Open Collections

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Sex and gender differences in symptoms of acute coronary syndromes Mackay, Martha Helen 2010

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

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

Item Metadata

Download

Media
24-ubc_2010_spring_mackay_martha.pdf [ 2.82MB ]
Metadata
JSON: 24-1.0069113.json
JSON-LD: 24-1.0069113-ld.json
RDF/XML (Pretty): 24-1.0069113-rdf.xml
RDF/JSON: 24-1.0069113-rdf.json
Turtle: 24-1.0069113-turtle.txt
N-Triples: 24-1.0069113-rdf-ntriples.txt
Original Record: 24-1.0069113-source.json
Full Text
24-1.0069113-fulltext.txt
Citation
24-1.0069113.ris

Full Text

 SEX AND GENDER DIFFERENCES IN SYMPTOMS OF ACUTE CORONARY SYNDROMES  by  MARTHA HELEN MACKAY B.S.N., University of British Columbia, 1986 M.S.N., University of British Columbia, 1997  A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF  DOCTOR OF PHILOSOPHY in THE FACULTY OF GRADUATE STUDIES (Nursing)  THE UNIVERSITY OF BRITISH COLUMBIA (Vancouver)    January, 2010  © Martha Helen Mackay, 2010  ii ABSTRACT  Background: Better understanding of acute coronary syndrome (ACS) symptoms is needed to improve diagnosis. Prior research has suggested sex or gender differences in ACS symptoms, but these studies have been fraught with methodological issues. Using percutaneous coronary intervention (PCI), specifically angioplasty balloon inflation to model myocardial ischemia, this study examined whether sex or gender differences exist in reported symptoms of ACS, and other predictors of ACS symptoms. Methods: Consecutive patients having non-emergent PCI were prospectively recruited. Hemodynamic instability, left bundle branch block and total occlusion were exclusion criteria. Prior to PCI, descriptions of prior symptoms that had led to referral for PCI were obtained. Balloon inflation was sustained for two minutes unless a clinical reason to deflate occurred. During inflation, subjects were questioned about current symptoms. Concurrent ECG data were collected. Findings: Of the final sample of 305 (39.7% women; mean age 63.9 years (SD = 10.6), 245 (83%) had ECG-evident ischemia during inflation. No sex/gender differences were found in rates of reporting chest discomfort or “typical” symptoms, regardless of ischemic status. Women were significantly more likely to report throat, jaw and neck discomfort, as well as only non-chest discomfort. Controlling for age, diabetes, urgency of procedure, prior MI or prior PCI increased the sex/gender effect. Increased age, urgency, prior MI and PCI were also covariates with specific symptoms. Conclusions: This prospective study with ECG confirmation of ischemia suggests women and men have similar rates of chest discomfort and other “typical” symptoms during ACS. However women are more likely to experience throat, jaw and neck discomfort. Although other factors  iii also influence reported symptoms, they do not diminish the effect of sex/gender. There has been suggestion in both the popular press and patient education materials that women experience ACS very differently from men. Therefore, it is important that clear educational messages be crafted to ensure both women and health professionals realise that classic symptoms of ACS are equally common in women and men.  iv TABLE OF CONTENTS ABSTRACT.................................................................................................................................................ii TABLE OF CONTENTS........................................................................................................................iv LIST OF TABLES ..................................................................................................................................viii LIST OF FIGURES...................................................................................................................................x ACKNOWLEDGEMENTS ...................................................................................................................xi CHAPTER 1: SEX AND GENDER DIFFERENCES IN SYMPTOMS OF ACUTE CORONARY SYNDROMES..................................................................................................................1 Background .............................................................................................................................................2 Sex and Gender: Concepts and Terminology ....................................................................................4 Context for the Study ............................................................................................................................6 Research Questions................................................................................................................................6 Significance..............................................................................................................................................8 CHAPTER 2: REVIEW OF THE LITERATURE .............................................................................9 Why Study Sex and Gender Differences?...........................................................................................9 Sex/Gender Differences in Symptoms of Acute Coronary Syndromes ......................................12 Published Reviews of the Literature..............................................................................................15 Other Studies Published from 2004-2009 ....................................................................................21 Causes of Sex/Gender Differences in Symptoms of ACS.............................................................26 Other Factors Associated with Symptoms of Acute Coronary Syndrome..................................28 Age .....................................................................................................................................................29 Diabetes.............................................................................................................................................29 Ethnicity and Race ...........................................................................................................................30 Alternative Approaches to Symptom Analysis ................................................................................32  v Sex/Gender Differences in Pain Experience...................................................................................34 Sex Differences in Clinical Pain.....................................................................................................34 Proposed Underlying Mechanisms................................................................................................35 Sex and Gender Differences in Non-Cardiac-Specific Symptoms ...............................................38 Shortness of Breath .........................................................................................................................38 Diaphoresis .......................................................................................................................................39 Nausea and Vomiting ......................................................................................................................39 CHAPTER 3: METHODS .....................................................................................................................41 Objective................................................................................................................................................41 Ethics Approval....................................................................................................................................41 Sample and Site.....................................................................................................................................42 Inclusion Criteria..............................................................................................................................42 Exclusion Criteria.............................................................................................................................42 Screening, Recruitment and Consent ................................................................................................44 Data Collection Procedures ................................................................................................................44 Measures ................................................................................................................................................46 Demographic Data...........................................................................................................................46 Clinical and Procedural Data..........................................................................................................47 Symptom and Ischemia Monitoring..............................................................................................49 Analysis ..................................................................................................................................................51 Characterization of the Sample ......................................................................................................51 Frequency and Nature of Symptoms ............................................................................................53 Identification of Myocardial Ischemia ..........................................................................................56 Sex/Gender Differences in Symptom Profiles............................................................................57  vi Analysis for Latent Classes .............................................................................................................60 CHAPTER 4: FINDINGS .....................................................................................................................62 Study Sample .........................................................................................................................................62 Margin of Error for the Sample .........................................................................................................64 Demographic Characteristics of the Sample ....................................................................................65 Baseline Clinical Characteristics .........................................................................................................69 Procedural Characteristics ...................................................................................................................73 Subjective Experience of Myocardial Ischemia during Balloon Inflation....................................77 Frequency of Reporting Any Symptoms ......................................................................................77 Number of Symptoms Reported ...................................................................................................78 Intensity of Reported Symptoms...................................................................................................79 Location and Type of Reported Symptoms.................................................................................83 Sex/Gender Differences in the Location or Type of Symptoms Reported............................92 Concordance of Reported Baseline Symptoms with Symptoms Reported During Balloon Inflation .......................................................................................................................................... 106 Exploration for Latent Classes.................................................................................................... 108 Summary ............................................................................................................................................. 114 CHAPTER 5: DISCUSSION.............................................................................................................. 115 Summary of Findings........................................................................................................................ 115 Strengths and Limitations ................................................................................................................ 117 Strengths......................................................................................................................................... 117 Limitations ..................................................................................................................................... 