), 1061-1065. Moher, D., Schultz, K., & Altaian, D. (2001). The CONSORT statement: revised recommendations for improving the quality of reports of parallel-group randomized trials. The Lancet, 357, 1191-1194. Mongeon, M., & Allard, R. (1995). Controlled study of a regular telephone support program given by volunteers on the establishment of breastfeeding. Canadian Journal ofPublic Health, 86(2), 124-127. Montgomery, K. (1999). Implementing breastfeeding education in the academic setting. Journal ofHuman Lactation, 15(2), 145-147. Montgomery, K. (2000). Maternal-newborn nursing: 13 challenges that influence excellence in practice. Journal of Perinatal Nursing, 10(1), 31-40. Morrell, C , Spiby, H., Stewart, P., Walters, S., & Morgan, A. (2000). Costs and effectiveness of community postnatal support workers: randomized, controlled trial. British Journal of Medicine, 321, 593-598.  168  Morrow, A., Guerrero, L, Shults, J., Calva, J., Lutter, C., Bravo, et al. (1999). Efficacy of homebased peer counselling to promote exclusive breastfeeding: a randomised controlled trial. The Lancet, 353, 1226-1231. Mozingo, J., Davis, M., Droppleman, P., & Merideth, A. (2000). It wasn't working: a woman's experiences of short term breastfeeding. American Journal of Maternal Child Nursing, 25(3), 120-126. Murray, L., & Carothers, A. (1990). The validation of the Edinburgh Post-natal Depression Scale on a community sample. British Journal of Psychiatry, 157, 288-290. Murray, L., & Cox, J. (1990). Screening for depression during pregnancy with the Edinburgh Depression Scale. Journal of Reproductive and Infant Psychology, 8, 99-107. Newcombe, P., Storer, B., Longnecker, M., Mittendorf, R., Greenberg, E., Clapp, R. et al. (1994). Lactation and a reduced risk of premenopausal breast cancer. New England Journal of Medicine, 330, 81 -87. Nichols, F. (2000). History of the women's health movement in the 20 Century. Journal of Obstetric, Gynecologic, and Neonatal Nursing, 29(1), 56-64. th  Novotny, R., Hla, M., Kieffler, E., Park, C , Mor, J., & Thiele, M. (2000). Breastfeeding duration in a multiethnic population in Hawaii. Birth, 27(2), 91-96. Nunnally, J. (1978). Psychometric Testing. New York: McGraw-Hill. O'Campo, P., Faden, R., Gielen, A., & Wang, M. (1992). Prenatal factors associated with breastfeeding duration: recommendations for prenatal interventions. Birth, 19(4), 195201. O'Hara, M., & Swain, A. (1996). Rates and risk of postpartum depression - a meta-analysis. International Review of Psychiatry, 8, 37-54. Palda, V., Guise, J-M., & Wathen, N. (2004). Interventions to promote breast-feeding: applying the evidence to clinical practice. Canadian Medical Association Journal, 170(6), 976978. Papinczak, T., & Turner, C. (2000). An analysis of personal and social factors influencing initiation and duration of breastfeeding in a large Queensland maternity hospital. Breastfeeding Review, 5, 25-33.  169  Parmar, M., & Machin, D. (1995). Survival analysis: A Practical Approach. Chichester, NY: John Wiley & Sons. Patton, C , Beaman, M., Csar, N., & Lewinski, C. (1996). Nurses' attitudes and behaviors that promote breastfeeding. Journal ofHuman Lactation, 12(2), 111-115. Perez-Escamilla, R., Pollitt, E., Lonnerdal, B., & Dewey, K. (1994). Infant feeding policies in maternity wards and their effect on breast-feeding success: An analytical overview. American Journal ofPublic Health, 84(1), 89-97. Piper, B., Lindsey, A., Dodd, M. (1989). The development of an instrument to measure the subjective dimension of fatigue. In S.G. Funk, E.M. Tornquist, M.T. Champagne, L.A. Copp, & R.A. Wiese (Eds.), Key aspects of comfort: Measurment ofpain, fatigue, and nausea, (pp. 199-208). New York: Springer. Pocock, S. (1983). Clinical Trials: A Practical Approach. Chichester, NY: John Wiley & Sons. Pocock, S., Clayton, T., & Altaian, D. (2002). Survival plots of time-to-event outcomes in clinical trials: good practice and pitfalls. Lancet, 359, 1686-1689. Porteous, R., Kaufman, K., & Rush, J. (2000). The effect of individual professional support on duration of breastfeeding: A randomized controlled trial. Journal ofHuman Lactation, 16(4), 303-308. Pugh, L., & Milligan, R. (1998). Nursing intervention to increase the duration of breastfeeding. Applied Nursing Research, 11(4), 190-194. Pugh, L., & Milligan, R. (1993). A framework for the study of childbearing fatigue. Advances in Nursing Science, 15(4), 60-70. Pugh, L., Milligan, R., & Brown, L. (2001). The breastfeeding support team for low-income, predominantly-minority women: a pilot intervention study. Health Care for Women International, 22(5), 501-515. Pugh, L., Milligan, R., Frick, K., Spatz, D., & Bronner, Y. (2002). Breastfeeding duration, costs and benefits of a support program for low income breastfeeding women. Birth, 29, 95100. Pugh, L., Milligan, R., Parks, P., Lenz, E., & Kitzman, H. (1999). Clinical approaches in the assessment of childbearing fatigue. Journal of Obstetric, Gynecologic, and Neonatal Nursing, 28(1), 74-80.  170  Pugin, E., Valdes, V., Labbok, M., Perez, A., & Aravena, R. (1996). Does prenatal breastfeeding skills group education increase the effectiveness of a comprehensive breastfeeding promotion program? Journal ofHuman Lactation, 72(1), 15-19. Raj, V., & Plichta, S. (1998). The role of social support in breastfeeding promotion: A literature review. Journal ofHuman Lactation, 7^(1), 41-45. Ratner, P., Johnson, J., & Bottorff, J. (1999). Smoking relapse and early weaning among postpartum women: Is there an association? Birth, 26, 76-82. Redman, S., Watkins, J., Evans, L., & Lloyd, D. (1995). Evaluation of an Australian intervention to encourage breastfeeding in primiparous women. Health Promotion International, 10 (2), 101-113. Reifsnider, E., & Eckhart, D. (1997). Prenatal breastfeeding education: Its effect on breastfeeding among WIC participants. Journal ofHuman Lactation, 13(2), 121-125. Renfrew, M., & Lang, S. (1994). Early versus delayed initiation of breastfeeding. (Cochrane Review). In: The Cochrane Library, Issue 3, Oxford: Update Software. Riordan, J., Gross, A., Angeron, J., Krumwiede, B., & Melin, J. (2000). The effect of labor pain relief medication on neonatal suckling and breastfeeding duration. Journal ofHuman Lactation, 16(1), 7-12. Riordan, J., Woodley, G., & Heaton, K. (1994). Testing validity and reliability of an instrument which measures maternal evaluation of breastfeeding. Journal ofHuman Lactation, 10(4), 231-235. Rodriguez-Palmero, M., Koletzko, B., Kunz, C , & Jensen, R. (1999). Nutritional and biochemical properties of human milk, Part II: Lipids, micronutrients, and bioactive factors. In C. Wagner & D. Purohit (Eds.), Clinical Aspects ofHuman Milk and Lactation. Clinics in Perinatology. (26, 2, pp.335-360), Philadephia: W.B. Saunders. Rogan, W., & Gladen, B. (1993). Breastfeeding and cognitive development. Early Human Development, 31, 181-193. Rossiter, J. (1994). The effect of a culture-specific education program to promote breastfeeding among Vietnamese women in Sydney. International Journal ofNursing Studies, 31(4), 369-379. Rowe-Murray, H., & Fisher, J. (2002) Baby Friendly Hospital Practices: Cesarean Section is a persistent barrier to early initiation of breastfeeding. Birth, 29(2), 124-131.  