118 Current Findings in Relation to Other Evidence ......................................................................... 124 No Sex/Gender Difference in Chest Discomfort or Other Typical Symptoms................. 124  vii Sex/Gender Differences in the Number and Intensity of Symptoms.................................. 128 Sex/Gender Differences in Other Symptoms.......................................................................... 131 Two Classes of Symptom Expression ....................................................................................... 132 Implications for Practice .................................................................................................................. 134 Patient and Public Education...................................................................................................... 134 Clinical Assessment....................................................................................................................... 136 Professional Education ................................................................................................................ 137 Implications for Research ................................................................................................................ 139 Conclusions ........................................................................................................................................ 141 REFERENCES...................................................................................................................................... 143 APPENDIX A:  Research Ethics Boards Approvals....................................................................... 164 APPENDIX B:  Informed Consent Forms....................................................................................... 168 APPENDIX C:  Data Collection Tools ............................................................................................. 177 APPENDIX D: Raw Symptom Descriptors and Codes ................................................................. 184  viii LIST OF TABLES Table 1  Summary of Symptom Classification, by Guideline .............................................................14 Table 2  Summary of Studies Examining Sex/Gender Differences in ACS or MI, 2004-2009....23 Table 3.  Margins of Error for Point Estimates of Frequency, by Sample Size ..............................65 Table 4  Demographic Characteristics of Sample ................................................................................66 Table 5  Comparison of Sample Demographic Characteristics with British Columbia Population who Self- Reported Having Heart Disease ...........................................................................................69 Table 6  Baseline Clinical Characteristics of Sample............................................................................71 Table 7  Procedural Characteristics of Sample .....................................................................................74 Table 8  Presence of Any Reported Symptoms during Balloon Inflation, by Gender ...................78 Table 9  Number of Reported Symptoms at Baseline, by Gender....................................................79 Table 10  Number of Reported Symptoms during Balloon Inflation, by Gender ..........................79 Table 11  Intensity of Main Symptom, by Gender (including those reporting no inflation discomfort).................................................................................................................................................80 Table 12  Intensity of Main Symptom, by Gender (excluding those reporting no discomfort at inflation) .....................................................................................................................................................80 Table 13  Location/Type and Frequency of Reported Main Symptoms, by Gender.....................85 Table 14  Location/Type and Frequency of All Reported Symptoms, by Gender ........................89 Table 15  Frequency of Chest Discomfort Reported as Main Symptom at Baseline, by Gender 93 Table 16  Frequency of Ever-Reported Chest Discomfort at Baseline, by Gender .......................94 Table 17  Frequency of Chest Discomfort Reported as Main Symptom during Balloon Inflation, by Gender...................................................................................................................................................94 Table 18  Frequency of Ever-Reported Chest Discomfort during Balloon Inflation, by Gender94 Table 19  Frequency of Any “Typical” a Symptom Reported as Main Symptom at Baseline, by Gender ........................................................................................................................................................95 Table 20  Frequency of “Typical” Ever-Reported Symptoms at Baseline, by Gender ..................95 Table 21  Frequency of Any “Typical” a Symptom Reported as Main Symptom during Balloon Inflation, by Gender .................................................................................................................................96 Table 22 Frequency of “Typical” Ever-Reported Symptoms during Balloon Inflation, by Gender .....................................................................................................................................................................96 Table 23  Frequency of Any of “Milner-Typical”  Symptoms Reported as Main Symptom at Baseline, by Gender ..................................................................................................................................97  ix Table 24  Frequency of Any “Milner-Typical” Symptoms Reported as Main Symptom at Balloon Inflation, by Gender .................................................................................................................................97 Table 25  Frequency of Any “Milner-Typical” Ever-Reported Symptoms during Balloon Inflation, by Gender .................................................................................................................................97 Table 26  Frequency of Reporting Only Non-Chest-Discomfort Symptoms at Baseline, by Gender ........................................................................................................................................................98 Table 27  Frequency of Reporting Only Non-Chest-Discomfort Symptoms during Balloon Inflation, by Gender .................................................................................................................................98 Table 28  Association of Gender with Selected Symptoms Reported at Baseline (unadjusted and adjusted; men as referent) ..................................................................................................................... 100 Table 29  Association of Gender with Selected Symptoms Reported during Balloon Inflation (unadjusted and adjusted; men as referent) ........................................................................................ 104 Table 30  Concordance of Reported Baseline and Balloon Inflation Symptoms......................... 107 Table 31  Percentage of Participants Reporting Selected Symptoms at Baseline, by Class ........ 109 Table 32  Unadjusted and Adjusted Association of Class Membership with Selected Socio- Demographic and Clinical Variables (Class 2 as referent) ............................................................... 113  x LIST OF FIGURES Figure 1. Details of all phases of participant screening and enrollment. ..........................................63 Figure 2  Intensity of main reported symptom at baseline, by gender (with and without those reporting no discomfort)..........................................................................................................................81 Figure 3  Intensity of main reported symptom during balloon inflation, by gender (with and without those reporting no discomfort). ...............................................................................................82 Figure 4  Proportion of participants reporting selected symptoms, by class. ............................... 111  xi ACKNOWLEDGEMENTS  I would like to acknowledge the expert support that I received from my dissertation committee members, Dr. Joy Johnson and Dr. Karin Humphries, which was given patiently and generously. Their suggestions and ideas were invaluable and always offered in a constructive and respectful way. I have learned a great deal from them. I would particularly like to acknowledge my research supervisor, Dr. Pam Ratner. Her mastery of all aspects of research is formidable and a real inspiration to me. Her obvious love of research, her modelling of impeccable standards at every step, her unwavering enthusiasm and commitment to my learning goals and progress, and her ability to combine instruction and coaching with collegiality all combined to give me a very privileged learning experience for which I am enormously thankful.  1 CHAPTER 1: SEX AND GENDER DIFFERENCES IN SYMPTOMS OF ACUTE CORONARY SYNDROMES  “Many of the symptoms of heart disease are often ignored, unrecognised or misdiagnosed, because women’s symptoms are completely different than men’s.” - The Oprah Winfrey Show website www.oprah.com: “The facts about heart disease for women”. For decades, cardiovascular diseases have topped all others as a cause of mortality amongst Canadians. In 2004, the most recent year for which Statistics Canada data are available, mortality from heart disease was highest amongst all cardiovascular diseases. Acute coronary syndrome (ACS) is the umbrella term for the clinical syndromes of unstable angina (UA) and the two main categories of myocardial infarction: non-ST-elevation myocardial infarction (NSTEMI) and ST-elevation MI (STEMI). ACS is most commonly caused by atherosclerotic coronary artery disease (CAD). Not surprisingly, ischemic heart disease, of which CAD is one type, is the most common cause of mortality from all heart diseases, leading to an annual death rate of 96.2 per 100,000 population (Statistics Canada, 2007a). Not only is there high mortality from heart disease, it also exacts a toll on quality of life. For example, people with heart disease are about six times more likely to report difficulty ambulating compared with those without heart disease (Manuel, Leung, Nguyen, Tanuseputro, & Johansen, 2006). Only recently have sex and gender differences in the epidemiology and other aspects of heart disease been appreciated. Although the age-standardised death rate is higher for men than women (223.7 vs 137.9 per 100,000 population), the absolute number of women dying is higher than men (186,579 vs. 184,282) (Statistics Canada, 2007a). And, while it is encouraging that the overall rate of death from cardiovascular causes has been declining steadily since the late 1970s, this reduction has been less pronounced amongst women (whose death rate dropped 4.3%) compared with men (whose death rate dropped 5.5%) (Statistics Canada, 2007b). As for  2 measures of health-related quality of life, women consistently report more severe pain, discomfort, activity restriction and disability secondary to their heart disease compared with men (Pilote et al., 2007). While this comparison is not meant to engage in the unproductive exercise of pitting men’s health issues against women’s, it is meant to underscore the fact that, contrary to beliefs that have been held by the lay public and perhaps even health professionals, until recently (Finnegan et al., 2000; Mosca, Ferris, Fabunmi, & Robertson, American Heart Association, 2004), heart disease in general, and ischemic heart disease in particular, causes a significant burden for women. Accordingly, it should be of concern to Canadian health professionals. A key to improving disability and mortality rates for atherosclerotic CAD lies in the ability to improve diagnosis, and thereby improve the timeliness of treatment and outcomes for ACS. Background Although some researchers have found no differences between men’s and women’s treatment-seeking behaviour (Morgan, 2005; Moser, McKinley, Dracup, & Chung, 2005; Zerwic, Ryan, DeVon, & Drell, 2003), it has been demonstrated in relatively more studies that women delay seeking treatment for ACS longer than do their male counterparts (Dracup & Moser, 1991; Lefler & Bondy, 2004; Meischke, Ho, Eisenberg, Schaeffer, & Larsen, 1995; Ottesen, Kober, Jorgensen, & Torp-Pedersen, 1996). This potential delay has clinical importance because favourable outcomes from ACS, and more specifically ST-elevation MI, are exquisitely dependent on timely treatment to achieve reperfusion (Jacobs, Antman, Faxon, Gregory, & Solis, 2007). Recently Lefler and Bondy (2004) synthesized 48 studies published between 1995 and 2004 that examined reasons for women’s treatment-seeking delay. In their synthesis, the most common reasons for delaying that emerged were related to the nature of the symptoms experienced. This body of literature, coupled with irrefutable evidence for the benefit of timely  3 reperfusion, naturally has led researchers to focus attention on symptoms of myocardial ischemia experienced by women: are they actually different from men’s? If so, how and why do they differ? Possible explanations for differences in the diagnosis, treatment and outcomes of CAD have been postulated, and have included sex-based physiology, psychosocial influences, and provider bias (Rosenfeld, 2006). The first two explanations may also shed light on the reasons for symptom differences, but the more fundamental questions of whether and exactly how women’s symptoms differ from men’s, despite a vast array of studies, have not been fully answered. Some of the problems with past research have been related to retrospective designs, with heavy reliance on review of the health record as well as the patient’s recall; lack of verifiability that ischemia was present at the time of the event the participant described; lack of standardisation in data collection; problematic collapsing or grouping of variables; selection bias introduced through inclusion of only patients presenting with chest pain; and the influence of the physician’s pre-test suspicion of ACS on the acquisition and documentation of the history (Canto et al., 2007). And so, even though there have been many studies and at least three excellent reviews of this body of literature (Canto et al., 2007; Chen, Woods, & Puntillo, 2005; DeVon & Zerwic, 2002), and although there is acknowledgement among cardiovascular health professionals of the importance of a full understanding of this phenomenon (Shaw et al., 2006), conclusive answers elude us, yet. An excellent opportunity exists during percutaneous coronary intervention (PCI) procedures to compare the symptoms experienced by men and women at the moment of confirmed ischemia. The PCI procedure entails temporarily obstructing coronary blood flow during inflation of a balloon at the tip of the angioplasty catheter, which causes ischemia distal to the balloon. The purpose of the research presented here was to evaluate, in what was anticipated  4 to be a controlled simulation of ACS, the qualitative nature of patients’ experiences during myocardial ischemia, and to identify whether there are sex/gender differences in this experience. Sex and Gender: Concepts and Terminology A brief discussion of the concepts of gender and sex is warranted at this point. Sex refers to “the biological characteristics such as anatomy (e.g., body size and shape) and physiology (e.g., hormonal activity or functioning of organs) that distinguish males and females" (Health Canada, 2003, p. 8). Another definition is: “the classification of living things, generally as male or female according to their reproductive organs and functions assigned by chromosomal complement” (Wizemann & Pardue (Committee on Understanding the Biology of Sex and Gender Differences, Board on Health Sciences Policy, Institute of Medicine), 2001). Significantly, although it is customary to invoke a binary conceptualisation of sex, it is increasingly acknowledged that this may be inadequate, and that the notion of a continuum of sex characteristics may be more helpful in understanding the wide variety of human sex characteristics. Gender, on the other hand, “…refers to the socially prescribed and experienced dimensions of “femaleness” or “maleness” in a society, and is manifested at many levels” (Johnson, Greaves, & Repta, 2007, p. 5). It also has been defined as “a person’s self- representation as male or female, or how that person is responded to by social institutions based on the individual’s gender presentation. Gender is rooted in biology and shaped by environment and experience”(Wizemann & Pardue, (Committee on Understanding the Biology of Sex and Gender Differences, Board on Health Sciences Policy, Institute of Medicine), 2001). Notions of socially-determined roles and responsibilities are also included in the concept of gender (World Health Organization, 1998), as well as the relational aspect, such that comparison of one gender to the other(s) is always inferred (Spitzer, n.d.). Perhaps it is most important to acknowledge that  5 sex and gender are intertwined in their complex effects on health and illness, thereby making separation along biological and socio-cultural lines somewhat arbitrary, unrealistic, and open to criticism (Spitzer). Genetic, hormonal, physiologic and experiential factors combine throughout the lifespan to produce the phenotype we call an “individual” (Wizemann & Pardue, (Committee on Understanding the Biology of Sex and Gender Differences, Board on Health Sciences Policy, Institute of Medicine)). Nonetheless, given that the primary focus of this research was on patients’ symptoms, which by their very nature are subjective, and constructed in concert with social and cultural factors, a focus on gender, not sex, was taken. However, because there are known to be anatomic and physiologic differences between men’s and women’s cardiovascular systems that could contribute to different symptom experiences (Merz et al., 2006), and because sex and gender are part of a single system (not separate spheres) of complex interactions, differences due to sex (biological) factors were also considered when appropriate, for example with respect to co-morbidities or coronary anatomy during angiography. Accordingly, the terms sex and gender are used in this dissertation as follows. The term “sex differences” will be used to describe sex-based differences that are related primarily to biological (normal physiological or pathophysiological) traits. The term gender differences is reserved to describe differences that are affected primarily by behavioural, social or cultural influences, including self-representation as male or female, and responses of social institutions to individuals based on their gender presentation. Since it is usually impossible to know a priori whether the cause for a difference is biological or socio-cultural, the combination term “sex/gender” is employed in situations in which either a combination of biological, socio- cultural and behavioural influences is thought to be operant, or it is unclear which might predominate. This usage is in line with that of World Health Organization (1998), Health Canada (2003) and the Institute of Medicine (Wizemann & Pardue, 2001), as discussed above.  6 Context for the Study A larger research team, of which I am a member, is conducting a program of research aiming to elucidate sex and gender differences in treatment-seeking for ACS. The first study in this program examined the knowledge of myocardial infarction (MI) symptoms possessed by a representative sample of men and women living in the Greater Vancouver area, utilizing a telephone-administered survey. The number of symptoms correctly identified by respondents was measured as well as the percentage of respondents correctly identifying the cause of the symptoms, and relationships between such knowledge and socio-demographic factors, prior history of an MI, and worry about and perceived risk of having an MI, were explored (Ratner, Johnson, Mackay, Tu, & Hossain, 2008; Ratner, Tzianetas, Tu, Johnson, Mackay, Buller, Rowlands, & Reime, 2006). Further studies are underway to examine (using both quantitative and qualitative approaches to data collection and analysis) whether men and women who have had symptoms of ACS differ in their treatment-seeking behaviour, and to uncover what factors (e. g., psychosocial, clinical, cognitive or social) may be at play in producing these differences. The study described here attempted to add an important measure of objectivity to an area of investigation that has been necessarily fraught with the limitations of subjectivity. It is anticipated that these findings will contribute to the overall research program of sex and gender differences in treatment-seeking behaviour, by clarifying whether there are sex/gender differences in initial symptoms of ACS, which could prove to be a factor in the observed differences in men’s and women’s treatment-seeking decisions. Research Questions The design and conduct of this research was guided by the following principal questions: 1. What is the subjective experience of myocardial ischemia during a period of known, controlled myocardial ischemia, with respect to:  7 a) perception of any symptoms b) number of symptoms c) intensity of symptoms, and d) location and type of symptoms? 2. Are there sex/gender differences in any of the above four aspects of the myocardial ischemia experience? 3. Do certain symptoms occur in clusters? 4. What are the characteristics of members of particular groups in which symptoms occur in common clusters? Two other issues led to additional research questions. First, the research protocol entailed collection of data relating to the symptoms that had prompted referral for PCI. So, although a major strength in the design of the study was the ability to confirm ischemia, there was nonetheless a desire to avoid squandering the opportunity to explore these abundant data. Second, because there was an inherent assumption that PCI balloon inflation would provide a valid simulation of ACS from which to evaluate symptoms, it was prudent to test this assumption. Therefore, the following secondary research questions were also addressed: 5. What is the subjective experience of presumed myocardial ischemia as later recalled by patients scheduled to undergo angiogram or PCI, with respect to: a) perception of any symptoms b) number of symptoms c) intensity of symptoms, and d) location and type of symptoms?  8 6. Are there sex/gender differences in any of the above four aspects of the recalled myocardial ischemia experience? 7. Do certain recalled symptoms occur in clusters? 8. What are the characteristics of members of particular groups in which recalled symptoms occur in common clusters? 9. Is there concordance between the recalled subjective experience of presumed myocardial ischemia and the subjective experience of confirmed myocardial ischemia described during PCI?  Significance Since delays in seeking definitive treatment for ACS are associated with poorer prognosis and increased mortality, recognising ACS symptoms early, which could mitigate treatment- seeking delay, is critical to survival. A better understanding of sex/gender differences in the symptom characteristics of ACS is needed to determine if sex- or gender-specific strategies for diagnosis are necessary to improve accuracy and timeliness of diagnosis, which in turn should reduce morbidity and mortality related to ACS.  