171  Rubin, R. (1967a). Attainment of the maternal role. Part I. Processes. Nursing Research, 16(3), 237-245. Rubin, R. (1967b). Attainment of the maternal role. Part II. Models and referrants. Nursing Research, 16(4), 342-346. Ruchala, P., & Halstead, L. (1994). The postpartum experience of low-risk women: A time of adjustment and change. Maternal-Child Nursing Journal, 22($),%3-%9. Saadeh, R., & Akre, J. (1996). Ten steps to successful breastfeeding: A summary of the rationale and scientific evidence. Birth, 23(3), 154-160. Schafer, E., Vogel, M., Viegas, S., & Hausafus, C. (1998). Volunteer peer counselors increase breastfeeding duration among low-income women. Journal of Human Lactation, 25, 101-106. Schmied, V., & Barclay, L. (1999). Connection and pleasure, disruption and distress: women's experience of breastfeeding. Journal of Human Lactation, 15(4), 325-333. Sciacca, J., Phipps, B., Dube, D., & Ratliff, M. (1995). Influences on breast-feeding by lowerincome women: An incentive based, partner-supported educational program. Journal of Dietetic Association, 95(3), 323-328. Scott, J., & Binns, C. (1999). Factors associated with the initiation and duration of breastfeeding: a review of the literature. Breastfeeding Review, 7(1), 5-16. Sears, W. (1992). The father's role in breastfeeding. NAACOG's Clinical Issues in Perinatal and Women's Health Nursing, 3(4), 713-716. Seguin, L., Potvin, L., St. Denis, M., & Loiselle, J. (1999). Depressive symptoms in the late postpartum among low socioeconomic status women. Birth, 26(3), 157-163. Serwint, J., Wilson, M., Vogelhut, J., Repke, J., & Seidel, H. (1996). A randomized controlled trial of prenatal pediatric visits for urban, low-income families. Pediatrics, 98(6), 10691075. Sheehan, A. (1999). A comparison of two methods of antenatal breast-feeding education. Midwifery, 15, 274-282. Sheeshka, J., Potter, B., Norrie, E., Valaitis, R., Adams, G., & Kuczynski, L. (2001). Women's experiences breastfeeding in public places. Journal of Human Lactation, 17(1), 31-38.  172  Shoham-Yakubovich, I., Pliskin, J., & Carr, D. (1990). Infant feeding practices: An evaluation of the impact of a health education course. American Journal of Public Health, 80(6), 732734. Sikorski, J., & Renfrew, M. (1998). Support for breastfeeding mothers. (Cochrane Review). In: The Cochrane Library, Issue 4, Oxford: Update Software. Sikorski, J, Renfrew, M., Pindoria, S., & Wade, A. (2004). Support for breastfeeding mothers (Cochrane Review). In: The Cochrane Library, Issue 1, Chichester, UK: John Wiley & Sons, Ltd. Singh, D. (2000). Postnatal care needs not being met. British Journal ofMidwifery, 5(8), 473474. Smith, M. (1989). Postnatal concerns of mothers: An update. Midwifery, 3, 182-188. Spitzer, R., Endicott, J., & Robins, E. (1975). Research Diagnostic Criteria. Instrument No. 58. New York: New York Psychiatric Institute. SPSS. Statistical Program for Social Scientists. (Pro Stats, Version 11), Chicago, IL: SPCC, Inc. Stainton, C , Murphy, B., Higgins, P., Neff, J., Nyberg, K., & Ritchie, J. (1999). The needs of postbirth parents: An international, multisite study. The Journal of Perinatal Education, 8(3), 21-29. Stamp, G. (1998). A primary healthcare team had low sensitivity for detecting postnatal depression [commentary on Hearn G., Iliff A„ Jones I., et al. Postnatal depression in the community. British Journal of General Practice 48, 1064-6], Evidence-Based-Nursing, 1(4), 121. Statistics Canada. (1994/1995). National Population Health Survey Supplement. Ottawa Statistics Canada. (1996). National Longitudinal Survey on Children and Youth. Ottawa Susin, L., Giugliani, E., Kummer, S., Maciel, M., Simon, C , & da Silveira, L. (1999). Does parental breastfeeding knowledge increase breastfeeding rates? Birth, 26(3), 149-156. Susman,V., & Katz, J. (1988). Weaning and depression: another postpartum complication. American Journal of Psychiatry, 145, 498-501.  173  Taddei, J., Westphal, M., Venancio, S., Bogus, C , & Souza, S. (2000). Breastfeeding training for health professionals and resultant changes in breastfeeding duration. Sao Paulo Medical Journal,118(6), 185-191. Taj, R., & Sikander, K. (2003). Effects of maternal depression on breast-feeding. Journal of Pakistan Medical Association, 53(1), 8-11.  Tammentie, T., Tarkka, M.J., Astedt-Kurki, P., & Paavilaninen, E. (2002). Sociodemographic factors of families related to postnatal depressive symptoms of mothers. International Journal ofNursing Practice, 8, 240-246.  Tamminen, T. (1988). The impact of mother's depression on her nursing experiences and attitudes during breastfeeding. Acta Paediatrica Scandinavica, 344(77), 87-94. Tarkka, M., Paunonen, M., & Laippala, P. (1999). Factors related to successful breast feeding by first-time mothers when the child is 3 months old. Journal of Advanced Nursing, 29(1), 113-118. The Breastfeeding Committee for Canada. (2003). Guidelines for the WHO/UNICEF BabyFriendly Initiative (BFI) in Canada. The ten steps and practice indicators for Baby-  Friendly Hospitals. Retrieved May 17, 2004 from www.breastfeedingCanada.ca Thornton, R., & Nardi, P. (1975). Dynamics of role acquisition. American Journal of Sociology, 80, 870-885. Tiedje, L, & Darling-Fisher, C. (2003). Promoting father-friendly healthcare. MCN The American Journal ofMaternal/Child Nursing, 28(6), 350-357.  Tulman, L., & Fawcett, J. (1988). Return of functional ability after childbirth. Nursing Research, 57,77-81. UNICEF UK Baby Friendly Initiative. Health benefits of breastfeeding. Retrieved March 4, 2003 from http// www.babyfriendly.org.uk/health.asp. Vari, P., Camburn, J., & Henly, S. (2000). Professionally mediated peer support and early breastfeeding success. The Journal of Perinatal Education, 9(1), 22-30. Victora, C , Tomasi, E., Olinto, M, & Barros, F. (1993). Use of pacifiers and breastfeeding duration. Breastfeeding Review, 2(8), 382-383.  174  Walker, M.L. (2002). Telephone based peer support increased duration of breastfeeding in Primiparous mothers [commentary on Dennis, C-L., Hodnett, E., & Gallop, R., et al., (2002). The effect of peer support on breast-feeding duration among primiparous women: a randomized controlled trial. Canadian Medical Association Journal, 166, 21-28], Evidence-Based-Nursing, 5(3), 75.  Wambach, K. (1998). Maternal fatigue in breastfeeding primiparae during the first nine weeks postpartum. Journal ofHuman Lactation, 14(3), 219-229.  Wiles, L. (1984). The effect of prenatal breastfeeding education on breastfeeding success and maternal perception of the infant. Journal of Obstetric and Gynecologic Nursing, JulyAugust, 253-257. Williams, N. (1997). Maternal psychological issues in the experience of breastfeeding. Journal ofHuman Lactation, 13(1), 57-60.  World Health Organization. (1998). Evidence for the Ten Steps to Successful Breastfeeding. Geneva: Department of Child Health and Development. Retrieved September 15, 2003  from http//www.who.org. World Health Organization (1990). Innocenti declaration on the protection, promotion and support of breastfeeding. In Breastfeeding in the 1990's: A global initiative meeting in  Florence, Italy, and New York. Geneva, Switzerland: Author. World Health Organization and Wellstart International. (1999). The Baby-friendly Hospital Initiative. Monitoring and reassessment: Tools to sustain progress. Geneva, World  Health Organization, 1999 (Document WHO/NHD/99.2). Retrieved April 6, 2004 from http://whqlibdoc.who.int/hq/1999AVHO_NHD_99.2_(sectionI-II).pdf Wright, A., Rice, S., & Wells, S. (1996). Hospital practices to increase the duration of breastfeeding. Pediatrics, 97(5), 669-675. Young, M., Bresnitz, E., & Strom, B. (1983). Sample size nomograms for interpreting negative clinical studies. Annals ofInternal Medicine, 99, 248-251. Zahr, L. (1991). The relationship between maternal confidence and mother-infant behaviors in premature infants. Research in Nursing & Health, 14, 279-286. Zahr, L. (1993). The confidence of Latina mothers in the care of their low birth weight infants. Research in Nursing and Health, 16, 335.