Therefore, with its unique prospective design, and other strategies that address the limitations of previous research, this study could make an important contribution to our knowledge of experiences of ACS in men and women. This study also provides invaluable evidence to inform the design of educational initiatives for both patients and health professionals regarding the recognition or diagnosis of ACS.  9 CHAPTER 2: REVIEW OF THE LITERATURE Why Study Sex and Gender Differences? Over the past several decades, scientists have developed increasing awareness that sex- based differences in biology exert an effect on physiology and pathophysiology. Undoubtedly, these have effects on health, illness and behaviour. Sex-based physiological differences that manifest as differences in health and illness between men and women include dissimilar brain functioning, which leads to differences in cognitive abilities between the sexes (such as verbal, perceptual, spatial and quantitative skills); differences in energy metabolism, which lead to different rates of endocrine disorders and obesity; and differences in chemical absorption, which lead to differences in the effects of drugs, to name only a few (Wizeman & Pardue (Committee on Understanding the Biology of Sex and Gender Differences, Board on Health Sciences Policy, Institute of Medicine), 2001). In atherosclerotic CAD in particular, there are many sex- and gender-based differences in etiology (for example: differences in exposure to toxins such as smoke; in the prevalence of familial hypercholesterolemia; in the impact of certain risks at different ages; and in the prevalence of modifiable risk factors, such as hypertension and obesity), clinical presentation (for example, women experience myocardial infarction (MI) as their initial presentation of CAD less often than do men), and prognosis (for example, women under 65 years of age have higher mortality from MI than men of the same age) (Wizeman & Pardue, 2001). Importantly, the authors of the Institute of Medicine’s (Wizeman & Pardue, 2001) comprehensive examination of issues related to the contributions of sex to health and illness concluded, among other things, that such sex differences in health cannot be explained by the influence of sex hormones alone, and that they have impacts that reach far beyond reproductive  10 health. Other possible mechanisms for the array of sex differences observed in health and illness have been explored. Among these are fetal responses to in-utero malnutrition or maternal stress, perhaps promoting later development of atherosclerosis (Barker, 2000); genetic causes for differences in drug absorption and excretion; genetic and molecular factors related to regulation of appetite, energy expenditure and causes of obesity; genetic and environmental influences on bone mass; and neuro-anatomic and physiologic differences (including the brain’s role in regulating non-sex hormones), possibly explaining differences in behaviour and perception (Wizemann & Pardue, 2001). The overarching question driving the report’s several recommendations was: “How can information on sex differences be translated into preventive, diagnostic and therapeutic practice?” (p. 14). One recommendation was based on the assertion that not only the organism as a whole, but every cell, has a sex. As such, they urged that studies of sex differences at the cellular level should be a priority for the future. Also, given the sophistication that has been achieved in the study of sex differences, they suggested that it was time to move beyond descriptive research to hypothesis-testing mechanistic studies, with the goal of translating such knowledge into positive outcomes for patients and the health care system. Another recommendation is that all health-related research should include sex as a biological variable, particularly studies of diseases that affect both sexes (e.g., CAD), and control for factors such as physiology (e.g., phase of menstrual cycle). In examining how sex affects behaviour and perception, as well as health, the report’s authors repeatedly argued that it is not sex hormones alone that are responsible for observed sex differences, and that, as suggested by Spitzer (n.d.), Johnson et al. (2007), and Knaak (2004), there is a complex, life-long, and evolving interaction between genetic, physiologic, social, cultural and environmental factors that results in the phenotype. They and others have urged  11 that research must be designed to “tease out” the respective roles of biology and environmental factors. With this context in mind, the following literature review presents studies of sex/gender differences in symptoms of ACS; studies of other (non-sex/gender) factors associated with symptoms of ACS; studies that have employed alternative approaches to analysis of symptoms of ACS; studies regarding sex/gender differences in the pain experience; and studies of sex/gender differences in symptoms that are associated with, but not specific to, ACS. The chapter concludes with a review of some of the research that has attempted to address the causes of differences. The earliest reference found in PubMed that alludes to symptoms associated with myocardial infarction (MI) and that also includes “sex factors” as a medical subject heading (MeSH) was dated 1969, entitled “On the clinical findings in myocardial infarct: Symptomatology and complications” (Reimann & Jahrmarker, 1969). Another study, from the same year, examined reasons for treatment-seeking delay, including sex/gender, and noted no sex/gender differences in this behaviour (Hackett & Cassem, 1969). Research interest in treatment-seeking behaviour has continued since the early 1970s, and by the late 1990s, this literature began to point to female sex being an important factor (Dracup & Moser, 1991; Finnegan et al., 2000; Goldberg et al., 2002; Meischke et al., 1995; Ottesen et al., 1996). Fuelled partly by overwhelming evidence of the importance of timely (early) delivery of reperfusion treatment on favourable outcomes (Chareonthaitawee et al., 2000; Fibrinolytic Therapy Trialists’ (FTT) Collaborative Group, 1994; Gersh & Anderson, 1993; Gruppo Italiano per lo Studio della Streptochinasi nell’Infarto Miocardico (GISSI), 1986; Hermens, Willems, Nijssen, & Simoons, 1992; Milavetz et al., 1998; Newby, Rutsch, Califf, & et al., for the GUSTO-1 Investigators, 1996), and perhaps also by clinical observations that women’s symptoms were different from  12 men’s, an increase in the number of studies of both women’s MI symptoms and sex/gender differences in MI symptoms occurred in the mid-to-late 1990s. A cursory bibliometric analysis reveals that from 1994 to 1999, there was an average of five papers published per year addressing sex/gender factors in the diagnosis of MI, and, beginning with a precipitous jump in 2000 to 16 papers, the average in the past eight years has been 18 per year. Certainly not all of these deal explicitly with symptoms, but the overall trend is likely reflective of trends in this specific area of inquiry. Sex/Gender Differences in Symptoms of Acute Coronary Syndromes  As a preface, a note on terminology found in the literature may be helpful. A review of terms used in 24 publications of original research focussed on sex and gender differences in the symptoms of ACS revealed the following: 11 used the term “typical”, “atypical”, or “classic” in conjunction with symptoms in the research question(s), the outcome of interest, the inclusion or exclusion enrollment criteria or the study instruments. Of these 11, only 7 defined their use of the terms, and of those, only 2 provided a reference for their definition. Eight of the 24 papers contained one or more of the terms within the discussion only, but none of these defined the terms. Further, five of six published reviews of the literature used one or more of the terms typical, atypical and classic, but only one defined them. In that review, Canto et al. (2007) refer to the widely-accepted definition of angina used in the World Health Organization’s Rose Questionnaire (Rose, 1962). Incidentally, it is interesting to note that the Rose definition and resultant questionnaire were developed after studying the descriptions of symptoms in a sample that was small by today’s standards (N = 74). These first subjects had angina, MI or chest discomfort without a cardiac diagnosis (N = 36, 15 and 23, respectively), and the validation sample for the questionnaire was 57 men. Canto and co-authors’ definition of typical symptoms,  13 which they prefaced with the statement, “the ‘typical’ (textbook) symptoms of angina are well- known” (p. 2406), also included associated symptoms, which the original Rose Questionnaire did not, but Canto et al. failed to specify whether symptoms would be considered typical if they occurred in isolation, that is, without chest pain or discomfort present. Finally, the author of one dissertation, focussing on sex/gender differences in ACS, did define the term, but did not provide a reference for the definition. Since the time when the Rose Questionnaire was introduced, there has been a trend towards developing consensus-based clinical practice guidelines within the cardiology community, usually jointly led in North America by the American Heart Association and the American College of Cardiology. Such guidelines exist for managing unstable angina/non-ST- elevation MI (Anderson et al., 2007), ST-elevation MI (Antman et al., 2004, Kushner et al., 2009), and chronic stable angina (Gibbons et al., 2003), in which the common or typical clinical presentation is described and guidelines are given for assessing the likelihood that a patient may have the condition. Similar guidelines have also been developed by the European Society of Cardiology (Bassand et al., 2007; The Task Force on the Management of Stable Angina Pectoris of the European Society of Cardiology, 2006; Van de Werf et al., 2008). It is a difficult task to muster a definition of typical symptoms from most of these guidelines. Some do not use the word “typical”, but do use the term “atypical”, which begs for at least a definition of typical. Other terms used in the documents to describe symptoms in relation to ACS are “suggestive”, “high likelihood”, “less typical”, “common”, “less common”, and “usual”. Sometimes several of these terms are used within the same guideline. Nevertheless, by gathering information from several sections of the documents, a list of symptoms that are considered typical of ACS can be created. Table 1 summarizes the symptoms that are considered typical (or equivalent) for each of the aforementioned guidelines.  14 Table 1  Summary of Symptom Classification, by Guideline Guideline Symptom ACC/AHA STEMI ACC/AHA UA/NSTEMI ACC/AHA CS Angina ESC STEMI ESC NSTEMI ESC CS Angina Chest (pain, pressure, tightness, squeezing) T T T T T T Arm(s) T T1 T2 T1 T1 T Jaw, Throat, Neck T T1 T2 T1 T1 T Shoulder   T2   T Intrascapular T T  T  T Back  T1 T2 Epigastrium/Indigest’n/Heartburn T T T2 T A-acs3, 4, 5 T SOB T1 T1 T2 C T1, 3, 4, 5 T Unexplained Fatigue/Weakness T C A-ang C  T Diaphoresis T T1 A-ang C T1 T Nausea & Vomiting/Burping T T1 A-ang T1 T1 T Dizziness/Lightheadedness/ Faintness T C  C T T ACC/AHA, American College of Cardiology/American Heart Association; ESC, European Society of Cardiology; STEMI, ST-elevation myocardial infarction; NSTEM, non- ST-elevation myocardial infarction; CS, chronic stable; SOB, shortness of breath; T, typical, suggestive, or high likelihood of ACS; A-ang, atypical of chronic stable angina; A-acs, atypical of ACS; C, common (lower category than typical implied); 1considered typical only if discomfort radiates from chest or symptom occurs in presence of chest discomfort; 2considered typical only if associate with exertion; 3stated more common in women; 4stated more common in elderly; 5stated more common in diabetic patients   15 The summary presented in Table 1 has been generalised somewhat; for example, in some sections of the AHA/ACC guidelines, it is stated that associated symptoms are only considered typical in the presence of chest discomfort, but in other sections this condition is not stipulated. There is perfect consensus across the six guidelines for only one symptom: chest discomfort. Fairly high consensus is present for arm, jaw, throat, or neck discomfort. Consensus is clearly not present for any other symptoms, which may explain why only 1 of the 30 works mentioned previously (24 individual studies and 6 reviews) cited any of these six guidelines in their definition of typical. It is extraordinary and unfortunate that a common definition of typical symptoms could not be devised and agreed upon from these guidelines, since they are consulted and followed for almost all aspects of clinical care for ACS patients. This gap likely exemplifies the current poor understanding of what constitutes typical symptoms. For the remainder of this review, I use whichever term was employed by the author, and define the term if a definition was supplied. Published Reviews of the Literature There have been six reviews of the literature specific to sex and gender differences in symptoms of ACS and MI, published in 1999, 2002 (2), 2004, 2005 and 2007. I provide an overview of only three of these, since one (Patel, Rosengren, & Ekman, 2004) was eventually exposed (Califf & Mark, 2005) as containing substantial content that had been plagiarised from DeVon and Zerwic’s (2002) review, and the other two (Herlitz, Bang, Karlson, & Hartford, 1999; Kyker & Limacher, 2002) did not review any studies focusing on symptoms of ACS that had not been included in the three that are reviewed below, and reached essentially the same conclusions. DeVon and Zerwic (2002) reviewed 12 studies published between 1989 and 2000. The authors argued that it was necessary to include studies of both ACS and MI symptoms because  16 women have fewer MIs compared with men, so they would tend to be underrepresented in studies of MI only. Their sample included five studies of ACS patients and seven studies that included only MI patients. Overall, the findings were that men and women experience the same symptoms, but not in the same proportion. Notably, rates of chest pain did not differ between men and women in 10 of the 12 studies, although two studies of MI patients did find significantly lower rates of chest pain among women than among men. Most other symptoms were experienced more often by women than by men, except diaphoresis, which was experienced more often by men. Some of these symptoms would commonly be classified as typical (e.g., jaw pain, dyspnea), and some as atypical (e.g., back pain, unexplained shortness of breath, indigestion, cough, palpitations, weakness, dizziness and fatigue). Importantly, where differences were found, the direction of the difference remained the same across all of the studies. Although not all of the studies identified sex or gender differences, the authors noted that the wide variability in the groups studied, geographical settings, and methodologies may explain the disparity. Also, statistical control for differences in age, ethnicity, co-morbidities and diagnoses was often not obtained. This review was somewhat superficial in that it did not thoroughly critique the methods of the reviewed studies. Nonetheless, the authors concluded that the total picture of symptom expression, as opposed to inventories of individual symptoms, is important, and called for further study to identify whether patterns exist.  