-342.  Zinn, B. (2000). Supporting the employed breastfeeding mother. Journal ofMidwifery & Women's Health, 45(3), 216-226.  176  Bibliography Aaronson, L.S., Teel, C.S., Cassmeyer, T., Neuberger, G., Pallikkathayil, L., Pierce, J., et al. (1999). Defining and measuring fatigue. Image: Journal ofNursing Scholarship, 31(1), 45-49. Adams, D., & Hewell, S. (1997). Maternal and professional assessment of breastfeeding. Journal ofHuman Lactation, 13(4), 279-283. Albernaz, E., Giugliani, E., & Victora, C. (1998). Supporting breastfeeding: a successful experience. Journal ofHuman Lactation, 14(4), 283-285. Avery, M.D., Duckett, L., & Frantich, C.R. (2000). The experience of sexuality during breastfeeding among primiparous women. Journal ofMidwifery & Women's Health, 45(3), 227-237. Ayers, J.F. (2000). The use of alternative therapies in the support of breastfeeding. Journal of Human Lactation, 16(1), 52-56. Bliss, M., Wilkie, J., Acredolo, C , Berman, S., & Tebb, K. (1997). The effect of discharge pack formula and breast pumps on breastfeeding duration and choice of infant feeding method. Birth, 24(2), 90-101. Bloom, K., Goldbloom, R., Robinson, S., & Stevens, F. (1982). Factors affecting the continuance of breastfeeding. Acta Paediatr Scand, Supplement 300, 9-14. Bourgoin, G., Lahaie, N., Rheaume, B., Berger, M., Dovigi, C., Picard, L., et al. (1997). Factors influencing the duration of breastfeeding in the Sudbury region. Canadian Journal of Public Health, 88(4), 238-241. Brennan, S. (1998). Nursing and motherhood constructions: implications for practice. Nursing Inquiry, 5, 11-17. Bryant, C., Coreil, J., D'Angelo, S., Bailey, D., & Lazarov, M. (1992). A strategy for promoting breastfeeding among economically disadvantaged women and adolescents. NAACOG's Clinical Issues in Perinatal and Women's Health Nursing, 3(4), 723-730. Buchko, B., Pugh, L., Bishop, B., Cochran, J., Smith, L., & Lerew, D. (1993). Comfort measures in breastfeeding, primiparous women. Journal of Obstetric, Gynecologic, and Neonatal Nursing, 23(1), 46-52.  177  Buckner, E., & Matsubara, M. (1993). Support network utilization by breastfeeding mothers. Journal of Human Lactation, 9(4), 231-235. Caldwell, K. (1999). Reaching the goals of "Healthy People 2000" regarding breastfeeding. In C. Wagner & D. Purohit (Eds.), Clinical Aspects of Human Milk and Lactation. Clinics in Perinatology. (26, 2, pp. 527-537), Philadephia: W.B. Saunders. Caplan, L. (1999). When baby won't breastfeed: Contemporary impediments to successful lactation. Journal of Newborn Nursing, Spring, 12-21. Carceller, A., Rousseau, E., Chad, Z., & Bernard-Bonnin, A. (1995). Infant feeding practices: Are CPS guidelines followed? Canadian Journal of Public Health, 86(5), 301-304. Caulfield, L., Gross, S.M., Bentley, M.E., Bronner, Y., Kessler, L., Jensen, J., et al. (1998). WIC based interventions to promote breastfeeding among African-American women in Baltimore: Effects on initiation and continuation. Journal of Human Lactation, 14(\), 15-22. Centuori, S., Burmaz, T., Ronfani, L., Fragiacomo, M., Quintero, S., & Pavan, C. et al. (1999). Nipple care, sore nipples, and breastfeeding: a randomized trial. Journal ofHuman Lactation, 15, 125-130. Chao, E. (1998). From the editor... Special issue on infant feeding practices. Journal of the Canadian Dietetic Association, 59(1), 8-9 Chen, C. (1993). Effects of home visits and telephone contacts on breastfeeding compliance in Taiwan. Maternal-Child Nursing Journal, 21(3), 82^90. Chezem, J., & Friesen, C. (1999). Attendance at breast-feeding support meetings: Relationship to demographic characteristics and duration of lactation in women planning postpartum employment. Journal of the American Dietetic Association, 99, 83-85. Chute, G. (1992). Promoting breastfeeding success: An overview of basic management. NAACOG's Clinical Issues in Perinatal and Women's Health Nursing, J(4),570-582. Cronenwett, L., Stukel, T., Kearney, M., Barrett, J., Covington, C , Del Monte, K., et al. (1992). Single daily bottle use in the early weeks postpartum and breastfeeding outcomes. Pediatrics, 90(5), 760-766. Crow, S.M. (1999). Secrets to breastfeeding survival. American Baby, 61(1), 54-60.  178  Driscoll, J. (1992). Breastfeeding success and failure: Implications for nurses. NAACOG's Clinical Issues in Perinatal and Women's Health Nursing, 3(4), 565-569. Duckett, L. (1992). Maternal employment and breastfeeding. NAACOG's Clinical Issues in Perinatal and Women's Health Nursing,3(A),10\-l\2. Evers, S. (1998). Influences on breastfeeding rates in low income communities in Ontario. Canadian Journal of Public Health, 89(3), 203-207. Ewing, G., & Morse, J. (1988). Paradoxical priorities in breastfeeding research: challenges for new directions. The Australian Journal ofAdvanced Nursing, 6(2), 24-28. Fetherston, C. (1995). Factors influencing breastfeeding initiation and duration in a private Western Australian maternity hospital. Breastfeeding Review, 3(1), 9-14. Fodor O'Brien, M., Sterken, E., & Shears, E. (1999). Breastfeeding protection, promotion and support - Forward to the Breastfeeding Statement of the Breastfeeding Committee for Canada. Journal of the Canadian Dietetic Association, 59(1), 10. Ford, R., Mitchell, E., Scragg, R., Steward, A., Taylor, B. & Allen, E. (1994). Factors adversely associated with breast feeding in New Zealand. Journal of Paediatrics and Child Health, 30, 483-489. Gagnon, A., Edgar, L., Kramer, M., Papageorgiou, A., Waghorn, K., & Klein, C. (1997). A randomized trial of a program of early postpartum discharge with nurse visitation. American Journal of Obstetrics and Gynecology, 176,(1), 205-211. Gottlieb, L. (1978). Maternal attachment in primiparas. JOGNNursing, 39-44.  January-February,  Greiner, T. (1999). Factors associated with the duration of breastfeeding may depend on the extent to which mothers of young children are employed. Acta paediatr, 89, 1311-1319. Guldan, G.S., Maoyu, Z, Guo, Z., Junrong, H. & Yi, Y. (1995). Breastfeeding practices in Chengdu, Sishuan, China. Journal of Human Lactation, 11(1), 11-15. Hanson, L. (1998). Breastfeeding provides passive and likely long-lasting active immunity. Annals ofAllergy, Asthma & Immunology, 81, 523-529 Henderson, A., Stamp, G., & Pincombe, J. (2001). Postpartum positioning and attachment education for increasing breastfeeding: a randomized trial. Birth, 28, 236-242.  179  Hewat, R., & Elliis, D. (1986). Similarities and differences between women who breastfeed for short and long duration. Midwifery, 2, 37-43. Hills-Bonczyk, S., Tromiczak, K., Avery, M., Potter, S., Savik, K., & Duckett, L. (1994). Women's experiences with breastfeeding longer than 12 months. Birth, 21(4), 206-212. Hoddinott, P., & Pill, R. (1999). Nobody actually tells you: a study of infant feeding. British Journal ofMidwifery, 7(9) 558-565. Ineichen, B., Pierce, M., & Lawrenson, R. (1999). Young people's views on breastfeeding. British Journal ofMidwifery, 7(5), 309-311. Jamieson, L. (1999). New developments in breastfeeding initiatives. British Journal of Nidwifery, 7(5), 308.  Johnson, T., Brennan, R., & Flynn-Tymkow, C. (1999). A home visit program for breastfeeding education and support. Journal of Obstetric, Gynecologic, and Neonatal Nursing, 28(5), 480-485. Jones, D. (1986). Attitudes of breast-feeding mothers: A survey of 649 mothers. Social Science and Medicine, 23(11), 1151 -1156. Jordan, P., & Wall, V. (1993). Supporting the father when an infant is breastfed. Journal of Human Lactation, 3(1), 31-34. Kaufman, K.J., & Hall, L.A. (1989). Influences of the social network on choice and duration of breastfeeding in mothers of preterm infants. Research in Nursing and Health, 12, 149159. Kaufmann, T. (2000). Infant feeding: politics vs pragmatism? Royal College ofMidwives, 2(8), 244-245. Keith, K. (1997). Infant feeding choice among first-time mothers. Scholarly inquiry for Nursing Practice: An International Journal, 11(3), 199-229. Kendall-Tuckett, K.A., & Sugarman, M. (1995). The social consequences of long-term breastfeeding. Journal of Human Lactation, 11(3), 179-183. Kiehl, E.M., Anderson, G.C, Wilson, M.E., & Fosson, L. (1996). Social status, mother-infant time together and breastfeeding duration. Journal ofHuman Lactation, 12(3), 201-206.  180  King, D. (1999). Statistics - breastfeeding. International Journal of Childbirth Education,  14(2),  8-9. Kistin, N., Abramson, R., & Dublin, P. (1994). Effect of peer counselors on breastfeeding initiation, exclusivity and duration among low income urban women. Journal of Human Lactation, 10(1), 11-15.  Ko, M., & Schulken, E. (1998). Factors related to smoking cessation and relapse among pregnant smokers. American Journal of Health Behavior, 22(2), 83-89.  Koniak-Griffin, D. (1993). Maternal role attainment. IMAGE: Journal of Scholarship in Nursing, 25(3), 257-262. Labarere, J., Bellin, V., Fourny, M., Gagnaire, J-C., Francois, P., & Pons, J-C. (2003). Assessment of a structured in-hospital educational intervention addressing breastfeeding: a prospective randomized open trial. BJOG: An International Journal of Obstetrics and Gynecology, 110(9), 847-852.  Langley, C. (1998). Successful breastfeeding: what does it mean? British Journal of Midwifery, 6(5), 322-325. Lee, N. (1997). Observations based upon multiple telephone contacts with new breastfeeding mothers. Journal of Human Lactation, 13(2), 147-150.  Leff, E. (1988). Comparison of the effectiveness of videotape versus live group infant care classes. Journal of Obstetric, Gynecolgic, and Neonatal Nurses, 338-344.  Leff, E., Schriefer, J., Hagan, J., & DeMarco, P. (1995). Improving breastfeeding support: A community health improvement project. Journal on Quality Improvement, 21, 10, 521529. Lothian, J. (1995). It takes two to breastfeed - The baby's role in successful breastfeeding. Journal of Nurse-Midwifery,  40(A), 328-334.  Lowe, N. (1998). Breastfeeding information and support services offered by Melbourne hospitals in antenatal classes. Breastfeeding Review, 6(1), 23-28. McKenna, J., Mosko, S & Richard, C. (1997). Bedsharing promotes breastfeeding. Pediatrics, 100(2), 214-219. MacMullen, N., & Dulski, L. (2000). Factors related to sucking ability in healthy newborns. Journal of Obstetric, Gynecologic, and Neonatal Nurses, 29 (A), 390- 396.  181  Margolis, L.H. & Schwartz, J.B.(2000). The relationship between the timing of maternal postpartum hospital discharge and breastfeeding. Journal ofHuman Lactation, 16(2), 121-128. Matthews, M., Webber, K., McKim, E., Banoub-Baddour, S., & Laryea, M. (1995). Infant feeding practices in Newfoundland and Labrador. Canadian Journal ofPublic Health, 86(5), 296-299. McBride, C , Pirie, P., & Curry, S. (1992). Postpartum relapse to smoking: a prospective study. Health Education Research, 7(3), 381-390. Mullen, P., Richardson, M., Quinn, P., & Ershoff, D. (1997). Postpartum return to smoking: who is at risk and when. American Journal ofHealth Promotion, 11(5), 323-330. Publication Manual of the American Psychological Association. (5th ed.).(2001). Washington, D.C. Raisler, J. (2000). Midwives helping mothers to breastfeed: food for thought and action. Journal ofMidwifery & Women's Health, 45(3), 202-204. Raisler, J. (2000). Against the odds: breastfeeding experiences of low income mothers. Journal ofMidwifery & Women's Health, 45(3), 253-263. Ratner, P., Johnson, J., Bottorff, J., Dahinten, S., & Hall, W. (2000). Twelve month follow-up of a smoking relapse prevention intervention for postpartum women. Addictive Behaviors, 25,81-92. Riordan, J. (1997). The cost of not breastfeeding: A commentary. Journal ofHuman Lactation, 13(2), 93-97. Robertson, C , & Goddard, D. (1997). Monitoring the quality of breastfeeding advice. Health Visitor, 70(11), 422-424. Rossiter, J., & Yam, B. (2000). Breastfeeding: How could it be enhanced? The perceptions of Vietnamese women in Sydney, Australia. Journal ofMidwifery & Women's Health, 45(3), 202-204. Sable, M. & Patton, C. (1998). Prenatal lactation advice and intention to breastfeed: selected maternal characteristics. Journal ofHuman Lactation, 14(1), 35-40. Saunders, S., & Carroll, J. (1988). Post-partum breastfeeding support: Impact on duration. Journal of the American Dietetic Association, 88(2), 213-215.  182  Schafer, E., Vogel, M., Viegas, S. & Hausafus, C. (1998). Volunteer peer counsellors increase breastfeeding duration among rural low-income women. Birth, 25(2), 101-106. Shaw, E. & Kaczorowski, J. (1999). The effect of a peer counseling program on breastfeeding initiation and longevity in a low-income rural population. Journal ofHuman Lactation, 75(1), 19-25. Shrago, L. (1992). The breastfeeding dyad: early assessment, documentation and intervention. NAACOG's Clinical Issues in Perinatal and Women's Health Nursing, J(4),583-597. Shrago, L., & Bocar, D. (1990). The infant's contribution to breastfeeding. Journal of Obstetric, Gynecologic, and Neonatal Nursing, 19(3), 209-215. Snell, B., Krantz, M., Keeton, R., Delgado, K., & Peckham, C. (1992). The association of formula samples given at hospital discharge with the early duration of breastfeeding. Journal ofHuman Lactation, 8(2), 61-72. Stamp, G., & Crowther, C. (1995). Breastfeeding-Why Start? Why stop? A prospective survey of South Australian women. Breastfeeding Review, 3(1), 15-19. Starbird, E. (1991). Comparison of influences on breastfeeding initiation of firstborn children: 1960-69 vs 1970-79. Social Science and Medicine, 33(5), 627-634. Svedulf, C , Engberg, B., Berthold, H. & Hoglund, I. (1998). A comparison of the incidence of breast feeding two and four months after delivery in mothers discharged within 72 hours and after 72 hours post delivery. Midwifery, 14, 37-47. Turtle, C. (2000). An open letter to the WIC Program: The time has come to commit to breastfeeding. Journal of Human Lactation, 16(2), 99-103. Valdes, V., Pugin, E., Labbok, M., Periz, A., Cataian, S., Aravena, R. & Adler, M. (1995). The effects on professional practices of a three day course on breastfeeding. Journal of Human Lactation, 11(3), 185-190. Van Esterik, P. (1989/ Beyond the breast-bottle controversy. New Brunswick, New Jersey: Rutgers University Press. Vogel, A., & Mitchell, E. (1998). The establishment and duration of breastfeeding. Part 2: Community influences. Breastfeeding Review, 6(1), 11-16.  