Chen et al. (2005) reviewed 11 studies published between 1989 and 2002. The studies included were those that examined symptoms of MI only, and only those that compared men’s symptoms with women’s. Of the 11 studies, 7 relied on reviews of medical records and 4 relied on open-ended interviews or structured questionnaires. Only one study failed to find any sex/gender differences in symptoms, but it was criticised for having a considerable selection bias, because it included only participants 75 years of age or younger, who had had symptom  17 duration of less than six hours. It is well known that women are often older than 75 years when presenting with an MI (Maynard & Weaver, 1992), and most researchers have concluded that they delay seeking treatment longer than do men (Dracup & Moser, 1991; Gibler et al., 2002; Lefler & Bondy, 2004; Meischke, Eisenberg, & Larsen, 1993; O'Donnell, Condell, Begley, & Fitzgerald, 2006; Ottesen et al., 1996), so these inclusion criteria would have effectively excluded “typical” women.  The remainder of the studies included in this review had variable findings with respect to symptom expression. The majority of the studies showed that women were more likely than men to report shortness of breath, nausea, vomiting, back, jaw, neck, arm or shoulder pain, or fatigue. Some, but not a majority, found that women were less likely to report chest pain or chest discomfort as their chief complaint, and most studies found that women were less likely to report diaphoresis. The authors suggested that reasons for this jumble of findings include reliance on retrospective reviews of health records; small sample sizes; the practice of collapsing reports of pain radiating to adjacent areas into chest pain; and exclusion of both patients without chest pain and patients over 75 years old. Some studies used the now-outdated World Health Organization criteria for the diagnosis of MI (Goldberg et al., 1998), which required at least two of the following three findings to be present: typical symptoms (e.g., chest pain); ECG changes consistent with ischemia (e.g., development of Q-waves); and elevated cardiac enzymes. Excluding women with atypical symptoms or non-Q-wave MI, the latter of which has been shown to be more prevalent in women than in men (Goldberg et al.), would potentially result in fewer truly representative female participants. This more thorough review concluded with a call for more research to confirm the existence of sex/gender differences, standardisation of symptom lists, prospective research designs, and more sophisticated analytic approaches.  18  Most recently, Canto et al. (2007) published a comprehensive review of 69 studies examining sex/gender differences in symptom expression of patients presenting with ACS. The review encompassed publications from 1970 to 2005, and also included studies that reported prevalence of unrecognised MI. Research addressing stable coronary disease, prodromal symptoms, and studies that did not focus primarily on symptom presentation were excluded. The authors did not undertake a meta-analysis, given the heterogeneity of study designs and methods, but confined their main analysis to a comparison of the proportion of participants without chest pain or discomfort. Amongst the nine large-cohort studies, all but one showed that a larger percentage of women than men did not present with chest pain or discomfort. The small-cohort and interview-based studies mirrored this finding: a greater percentage of women than men presented without chest pain in 19 out of 20 studies. The cumulative proportion of subjects without chest pain in women and men for the large cohorts was 37% and 27%, respectively, and 30% and 17%, respectively, in the smaller studies. However, it is important to note that overall, about one third of all subjects in the large-sample studies and one quarter in the smaller studies presented without this symptom, thereby representing a substantial number of men as well. The relative frequency of other symptoms (e.g., other locations of pain; and other non-pain symptoms such as nausea and vomiting, shortness of breath, indigestion, weakness or fatigue, or dizziness) was higher in women than in men, with the mean number of symptoms reported across all studies being 2.6 for women versus 1.8 for men. The occurrence of silent MI (having no symptoms) or unrecognised MI (not recognised as MI by a physician, which included both patients with no symptoms and those with symptoms other than chest discomfort that were unrecognised as cardiac in origin by either the patient or the physician) was similar in women and men, though MI was often not validated with biomarkers in these studies.  19 In this review, Canto et al. (2007) expressed concern that several investigators did not adjust statistically for differences in age, especially since women are usually older than men when hospitalised for MI. One study of 434,000 patients that did control for age (Canto et al., 2000) found that increasing age (10-year increments) was associated more strongly with absence of chest discomfort (odds ratio [OR], 1.36; 95% confidence interval [CI]: 1.34-1.37) than was female sex (OR, 1.09; 95% CI: 1.06-1.11). In three of the studies reviewed, the sex/gender difference in presenting without chest pain that was initially apparent disappeared when the investigators controlled for age. In one of these studies, both age of greater than 75 years and sex/gender were predictive of the absence of chest discomfort. Thus the authors concluded, as had previous reviewers, that this body of literature suffered from several limitations. These will be examined in detail, since they are common in this body of literature. First, the lack of standardisation in the method and content of data collected makes comparison across studies virtually impossible, and may affect the findings obtained. There are two main ways in which data have been typically collected for these studies: through the participants’ self-reports and through review of health records. It should be noted that most, if not all studies to date that have used the self-report method have entailed retrospective, not concurrent accounts of symptoms. Within the self-report approach there are several variations, and Shin, Martin and Howren (2009) noted that the way symptom data are collected affects what participants report. In their study comparing open-ended questioning to a combination of open- ended and closed-ended questioning (using a list of symptoms), significant differences between men’s and women’s symptom reports were found in the combination approach that were not evident in the data from open-ended questioning alone. In particular, men reported chest pain significantly more often than women in response to the combination method, but differences were not apparent when only open-ended questioning was employed. The study authors had a  20 tendency to conflate the clinical significance of this finding somewhat with the methodological significance; it is acknowledged, however, that, although they are different issues, they are certainly important for both domains. Nonetheless, for the purpose of this chapter, the methodological issue is of most relevance. The authors suggested that different methods of data collection may account for the inconsistency in findings regarding sex/gender differences in symptoms of ischemia, to date, but did not offer an explanation as to why there might be differential responses. Data retrieved from health records suffer from at least two limitations: there is no way to know exactly how the clinician obtained the information (i.e., what questions and how much probing were used?; were questions open- or closed-ended?); and there is no way of knowing if the clinician recorded all the information the patient gave. DeVon, Ryan, and Zerwic (2004) investigated this exact methodological dilemma in an attempt to quantify the agreement between data in the health record and data obtained through in-depth interviews. They found that the health record is an unreliable source of data for symptoms of ACS, with agreement for chest discomfort reasonable at 87% (kappa = 0.40), but for all other symptoms below 80%, and as low as 36% (kappa = 0.01) for fatigue. These authors acknowledged that there can be recall bias during interviews held sometime after an acute event; nonetheless, they concluded that  health records are a poorer source of data for ACS symptom research. Retrospective designs, as mentioned above, may necessitate reliance on the health record, but even when patients themselves are the source of data, their poor recall and other contextual influences on ability to report, such as pain or anxiety, create problems.  Other limitations in this body of research noted by Canto et al. (2007) were: questionable collapsing of symptom variables; exclusion of those who did not have chest pain; and the inability to test the specificity of a particular symptom when only participants with a particular diagnosis were  21 included. They also recognised that the influence of co-morbid conditions, which had often not been addressed, deserves further exploration. Although the Canto et al. review (2007) suggested that women more often present without chest pain, the authors declined to recommend a change in public or professional messages regarding the symptoms of ACS because the current messages encompass all of the common symptoms. Other Studies Published from 2004-2009 Using the Medical Subject Headings (MeSH), “myocardial ischemia/diagnosis” and “sex factors”, and limiting to years 2004-2009, 10 additional published reports were retrieved that had not been included in any of the foregoing reviews (see Table 1). Of these publications, one study included patients with unstable angina, five studied patients with MI, and the remaining three studies and one report of a large, prospective registry included patients along the entire continuum of ACS (UA, STEMI and NSTEMI). All were prospective and utilised interviews and questionnaires, or in the case of the registry, routine history-taking and review of the active clinical record, to obtain data. As with some past studies, four found that fewer women reported chest discomfort than did men (Arslanian-Engoren et al., 2006; Granot, Goldstein-Ferber, & Azzam, 2004; King & McGuire, 2007; Lovlien, Schei, & Hole, 2006), but six publications reported no difference between men and women in the frequency of reporting chest discomfort (DeVon, Ryan, Ochs, & Shapiro, 2008; Dey et al., 2009; Kosuge et al., 2006; Lovlien, Schei, & Gjengedal, 2006; Omran & Al-Hassan, 2006; Thuresson et al., 2005). And, similar to previous studies, several reported that more women reported non-chest discomfort than did men, although one study found that men reported left arm pain more often (Arslanian-Engoren et al., 2006) and two had equivocal findings concerning indigestion (DeVon et al., 2008; King & McGuire, 2007). Of the four studies that examined the number of symptoms reported, three  22 found no significant differences, and one found women to report more symptoms than did men. Finally, six of the ten studies presented data about the intensity of reported symptoms: three showed that women reported greater intensity than did men, one found no statistically significant difference, and two did not test for differences. Although a picture of women’s symptoms of ACS is beginning to appear, both from these more recent studies and from the foregoing reviews, it is not yet a clear one, and our knowledge of sex/gender differences in ACS is still lacking. Including the reviews by DeVon and Zerwic (2002) and Chen et al. (2005), and the individual studies, overall, about 70% of the studies (total N = >511,000) reported no difference in the frequency of reporting chest pain. Canto and colleagues’ (2007) large review is not included in this summary statistic because they did not test for differences in frequencies between studies. This proportion is not definitive enough for one to conclude that the question is answered. As well, findings regarding other non- chest symptoms are varied. More sophisticated analyses are beginning to implicate age and co- morbidities to an extent that cannot be ignored; however, since such analyses were only applied in one half of this latter group of studies, it remains to be seen how much variance in reported symptoms can be explained by these factors, and in what patient groups.  