183  Waldenstrom, U., & Nilsson, C. (1994). No effect of birth centre care on either duration or experience of breastfeeding, but more complications: findings from a randomised controlled trial. Midwifery, 10, 8-17. Walker, L. O., Crain, H., & Thompson, E. (1986). Maternal role attainment and identity in the postpartum period: Stability and change. Nursing Research, 35(2), 68-71. Walker, L., Crain, H., & Thompson, E. (1986). Mothering behavior and maternal role attainment during the postpartum period. Nursing Research, 35(6), 352-355. Wambach. K. (1997). Breastfeeding intention and outcome: A test of the theory of planned behavior. Research in Nursing & Health, 20, 51-59. Wandersman, L. & Wandersman, A. (1980). Social support in the transition to parenthood. Journal of Community Psychology, 8, 332-342. Wang, Y. & Wu, S. The effect of exclusive breastfeeding and development and incidence of infection in infants. Journal ofHuman Lactation, 12(1), 27-30. Williams, J. & Mueller, S. (1989). A message to the nurse from the baby. Journal ofHuman Lactation, 5(19), 3-5. Williams, N. (1997). Maternal psychological issues in the experience of breastfeeding. Journal ofHuman Lactation, 13(1), 57-60. Wilmoth, T. & Elder, J. (1995). An assessment of research on breastfeeding promotion strategies in developing countries. Social Science and Medicine, 41(A), 579-594. World Health Organization. (1989). WHO/UNICEF. Protecting, Promoting and Supporting Breastfeeding: The Special Role ofMaternity Services. Geneva, Switzerland. Wright, A., Bauer, M., Naylor, A., Sutcliffe, E., & Clark, L. (1998). Increasing breastfeeding rates to reduce infant illness at the community level. Pediatrics, 101(5), 837-844. Wylie, M. (1979). The effect of expectations on the transition to parenthood. Sociological Focus, 12(4), 323-329. York, R., Brown, L., Samuels, P., Finkler, S., Jacobsen, B., Persely, C , et al. (1997). A randomized trial of early discharge and nurse specialist transitional follow-up care of high risk childbearing women. Nursing Research, 46(5), 254-261. Zimmerman, D. (1999). You can make a difference: Increasing breastfeeding rates in an innercity clinic. Journal ofHuman Lactation, 15(3), 217-220.  Appendix I. Pre-Workshop Guide  Appendix II. Outline of Interactive Group Session - (90 Minutes plus Refreshment Break) Introductory Remarks • Participants and support persons welcomed • Purpose and scope of session presented by facilitator Learning the Art of Breastfeeding • • •  Discussion pursued using slide presentation to highlight content Participants asked to refer to their pre-workshop guides, as needed Facilitator guided the discussion to include a range of topics including: Anticipating the realities of becoming first-time mothers The importance of feeling confident you will be successful What mothers have reported as being helpful when breastfeeding What mothers have reported as being challenging when breastfeeding Strategies to consider in planning help once home after birth Thinking about how to resume aspects of one's lifestyle Anticipating it takes time to practice and learn to breastfeed Remembering the baby needs time to learn to breastfeed too. Refreshment Break (15 minutes) Space included poster boards and handouts featuring related information  •  Facilitated Discussion resumed topics including: What mothers have reported as reasons for weaning How you can determine if the baby is receiving enough milk Alternatives to consider such as rooming-in after birth Learning to recognize your infant's hunger cues Determining if the baby has achieved a good latch Talking with your partner and family about breastfeeding Practicing various breastfeeding positions using a doll Including the partner's assistance in positioning for breastfeeding Contacting help early if problems experienced Contacting help early to obtain reassurance things are going well Frequency of feedings and cluster feedings Maintaining a baby diary to record feeding and elimination patterns Ways to determine your baby is receiving enough milk Alternatives to try when the baby is crying or fussy How partners, family and friends can help When and where to call for assistance Topics the participants identified through questions or comments.  197  Appendix III. Antenatal Data Collection - Personal Information  1. Current Date  /  /  2. Expected Date of Delivery  /  /  2. Prenatal Class Location: 3. Name:  4. Age (in years)  5. Home Address: 6. Home Telephone Number:  7. Alternate phone number:  8. Email address: 9. Name of support person:  10. Relationship:  11. Highest level of education achieved: other: ;  less than high school  12. Are you likely to move with the next year?  Yes  high school  college  university  No  If yes, please note any additional contact information the research assistant may require to contact you and approximate date of the move: 13. Were you breastfed?  Yes  No  Not sure  14. On a scale of 1 to 10, how committed to breastfeeding do you rate yourself? 15. Think about who you view as having the most influence on your decision to breastfeed your baby. Please specify the relationship of that person to you: 16. In your opinion a scale of 1 to 10, how committed to breastfeeding is that person? 17. On a scale of 1 to 10, how supportive is your partner of you breastfeeding? 18. On a scale of 1 to 10, how much physical help do your receive from others (e.g. vacuuming, household chores, etc.) 19. Who provides you with the most physical help? 20. Who provides you with emotional support?  Family and Friends  Professional Help  receive support 21. If you identified professional help, please indicate the type of professional who helps you: Nurse Doctor Social Worker Counselor 22. What concerns do you have about breastfeeding?  Other:  Do not  198  Appendix IV. Parenting Sense of Competence Scale (Scored 1-6 to reflect extent of agreement from Strongly Agree to Strongly Disagree) Please indicate the number that most closely indicates how you think you will feel about parenting: I. The problems of taking care of the baby will be easy to solve once you know how your actions affect your baby, an understanding I have acquired. 2.1 will meet my own personal expectations for expertise in caring for my baby. 3.1 will make a fine model of a new mother to follow in order to learn what she would need to know in order to be a good parent. 4. Being a parent will be manageable, and any problems will be easily solved. 5. If anyone can find the answer to what is troubling my baby, I will be the one. 6. It will be easy to know when you are doing a good job or a bad job being a parent. 7. Considering I am expecting my first child, I feel thoroughly familiar with the role. 8.1 honestly believe I will have all the skills necessary to be a good mother to my baby. 9.1 know being a parent will be rewarding and am not feeling frustrated. 10.1 feel like I will be in control and not being manipulated. II. 1 feel as prepared to be a good mother as my mother would have been when she had me. 12.1 feel like I will be able to complete what I need to do. 13. When I wake up in the morning, I will like I have accomplished a lot. 14. My talents and interests will include being a parent. 15. Being a mother of an infant will be interesting and I am motivated to do a good job. 16.1 will be relaxed and calm as a parent. 17. Being a good mother will be a reward in itself.  