23 Table 2  Summary of Studies Examining Sex/Gender Differences in ACS or MI, 2004-2009 Author (Year) N (% Women)/ Population Chest Discomfort (Frequency/Intensity) Non-Chest Discomfort Non-Pain Symptoms Adjusted for Covariates Other Granot (2004) 61 (47)/UA Women < men/↑ intensity in women Overall more in women (e.g., stomach, back, chin) More women with dyspnea, palpitations, irritability, dizziness, nausea No Women were more often vague/couldn’t describe at all Thuresson (2006) 1939 (25)/all ACS No difference/↑ intensity in women More women with neck, jaw, back More vomiting, anxiety, fatigue, weakness in women Yes More anguish, fear Omran & Al- Hassan (2006) 83 (31)/MI No difference/N/A  Both with weakness, sweating, nausea & fatigue, but different frequency No Arslanian- Engoren (2006) 1941 (35)/all ACS Women < men/N/A Less left arm in women Less diaphoresis; more nausea Yes Sex/gender predicted diaphoresis (men) & nausea (women); no sex/gender diffs when controlled for co- morbidities & age Note: UA = unstable angina; ACS = acute coronary syndrome; STEACS = ST-elevation ACS; MI = myocardial infarction; STEMI = ST-elevation myocardial infarction; NSTEMI = non-ST-elevation myocardial infarction; N.A = not available   24 Table 2 (cont’d) Summary of Studies Examining Sex/Gender Differences in ACS or MI, 2004-2009 Author (Year) N (%Women)/ Population Chest Discomfort (Frequency/Intensity) Non-Chest Discomfort Non-Pain Symptoms Adjusted for Covariates Other Lovelien, Schei, & Gjengedal (2006) 82 (46)/MI No difference/N/A More women with arm, back, jaw, throat More nausea No Lovelien, Schei & Hole (2006) 533 (28)/MI Women < men/NA More women with back, jaw More nausea, dyspnea No >76 yo excluded; high prevalence HTN in those with no chest symptoms Kosuge (2006) 457 (23)/STEMI No difference/NA More women with jaw, throat, neck, back, L & R arm, L shoulder More nausea, less diaphoresis No More often vague characterisations King (2007) 60 (50)/MI Less R-sided  & central Less indigestion  Yes Note: Statements relate to women unless otherwise indicated. UA = unstable angina; ACS = acute coronary syndrome; STEACS = ST-elevation ACS; MI = myocardial infarction; STEMI = ST-elevation myocardial infarction; NSTEMI = non-ST-elevation myocardial infarction; N/A = not applicable  25 Table 2 (cont’d) Summary of Studies Examining Sex/Gender Differences in ACS or MI, 2004-2009 Author (Year) N (%Women)/ Population Chest Discomfort (Frequency/Intensity) Non-Chest Discomfort Non-Pain Symptoms Adjusted for Covariates Other DeVon (2008) 256 (44)/ACS No difference/N/A None STEMI: more women with indigest’n, nausea, palpitations, hand numbness, fatigue; dizziness; no STEMI: more women with weakness & cough Yes ↑ Intensity in atypical symptoms Dey (2009) 26,755 (29)/ACS No difference/N/A More women with jaw More women with nausea/vomiting Yes Note: Statements relate to women unless otherwise indicated. UA = unstable angina; ACS = acute coronary syndrome; STEACS = ST-elevation ACS; MI = myocardial infarction; STEMI = ST-elevation myocardial infarction; NSTEMI = non-ST-elevation myocardial infarction; N/A = not applicable    26 Causes of Sex/Gender Differences in Symptoms of ACS  The majority of the studies examined were observational, not mechanistic in their objectives, that is, they did not attempt to uncover possible explanations for any sex/gender differences in symptoms. However, some researchers have attempted to address this gap. The body of research related to sex and gender differences in pain intensity and in reporting specific locations of pain is described fully in a later section, including a discussion of the evidence for various proposed mechanisms.  In terms of cardiac symptoms, there is a growing body of research elucidating sex/gender differences in risk factors and the pathogenesis of ischemic heart disease, but there has been less specific focus on possible causes of observed sex/gender differences in symptoms. Nevertheless, it is widely accepted that pre-menopausal women have a lower prevalence of obstructive CAD than men do (Shaw et al., 2006), becoming equal to men’s only around the seventh decade of life. The lower prevalence is possibly due to differences in atherosclerotic storage, which results in less intrusive plaque (Merz et al., 2006). This has obvious and significant implications for the diagnosis of CAD in women, since the dominant diagnostic algorithms are specifically designed to detect obstructive CAD (e.g., stress-testing, coronary angiogram) and lead to therapy to mitigate coronary obstruction (percutaneous and surgical revascularisation). The lower prevalence of obstructive CAD among women has prompted some researchers to investigate subendocardial (as opposed to epicardial) ischemia (Panting et al., 2002) and alterations in myocardial metabolism (Buchthal et al., 2000) as explanatory mechanisms for the sex differences in the pathophysiology of ischemic heart disease. A diminished pool of endothelial progenitor cells, which function as a vascular “repair kit”, exists in some pre- and many post-menopausal women, and has also been suggested as a mechanism for the development of coronary atherosclerosis (Quyyumi, 2006). While these studies were not   27 designed to evaluate reasons for differences in symptoms, it seems plausible that a qualitatively different disease might also affect its symptom profile, though exactly how is not clear, as yet. The Women’s Ischemic Syndromes Evaluation group, in one of their reports of insights gained from their work (Shaw et al., 2006), also referred to the complex inter-relationships between sex hormones, metabolism, inflammation and vascular (dys)function in the pathogenesis of ischemic heart disease: Emerging data suggest a unique risk profile in women, including hypoestrogenemia coupled with the adverse effects of a protracted dysmetabolic state on promoting an inflammatory milieu and/or vascular or metabolic alterations that may provoke both symptoms and ischemia in the setting of non-obstructive CAD. (p. 15S)  In a study that addressed reasons for the differences between men and women in the symptoms of ACS, the relationship of ischemic symptoms to the phase of the menstrual cycle was examined (Methot et al., 2004). The investigators found that atypical presentation was more common amongst pre-menopausal women than amongst post-menopausal women, regardless of whether they were taking hormone replacement therapy. Possible physiologic reasons for sex differences in reported pain intensity have been postulated. Most of these relate to the multiple and, as yet poorly-understood, effects of sex hormones on pain (e.g., the hormone-dependent prevalence of certain types of pain; effects on the inflammatory response and thereby nocioception; and effects on endogenous opioids) (Fillingim, King, Ribeiro-Dasilva, Rahim-Williams, & Riley, 2009; Unruh, 1996). Causes of sex/gender differences in non-chest symptoms have also been explored, for example, the higher rate of throat and jaw pain reported by women. Several investigators have shown that women have a higher prevalence of temporomandibular joint pain and tooth, jaw and other types of orofacial conditions (Fillingim et al., 2009). It has been suggested that sex hormones are influential in this, since these differences are not apparent until after girls reach   28 puberty. However, since these conditions are primarily musculoskeletal in origin, this may not necessarily be a valid explanation for cardiac ischemic pain being referred to this area. Neuro- anatomical studies and interventions, though, have revealed that the neck and jaw region is innervated primarily by the vagus nerve (Foreman, 2007; Mork, Haufe, & Yancey, 2004), and there is also evidence to suggest that women have higher cardiac vagal activity than do men (Chambers & Allen, 2007). Some studies have attempted to identify possible psychosocial and behavioural causes for differences in ACS symptoms. Shin et al. (2009) found that the method of data collection used (i.e., checklists versus open-ended questions) affected what symptoms were reported. In particular, when open-ended questioning was employed, the frequency of women reporting chest discomfort increased (thus eliminating gender differences), compared to using only checklists. The investigators speculated that prescriptive checklists discourage reporting of symptoms that are not on the list, which would be likely to be those that had been deemed to be atypical. Finally, women’s tendency to report higher pain intensity has been shown to be related to socially-prescribed gender roles (including those of the clinician assessing the patient), higher levels of anxiety, higher prevalence of depression, and different coping mechanisms (Fillingim et al., 2009). Many questions remain to be answered in the area of the risks for, and pathophysiology of, ischemic heart disease in women. Still more are unanswered as to the causes of sex/gender differences in symptoms of ACS, but it is likely that at least some progress will be made in the latter as progress is made in the former. Other Factors Associated with Symptoms of Acute Coronary Syndrome While not the purpose of this study to examine them, it is important to note that characteristics other than, or in addition to sex/gender have been suggested as being possibly   29 associated with differences in ACS symptoms. Age has been evaluated as an important covariate in many studies, but only two studies were found that explored the association between age and ACS symptoms as a stated main purpose. Diabetes, ethnicity and race have also received some research attention as covariates of principal interest. A review of these studies follows. Age Stern et al. (2004) studied the symptoms of ACS patients in relation to age, admission ECG, medical management, and outcomes. The main finding with respect to symptoms and age was that a “no-anginal pain or atypical” presentation was significantly more prevalent with increasing age (divided into three groups: less than 65 years, 65-74 years, and 75 years or greater), even after adjusting for sex/gender. Diabetes Diabetes has garnered research attention because an association has been demonstrated between silent ischemia and diabetes (Langer, Freeman, Josse, Steiner, & Armstrong, 1991). Silent ischemia, in turn, is associated with poorer outcomes because of the inherent delay in seeking medical treatment (Chico, Tomas, & Novials, 2005). One mechanism that may explain the increased incidence of silent ischemia among patients with diabetes is the co-existence of cardiac autonomic neuropathy. This is a form of diabetic neuropathy that results in damage to the autonomic fibres that innervate certain cardiac structures, thereby affecting various cardiac functions, especially heart rate variability, and perhaps the transmission of pain messages (Manzella & Paolisso, 2005). Although there have been no definitive studies, it has been speculated, based on the time since diagnosis before microvascular changes develop in diabetic patients, that longer duration of diabetes is associated with a higher prevalence of cardiac autonomic neuropathy (DeVon, Penckofer, & Larimer, 2008).   