Appendix V. Breastfeeding Self Efficacy Scale (Scored 1-5 to reflect extent of confidence from Not at all to Always) Please note which of the following most closely describes how you think you will feel about breastfeeding... I will... 1. Ensure that my baby is properly latched on for the whole feeding. 2. Recognize when my baby is finished breastfeeding. 3. Manage the breastfeeding situation to my satisfaction. 4. Determine that my baby is getting enough milk. 5. Recognize the signs of a good latch. 6. Feel if my baby is sucking properly at my breast. 7. Maintain my milk supply by using the 'supply and demand' rule. 8. Finish feeding my baby on one breast before switching to the other breast. 9. Position my baby correctly at my breast. 10. Manage to keep up with my baby's breastfeeding demands. 11. Take my baby off the breast without pain to myself. 12. Hold my baby comfortably during breastfeeding. 13. Manage to breastfeed even if my baby is crying. 14. Monitor how much breast milk my baby is getting by keeping track of my baby's urine output and bowel movements. 15. Keep my baby awake at my breast during a feeding. 16. Feed my baby every 2-3 hours. 17. Deal with the fact that breastfeeding can be time consuming. 18. Accept the fact that breastfeeding may temporarily limit my freedom. 19. Comfortably breastfeed with my family members present. 20. Comfortably breastfeed in public places. 21. Count on my friends to support my decision to breastfeed 22. Keep feeling that I really want to continue to breastfeed my baby  Appendix V .  (Continued)  Breastfeeding Self Efficacy Scale  23. Keep wanting to breastfeed my baby. 24. Stay motivated to breastfeed my baby. 25. Breastfeed my baby without using formula as a supplement. 26. Motivate myself to breastfeed successfully. 27. Feed my baby with breast milk only. 28. Continue to breastfeed my baby for every feeding. 29. Depend on my family to support my decision to breastfeed. 30. Refrainfrombottle feeding for the first few weeks. 31. Successfully cope with breastfeeding like I have with other challenging tasks. 32. Be satisfied with breastfeeding experience. 33. Focus on getting through one feed at a time.  Appendix VI Maternal Breastfeeding Evaluation Scale (Scored 1-5 to reflect extent of agreement from Strongly Agree to Strongly Disagree) Please indicate which of the following best describes how you think you will feel about breastfeeding 1. With breastfeeding I will feel a sense of inner contentment. 2. Breastfeeding will be a special time with my baby. 3. My baby won't be interested in breastfeeding. 4. My baby will love to nurse. 5. It will be a burden being my baby's main source of food. 6.1 will feel extremely close to my baby when I breastfed. 7. My baby will be an eager breastfeeder. 8. Breastfeeding will be physically draining. 9. It will be important to me to be able to nurse. 10. While breastfeeding, my baby's growth will be excellent. 11. My baby and I will work together to make breastfeeding go smoothly. 12. Breastfeeding will be a very nurturing, maternal experience. 13. While breastfeeding, I will feel self-conscious about my body. 14. With breastfeeding, I will feel too tied down all the time. 15. While breastfeeding, I will worry about my baby gaining enough weight. 16. Breastfeeding will be soothing when my baby is upset or crying. 17. Breastfeeding will be like a high of sorts. 18. The fact that I could produce the food to feed my own baby will be very satisfying. 19. In the beginning, my baby will have trouble breastfeeding. 20. Breastfeeding will make me feel like a good mother. 21.1 will really enjoy nursing. 22. While breastfeeding, I will be anxious to have my body back. 23. Breastfeeding will make me feel more confident as a mother.  202 Appendix VI (Continued) Maternal Breastfeeding Evaluation Scale 24. My baby will gain weight really well with breastmilk. 25. Breastfeeding will make my baby feel more secure. 26.1 will easily fit my baby's breastfeeding with my other activities. 27. Breastfeeding will make me feel like a cow. 28. My baby will not relax while nursing. 29. Breastfeeding will be emotionally draining. 30 Breastfeeding will feel wonderful to me.  Appendix VII Multidimensional Assessment of Fatigue Scale (Items 1-14 scored 1-10 to reflect extent of intensity or frequency from not at all to a great deal.) Please indicate the number that most closely indicates how you have been feeling the past week: 1. To what degree have you experienced fatigue? 2. How severe is the fatigue which you have been experiencing? 3. To what degree has fatigue caused you distress? In the past week to what degree has fatigue interfered with your ability to: 4. Do household chores?  5. Cook?  6. Bathe or wash?  7. Dress?  8. Work?  9. Visit or socialize with friends or family?  10. Engage in sexual activity? 11. Engage in leisure and recreational activities/ 12. Shop and do errands? 13. Walk? 14. Exercise other than walking? 15. Over the past week, how often have you been fatigued? Every day Most, but not all days Occasionally, but not most days Hardly any days 16. To what degree has your fatigue changed during the past week? Increased Fatigue has gone up and down Stayed the same Decreased  Appendix VIII Edinburgh Postnatal Depression Scale Please indicate which of the following statements comes closest to how you have felt during the past week: 1.1 have been able to laugh and see the funny side of things. As much as I always could  . Definitely no so much now  . Not quite so much now  .Notat'all  2.1 have looked forward with enjoyment to things: As much as I ever did  _ . Definitely less than I used to  Rather less than I used to  _ .Hardly at all  3.1 have blamed myself unnecessarily when things went wrong: Yes, most of the time  _ .Not very often  Yes, some of the time  .No, never  4,1 have felt worried and anxious for no very good reason: _  No, not at all Hardly ever  _ . Yes, sometimes ,  _ . Yes, very often  5.1 have felt scared or panicky for no very good reason: Yes, quite a lot  _ . No, not much  Yes, sometimes  _ .No, not at all  6. Things have been getting on top of me: Yes, most of the time I haven't been able to cope at all Yes, sometimes I haven't been coping as well as usual No, most of the time I have coped quite well No, I have been coping as well as ever  Appendix VIII (Continued) Edinburgh Postnatal Depression Scale 7.1 have been so unhappy that I have had difficulty sleeping: Yes, most of the time  Not very often  Yes, sometimes  No, not at all  8.1 have felt sad or miserable: Yes, most of the time  Not very often  Yes, sometimes  Not, not at all  9.1 have been so unhappy that I have been crying: Yes, most of the time Yes, sometimes  Not very often • No not at all  10. The thought of harming myself has occurred to me: Yes, quite often  Hardly ever  Sometimes  Never  206  Appendix IX. Postnatal Data Collection - Personal Information and Feeding Data - 2 weeks 1.  Current date:  ;  Date of Delivery:  Delivery site: _ _ _ V G H LDRP:  Yes  SBGH  SVD  C/S  Length of hospital stay:  Analgesia during labor: Do you smoke?  narcotic  ;Yes  Vacuum  13-24 hrs  0-12 hrs  25-48 hrs  13-24 hrs epidural  > 48 hrs  25-36 hrs  entonox  other:  No  < while pregnant  Infant's Health:  Forceps  0-12 hrs  Length of labor (in hours):  2.  Other:  No  Delivery mode:  If yes,  HSC  > while pregnant  Excellent  Good  Fair  same as while pregnant Poor  If fair or poor, please explain:  Gender: Birth wt:  Male:  Female  If male, circumcised?  Current wt:  When last weighed, wearing:  Yes  No  Date last weighed clothes  Supplements offered to infant in hospital?  diaper Yes  no clothes or diaper No  If yes, how often did the infant take supplements while in hospital and from who?  If yes, what supplement(s)?:  3.  formula  water  _other:  Feeding Data: Expressed Breastmilk (EBM) A. Exclusive Breastfeeding or EBM: B. Primarily Breastfeeding or EBM, plus up to one alternate feeding per week: C. Mainly Breastfeeding or EBM, plus up to one alternate feeding per day: D. Partial Breastfeeding or EBM, plus mainly alternate feedings per day: E. Weaned (no breastfeeding or EBM for one week or more):  Date:  207  Appendix IX. (Continued) Postnatal Data Collection - Personal Information and Feeding Data - 2 weeks (Continued)  If A-D: who has helped you to continue to breastfeed your baby and how?  If A-D: approximate number of breastfeeds during the dav  during the night  If EBM, given per:  finger  cup  bottle _ dropper  teaspoon  other:  If E, "how did you know it was time to stop breastfeeding?" 4.  Who feeds the baby most of the time? LC other:  5.  Have you contacted any specific help in relation to breastfeeding (La Leche League, Breastfeeding Hotline, etc)? LLL Hotline CHN Friend Mother Other Family Member  6.  Dad  Nurse  Grandmother  Other:  Have you completed and mailed in the second package of research questionnaires? Yes  7.  Mom  No/Plan:  Replacements sent if needed  Any additional comments you would like to add in relation to the study?  THANK YOU FOR PARTICIPATING IN THIS RESEARCH  Appendix X. Postnatal Data Collection - Personal Information and Feeding Data -12 weeks 1.  Mom's current health:  Excellent  Good  Fair  Poor  If fair or poor, please explain: Do you smoke? If yes, 2.  Yes  No  < while pregnant  Infant's Health:  Excellent  > while pregnant Good  Fair  same as while pregnant  Poor  If fair or poor, please explain:  Birth wt:  •  Current wt:  When last weighed, wearing: 3.  Date last weighed clothes  diaper  no clothes or diaper  Feeding Data: Expressed Breastmilk (EBM) A. Exclusive Breastfeeding or EBM: B. Primarily Breastfeeding or EBM, plus up to one alternate feeding per week: _ C. Mainly Breastfeeding or EBM, plus up to one alternate feeding per day: D. Partial Breastfeeding or EBM, plus mainly alternate feedings per day: E. Weaned (no breastfeeding or E B M for one week or more): *  Date:  If A-D: who has helped you to continue to breastfeed your baby and how?  If A-D: approximate number of breastfeeds during the day  during the night  If E B M , given per:  finger  cup  bottle  dropper  teaspoon  If E, "how did you know it was time to stop breastfeeding?"  other:.  209 Appendix X . (Continued) Postnatal Data Collection - Personal Information and Feeding Data -12 weeks (Continued) 4.  Who feeds the baby most of the time? LC other:  5.  Have you contacted any specific help in relation to breastfeeding (La Leche League, Breastfeeding Hotline, etc)? LLL Hotline CHN Friend Mother Other Family Member  6.  Mom  Dad  Nurse  Grandmother  Other:  Have any solids or semi-solid foods been introduced?  No  Yes  If Yes, what has been offered to the infant, in what quantity and how often?  7.  Have you completed and mailed in the second package of research questionnaires? Yes  8.  No/Plan:  Replacements sent if needed  Any additional comments you would like to add in relation to the study?  THANK YOU FOR PARTICIPATING IN THIS RESEARCH  210  Appendix XI. Postnatal Data Collection - Personal Information and Feeding Data - 24 weeks 1.  Mom's current health  Excellent  Good  Fair  Poor  If fair or poor, please explain: Do you smoke? _ _ _ _ Yes If yes, 2.  No  < while pregnant  Infant's Health:  Excellent  > while pregnant Good  Fair  same as while pregnant  Poor  If fair or poor, please explain:  Birth wt:  Current wt:  When last weighed, wearing: 3.  Date last weighed clothes  diaper  no clothes or diaper  Feeding Data: Expressed Breastmilk (EBM) A. Exclusive Breastfeeding or EBM: B. Primarily Breastfeeding or EBM, plus up to one alternate feeding per week: _ C. Mainly Breastfeeding or EBM, plus up to one alternate feeding per day: D. Partial Breastfeeding or EBM, plus mainly alternate feedings per day: E. Weaned (no breastfeeding or E B M for one week or more):  Date:  If A-D: who has helped you to continue to breastfeed your baby and how?  If A-D: approximate number of breastfeeds during the day  during the night  If EBM, given per:  finger  cup  bottle  dropper  teaspoon  If E, "how did you know it was time to stop breastfeeding?"  other:.  211  Appendix XI. (Continued) Postnatal Data Collection - Personal Information and Feeding Data - 24 weeks 4.  Who feeds the baby most of the time? LC other:  5.  Have you contacted any specific help in relation to breastfeeding (La Leche League, Breastfeeding Hotline, etc)? LLL Hotline CHN Friend Mother Other Family Member  6.  Mom  Dad  Nurse  Grandmother  Other:  Have any solids or semi-solid foods been introduced?  No  Yes  If Yes, what has been offered to the infant, in what quantity and how often?  7.  Have you completed and mailed in the second package of research questionnaires? Yes  8.  No/Plan:  Replacements sent if needed  Any additional comments you would like to add in relation to the study?  THANK YOU FOR PARTICIPATING IN THIS RESEARCH  Appendix XII. Content of Intervention versus Usual Care  2! »  ! i & s  5- a O 5' a oo o a  jf|  p.  get  t r 5. S g> s o n o  g05  g; CD*  09  3. S € o 5 co ct m  3  e  o CD C>c < o CD CD ^ o o  g- "> o'  g  c? cr  3  ^•1  S era  •5 1=8 CD  era a  # 1 1  05 O  O  8 - ^ if Eo g ^  o  a. E*  to  215  Appendix XIII. Letter of Information THE  UNIVERSITY  OF  BRITISH  COLUMBIA  You are invited to be part of a nursing research study being conducted by Sandy Kluka RN, for her doctoral dissertation. By participating, you will be helping nurses and other health professionals to better understand how they may help new mothers who plan to breastfeed. Participation requires: 1. Signing a consent form once you have heard about the study, and had a chance to ask any questions and decide if you are interested in participating in the study. You're participation is completely voluntary, and you may choose to not continue to participate at any time by stating you wish to withdraw from the study. You may refuse to answer any question at any time. The care you are to receive both now, and after your baby is born, will not be affected by your decision about whether or not to participate in the study. 2. Completing a questionnaire that asks your name, age, education, contact information, and when you expect to deliver your baby. You will be asked to hand in that information tonight. It should not take longer than 5-10 minutes to fill in this information. 