30 Stephen, Darney, and Rosenfeld (2008) undertook a comprehensive review of the literature related to symptoms of ACS in women with diabetes. With their inclusion criteria of being published in English; comparing women and men; and comparing patients with diabetes and those without, relatively few studies (eight) qualified for the review. Overall these studies had inconsistent findings with respect to chest pain and non-chest pain: some found differences attributable to diabetes or sex/gender (only one study analysed a sub-sample of women with diabetes), and some did not. However, some investigators did report differences in associated symptoms and symptom characteristics: shortness of breath (SOB) was found to be a common ACS symptom in women with diabetes, and diabetes predicted atypical (and often painless) presentations by women, in some of the studies.  The authors concluded by calling for further research because of several noted limitations in the research, namely: selection bias (often including only patients with chest pain), narrow inclusion criteria (e.g., only MI), absence of control for co-morbid conditions, inclusion of patients with a history of CAD (as opposed to including only those experiencing the first episode of the illness), and absence of control for duration of diabetes. Both age and diabetes were examined for their association with symptoms of ACS by DeVon, Penckofer, and Larimer (2008). Their findings were that both diabetes and age were predictive of the absence of chest pain, though diabetes was a stronger predictor (ORdiabetes 0.46; 95% CI: 0.22-0.94) vs. ORage 0.96; 95% CI: 0.93-0.99). Participants with diabetes were also more likely to report unusual fatigue and shortness of breath. Ethnicity and Race1 Teoh, Lalondrelle, Roughton, Grocott-Mason & Dubry (2007) also studied the effect of  1 Race and ethnicity, while not synonymous, do overlap. However, use of these terms in health research has been inconsistent, and has been discouraged (Bhopal & Donaldson, 1998), in favour of more precise descriptions of comparator groups. For the purpose of this section of the review, however, I have elected to use the combined term “ethnicity/race”.   31 diabetes on the expression of ACS symptoms, although the main focus of their study was to compare Asian patients’ symptoms with that of a reference group, termed “white” by the authors. This large-sample study (N = 2905) demonstrated that both patients with diabetes and “white” patients had more frequent episodes of silent ischemia than did their counterparts. The investigators also reported that Asians more often had back discomfort and described a larger area and higher intensity of discomfort than did “white” people. They noted that this pattern of symptoms could understandably lead to investigation for a ruptured aortic aneurysm, thus delaying initiation of a reperfusion strategy. Investigators from the United States have compared African-American and Mexican- American patients with ACS, primarily with respect to treatment-seeking behaviour, not symptoms. Two studies comparing these groups’ reported symptoms were retrieved, however. Meshack et al. (1998) included 589 patients with confirmed or suspected MI in their study, and controlled for diabetic status and sex/gender, as well as self-reported Mexican-American ethnicity/race. More Mexican Americans reported chest and upper back pain than did participants of non-Hispanic ethnicity/race, as well as palpitations; arm pain or jaw pain were reported less frequently. Stratifying by sex/gender revealed that women were less likely to report chest pain, but more likely to report several other symptoms (e. g., dizziness, dyspnea, fatigue, upper back pain and palpitations). Increasing age predicted a lower likelihood of reporting many symptoms, whereas diabetes was not a significant predictor of any symptoms. The investigators concluded that although there were differences in symptom presentation between these two ethnic/racial groups, sex differences may be of greater importance. Although they did not include the importance of age in their conclusions, their findings also identify age as an important predictor.   32 Lee et al. (2000) studied the interaction of clinical symptoms and ethnicity/race (African- American versus “white”) on pre-hospital delay in seeking treatment for MI, with a sample of 128 MI patients (32% African-American). In terms of presenting symptoms, dyspnea and fatigue were reported significantly more often by the African-Americans than by the comparator group, termed “whites”. Though it is not germane to the particular purpose of the study described in this dissertation, it is important to the overall issue of treatment-seeking to note that the investigators found a significant interaction effect between ethnicity/race and symptoms on pre- hospital delay. However, they did not report the relationship between the type of symptoms experienced and length of pre-hospital delay. Without knowing exactly what symptoms they were responding to, it is difficult to draw conclusions about the delay. Richards, Funk, and Milner (2000), in a study of ACS patients, also explored ethnic/racial differences in symptom presentation and time to seek treatment. In their sample of 231 people (17% African-American), African-Americans were more likely to report shortness of breath and left-sided chest pain, but unlike Lee et al.’s (2000) study findings, no differences in pre-hospital delay were found. However, as in Lee and colleagues’ report, an analysis of the association between the type of symptom and pre-hospital delay was not reported. Although these studies of other clinical and socio-demographic factors are provocative, they do not fully explain or discount the sex/gender differences noted in other studies. Furthermore, none of the latter three studies of ethnicity/race presented a predictive model that included both sex/gender and ethnicity/race, so that each factor’s contribution to symptom reporting could be compared and understood. Alternative Approaches to Symptom Analysis  A different analytic approach to symptoms of MI was taken in a study by Ryan et al. (2007), who pooled data from nine studies of MI symptoms and treatment-seeking behaviour.   33 Only studies that employed face-to-face interviews for data collection were included, yielding a total sample of 1,073 patients. The investigators argued that MI symptoms do not occur in isolation, and, moreover, that patients evaluate a constellation of symptoms to make decisions about what is happening and whether to seek help. Therefore, they set out to determine whether there are identifiable “clusters” of symptoms in patients experiencing MI, and to examine relationships of those clusters with socio-demographic variables. Using latent class analysis, they found that five classes or clusters emerged; interestingly, none of the five contained all of the typical symptoms. In fact, in one cluster, participants had only moderate to low probability of having any common symptoms. The findings were consistent with previous research in that clusters with the most men had the least neck, jaw, or back pain, or nausea, vomiting, shortness of breath, and fatigue. However, one of these predominantly male clusters also had only a moderate chance of having chest pain, which is inconsistent with previous findings. Age, sex, and race proved to be significant predictors of cluster membership. Further analysis of the relationships of the clusters with the intensity of the symptoms revealed that those less than 65 years old, women, and those of non-white racial origin experienced more severe intensity of symptoms. Although this delineation of classes is provocative, the taxonomy is somewhat contrary to traditional “wisdom”, making it more difficult to easily appreciate its application to clinical practice. The authors called for further research to validate their number and symptom content of clusters, and their relationships with demographic and other clinical variables. The foregoing published reviews and the additional review of more recent literature regarding sex/gender and other differences in the symptoms of ACS indicate that there is still not certainty whether women experience chest pain as a presenting symptom less often than do men (although adjustment for age diminished sex/gender differences in many studies), but that women do seem to experience a greater number of symptoms than do men. However, the   34 limitations of many of the reviewed studies point to three recommendations, which the study described in this dissertation attempted to incorporate. First, there must be statistical control for age and co-morbidities, since these are emerging as important explanatory variables. Second, prospective designs must be used to counter the problems of poor or biased patient recall and incomplete documentation. Finally, clustering patterns of symptom expression must be explored, so that a picture of the total symptom experience can be painted for the public and professionals alike. When studying symptoms of ACS, the symptom of pain is inevitably encountered. Therefore, to inform the inquiry into sex/gender differences in ACS symptoms, a review of literature related to sex/gender differences in pain experiences is instructive. Sex/Gender Differences in Pain Experience An extensive critical review of recent clinical and experimental findings was recently published by Fillingim et al. (2009). Within their review of over 450 papers, several sections of the review had particular relevance to the clinical problem of symptoms of ACS, and will be summarized, in turn. Sex Differences in Clinical Pain  Overall most, but not all, epidemiological studies reported before the Fillingim et al. (2009) review pointed to a higher prevalence of pain in women than in men. Fillingim and colleagues’ review summarised more recent studies in several specific clinical problem areas (e.g., cancer management, neuropathic pain, musculoskeletal pain, headache), but not ischemic cardiac pain. However, it is useful to understand that recent findings continue to indicate that women are at increased risk for several types of chronic pain conditions, and also that they report higher levels of acute pain in relation to procedures.   35 Sex/Gender Differences in Responses to Experimental Ischemic Pain Stimuli  Of possible, but likely limited, relevance to ischemic cardiac pain, some experimental studies have employed an ischemic pain stimulus, most often using a tourniquet on the arm. However, the findings in terms of threshold (when pain is perceived), tolerance (how much pain is considered bearable) and pain ratings (what intensity is reported) have mostly shown no differences between men and women. Temporal Summation  Temporal summation (sometimes called “windup”) refers to increasing nociception (pain perception) with repeated exposure to a painful stimulus. Most experimental studies of this phenomenon have demonstrated that women exhibit greater temporal summation, though this has not been studied in a clinical setting. Anxiety has been suggested as a mediator of this effect. Proposed Underlying Mechanisms Fillingim et al. (2009) reviewed three main categories of proposed mechanisms for the observed sex and gender differences in pain responses: hormonal, mediation through N-methyl- D-aspartic acid (NMDA) receptors, and psychosocial mechanisms. They cautioned against the overly simplistic tendency to dichotomise these mechanisms as either biological or psychological, offering as an example stereotypic sex roles, which may seem like a psychosocial factor, but which are also heavily influenced by neurobiological processes. That caution is noted, but the dualistic classification is used for convenience. Hormonal influences reach beyond the reproductive system, and their effects on pain perception and responses have been studied extensively, using such designs as comparing pre- pubescent with post-pubescent pain prevalence, or the pain perceptions of transsexual patients before and after undergoing hormonal treatment. However, the effect of endogenous sex hormones appears complex, in that both waxing and waning levels of some hormones have been   36 shown to increase the prevalence of reported pain. Likewise the menstrual cycle often has been shown to be associated with variations in pain perception, usually increasing in the peri- menstrual phase, but some studies have not borne this out. Examination of the effect of exogenous