3. Taking home some further questions to complete and submit by mail (in pre-addressed and stamped envelopes). These questions ask how you are feeling about becoming a parent, and what you think it will be like to breastfeed your baby. Other questions ask about how tired you are feeling, your mood, and how much help you expect to receive. It may require up to 30 minutes to answer the questions. Your responses are to be mailed at four different times: within the next week, and then when the baby is two weeks, 3 months, and 6 months of age. 4. Your name does not need to be stated on the information you provide by mail. You will be contacted by a research assistant by telephone at each of the times after the baby is born to remind you about completing the questionnaires, and ask you some more details about how you and your baby are managing. For example, you would be asked how much the baby weighs, how you are feeling, and how well you think you are managing as a new parent. The phone call may require up to 10 minutes of your time. 5. Any information you provide will be kept confidential and used only for the purpose of the study. At no time will the information you provide be reported to anyone other than the person conducting the study, or a research assistant. The information will be stored in a locked cabinet file until it is no longer required, and then shredded. The responses you provide will not be stored with your name, but rather with a numerical code. 6. Participation in the study may or may not require your attendance (with a support person, if preferred) at one additional prenatal class that will further explore breastfeeding content,  216  Appendix XIII. Letter of Information (Continued) common problems experienced and resources available. The study design will determine who is to attend the extra class. There would not be an additional cost and it would not require more than 2 hours of your time to attend. I would contact those women (and their support person) who are to attend the class by telephone within the next two days to provide further details about the class time, location, etc. 7. There are no known risks to participating in the study. Some participants may benefit by attending the additional class or by filling out the questionnaires which may assist in their self evaluation of how they are managing as a new parent. 8. You may keep a copy of the consent form. You also may obtain a copy of the study results by completing the optional note listed on the consent form or telling me you would like to hear about the study outcomes. If you would like to participate in the study, please review the consent form that I have provided and return it to me. I would be happy to answer any questions that you may have about the study. Thank you for considering this opportunity.  217  Appendix XIV. Consent Form THE  UNIVERSITY  OF  BRITISH  C O L U M B I A  Principle Investigator: Sandy Kluka RN, PhD(c) - Graduate student, University of British Columbia (UBC) Purpose: A study is being conducted to better understand how nurses can assist first time mothers who plan to breastfeed. The study has been approved by an ethical review board for education and nursing. Women who meet the selection criteria will be asked to remain for an additional 15 minutes after one of the prenatal classes to be introduced to the study by the principal investigator or research assistant, and invited to participate. Study Procedures: I understand that if I agree to participate: -1 will be asked to complete a number of questions including my name, age and other personal information as well as information about how I feel about becoming a parent and breastfeeding my baby. It should take about 5 minutes to complete the questions. -1 will be given four addressed and stamped envelopes that each contain additional questions I am to answer at home and then mail to the Manitoba Nursing Research Institute within the next week and after I have the baby i.e. when the baby is 2 weeks, 3 months, and 6 months old. -1 will be contacted by telephone on three occasions after I have the baby by a research assistant who will ask me questions such as my baby's age and weight that I can answer by telephone. The call will require up to 10 minutes. The assistant will also remind me to complete and mail the additional questions. - An assistant will select a sealed envelope and inside it will say whether I am to attend an additional prenatal class. I may or may not be required to attend the extra session that will provide further information about what to expect as a first time breastfeeding mother. I will be phoned by the research nurse within the next two days if I am to attend. Confidentiality Any information resulting from this research study will be kept strictly confidential. All documents will be identified only by a code number and kept in a locked filing cabinet. Participants will not be identified by name in any reports of the completed study. Security of the computer records will also be maintained through use of numerical coding. Contact If I have any questions or desire further information with respect to this study, I may contact Sandy Kluka or a research assistant by calling (XXX) XXX-XXXX.  218  Appendix XIV. Consent Form (Continued) If I have any concerns about my treatment or rights as a research subject, I may contact the Director of Research Services at the University of British Columbia, (XXX) XXX-XXXX. Consent I understand that my participation in this study is entirely voluntary and that I may refuse to participate or withdraw from the study at any time without jeopardy to the care I am to receive now or after my baby has been born. I may refrain from answering any questions. I have received a copy of this consent form for my own records! I consent to participate in this study.  Signature of Subject  Date  Signature of Witness  Date  I would like the investigator to mail me a one page summary of the study findings upon completion: Yes  No  219  Appendix XV. Agreement of Research Assistant(s) and Statistician to Maintain Confidentiality  Regarding the University of British Columbia, PhD in Nursing dissertation research titled: "A randomized controlled trial to effect breastfeeding duration: A synthesis of prenatal nursing intervention strategies for primiparous women who are planning to breastfeed," being conducted by Sandy Kluka RN, PhD(c), I agree to the following: 1. To maintain confidentiality of any information collected in relation to the study at all times. 2. To store data in a designated and locked cabinet file for the duration of the study. 3. To convey information to prospective or actual study participants in a respectful, neutral and professional manner, as outlined by the principal investigator. 4. To notify the principal investigator of any concerns expressed by the study participants in relation to the conduct of the study. 5. To use identification codes only when entering, cleaning, validating or analyzing raw data. 6. To abide by any directive deemed necessary by the principal investigator in order to maintain confidentiality and anonymity, whenever possible.  Date:  Signature of Research Assistant or Statistician  Printed Name  

Cite

Citation Scheme:

        

Citations by CSL (citeproc-js)

Usage Statistics

Share

Embed

Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                        
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            src="{[{embed.src}]}"
                            data-item="{[{embed.item}]}"
                            data-collection="{[{embed.collection}]}"
                            data-metadata="{[{embed.showMetadata}]}"
                            data-width="{[{embed.width}]}"
                            async >
                            </script>
                            </div>
                        
                    
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:
http://iiif.library.ubc.ca/presentation/dsp.831.1-0091835/manifest

Comment

Related Items