hormones on pain has also shown differences in both directions. Sex hormones appear to influence the inflammatory response, with women generally displaying an increased inflammatory response and a correspondingly increased risk for various inflammatory conditions (e.g., rheumatoid arthritis, lupus, inflammatory bowel disease). This is important because inflammation lowers nociception thresholds. The influence of hormones on inflammation is complex, but acknowledged by these reviewers to be a potentially important mechanism to consider in relation to sex differences in response to pain. Sex hormones also modulate several endogenous mechanisms that are involved with pain processing, such as endogenous opioid systems, dopamine and serotonin, which could result in different nocioception between men and women. NMDA receptors, which are present in the central nervous system, enhance pain perception when activated. These too are affected by sex hormones, in particular estrogen, which researchers have argued may account for higher sensitisation to pain in women, and also temporal summation. The psychosocial mechanisms that may underlie gender differences in pain experience which have received the most research attention (and were therefore included in the review by Fillingim et al., [2009]), are gender roles and cognitive/affective variables such as coping and affective distress (e.g., anxiety and depression). The feminine role norm dictates that pain be accepted as a part of normal life, and seen as “…a monitor of health as well as a potential symptom of injury, illness or disease” (Unruh, 1996, p. 157). Studies reviewed by Fillingim et al. also demonstrated that the feminine norm   37 permits more open expression of pain, whereas the masculine norm includes greater tolerance or suppression of pain. Instruments to measure gender roles have been developed, including some that are specific to pain (e.g., the Gender Role Expectations of Pain) and used to explore this variable as a possible correlate of experimental pain responses. Most studies have found that pain tolerance, but not threshold, is increased in those who have higher masculinity relative to femininity scores. The gender of the experimenter also has been shown to affect the responses of men (reporting less intense pain to female experimenters than to male experimenters), but not of women; no clinical studies of this phenomenon (i.e., differing patient responses based on the gender of the health care professional to whom they were reporting) were reported in this review. Research has shown that women employ different and more numerous coping strategies than men do, including more catastrophising (imagining and dwelling on the worst possible outcome when one encounters a situation in which something unpleasant could happen), which in turn has been shown to mediate sex differences in pain, and higher use of emotional and social support. High levels of anxiety also are known to be related to increased clinical pain, and women have been shown in several studies to report higher anxiety levels compared to men. More recent studies, however, have indicated that the association of anxiety with pain is stronger for men than for women. Finally, depression and pain have been frequently linked in research reports, and the prevalence of depression is greater in women than in men. These facts notwithstanding, there has been no conclusive research on whether the effect of depression on pain perception is different for men and women. The pertinence of the exhaustive review by Fillingim et al. (2009) to the current study can be summarised as follows: women have a higher prevalence of most common forms of pain; evidence is mounting for many biological and psychosocial factors that may explain such sex and   38 gender differences, especially sex hormones, differing gender roles, and several cognitive and affective variables. However, a full explanation of these differences is not yet obvious. Sex/gender differences in other, less specific symptoms (that is, those whose occurrence is not limited to ACS, have also been explored. Sex and Gender Differences in Non-Cardiac-Specific Symptoms Shortness of Breath  Most of the literature on sex/gender differences in shortness of breath (SOB) was derived from studies of patients with respiratory disease. Conclusions about pathophysiological causes and effects are therefore of some, but limited, relevance to patients with ACS, since the underlying pathophysiology of SOB in ACS is more likely to be cardiac in nature (at least partly). Nevertheless, it is interesting to note that among respiratory patients, SOB is reported by more women than men, and this difference is not fully explained by the fact that women normally have smaller lung volumes (Becklake & Kauffmann, 1999). This has led to speculation about other causes, such as psychosocial and cultural factors affecting the perception of breathlessness. As examples, Becklake and Kauffman pointed out that depression is linked to increases in reporting of respiratory symptoms, including SOB, and women are known to have a higher prevalence of depression than men. They further suggested that the high value placed on athletic prowess in many cultures may negatively affect the likelihood of men reporting SOB. Patients with heart failure, though, are perhaps more relevant to the ACS population, and it has been observed that more women with heart failure report SOB than do their male counterparts (Riegel et al., 2003). It has also been reported that more women develop heart failure after MI than do men (Vaccarino, Parsons, Every, Barron, & Krumholz, 1999). Taken together, these observations may provide a possible explanation for the higher observed rate of women with ACS reporting SOB, compared with men, in many studies.   39 Diaphoresis  Sex and gender differences in diaphoresis have also been studied from non-cardiologic perspectives. Some studies relate to specific autonomic diseases, but those studying “normal” subjects seem to focus on the concept of thermoregulation during exercise or exposure to heat. Three studies (Hazelhurst & Claassen, 2006; Inoue et al., 2005; Stern, Salzer, Schuch, & Hornstein, 1998) demonstrated higher sweat production in “normal” men compared with “normal” women, and Inoue et al. also noted that women relied more on vasodilation for thermoregulation than men. Nausea and Vomiting  The incidence of post-operative nausea and vomiting has been found to be higher in women than men by several investigators (cited in Buchanan, Myles, & Cicuttini, 2009), and these differences are attenuated after the age of 50 years, suggesting that female hormones may be a mediating factor. Further evidence for this hypothesis has accumulated as studies of post- operative nausea have demonstrated variation by the stage of the menstrual cycle (Buchanan et al.). It is possible that post-operative nausea and vomiting can be extrapolated to nausea and vomiting associated with other conditions, such as the myocardial ischemia that is present during ACS. In addition to the findings in the anaesthesia literature, Airaksinen, Ikaheimo, Linnaluoto, Tahvanainen & Huikuri (1998) reported evidence suggesting increased vagal tone in women compared with men, during coronary occlusion caused by angioplasty balloon inflation. They found decreased heart rate and blood pressure amongst women, as opposed to increases in these indices in men, as well as increased ventricular ectopy in men, as opposed to no increase in women. These physiological responses point to increased vagal tone, and might be related to the increased tendency towards nausea and vomiting amongst women.   40 Although this brief survey of sex and gender differences in other, less specific symptoms is not exhaustive, there is a strong suggestion that female sex hormones may play a role in the differences between men and women, and that other neurological mechanisms, as yet not fully understood, may contribute. And, as with sex and gender differences in the pain experience, psycho-social and cultural factors also play a role in the expression of other symptoms. In summary, this review of the literature has uncovered methodological shortcomings in research to date, including an absence of confirmed ischemia amongst subjects, limited or no adjustment for important clinical and sociodemographic co-variates, and use of retrospective research designs. Thus, there remains uncertainty in our knowledge regarding sex and gender differences in the symptoms of ACS. The literature on gender, sex and pain in general, also leaves many unanswered questions about the mechanisms for any observed differences, though there are many promising factors to consider. The following study, which was the basis for this doctoral dissertation, sought to address many of these gaps by prospectively examining the subjective experience of known myocardial ischemia in a controlled setting, to determine whether there are sex/gender differences in symptoms of myocardial ischemia. The study also examined the utility of this particular method of assessing the subjective experience of myocardial ischemia.    41 CHAPTER 3: METHODS Objective This study aimed to examine whether there are sex/gender differences in the symptom expression of ACS. A period of known, controlled myocardial ischemia, such as occurs during balloon inflation during percutaneous coronary intervention (PCI) procedures, presents an exceptional opportunity to explore this issue. Ethics Approval Approval to conduct this study was obtained from the University of British Columbia/Providence Healthcare Research Ethics Board and the University of British Columbia Clinical Research Ethics Board. The utmost care was taken to ensure that the highest ethical standards were met in this study: participation in the study was voluntary, informed consent was obtained from all study participants, and all data obtained during the study were kept strictly confidential by means of password-protected electronic records, and locked filing systems. Consent was obtained as follows. After ascertaining which patients were eligible from discussion with the nurse in charge of the hospital nursing unit, all eligible patients were approached by a cardiology research coordinator to ascertain whether they were interested in learning more about the study. Further screening was then carried out to ensure interested patients met all of the inclusion criteria.  For those who did, the research coordinator then began the informed consent process before each participant was enrolled in the study. Patients had ample time to have all their questions addressed about the study and what their participation would involve. Because of the potential power of suggestion to influence participants’ responses to questions about sensations, the specific aim of the study was not revealed, but was kept relatively vague. The specific wording used in the consent process was: “The purpose of the study is to evaluate patients’ experiences during angioplasty” and “We are most interested in   42 what patients feel as the balloon is inflated (blown up) during the angioplasty”. No specific references were made to chest pain, discomfort or, importantly, the notion of sex/gender differences. This approach was revealed explicitly to all Research Ethics Boards during the approval process, and was approved. Sample and Site The study was conducted at two sites: St. Paul’s Hospital and Vancouver General Hospital, both of which are large, university-affiliated hospitals (475 and 955 beds, respectively). Each provides a high volume of tertiary and quaternary-level cardiac services, including approximately 1,300 PCIs each, per year. Inclusion Criteria Patients who were scheduled for urgent or elective PCI, were able to provide consent, could understand and read English, were 19 years old or older, and who met no exclusion criteria were considered for participation. Exclusion Criteria The following patients were excluded from consideration: those who were not able to consent for any reason (e.g., confusion, dysphasia, decreased level of consciousness; unable to understand or read English); had signs or symptoms suggestive of myocardial ischemia immediately before the procedure (including (i) complaints of chest, arm, neck, jaw, shoulder, epigastric or back pain; (ii) new, unexplained shortness of breath; (iii) new, unexplained dizziness or syncope; or (iv) ST-segment deviation greater than or equal to one millimetre in any electrocardiographic (ECG) lead (Anderson et al., 2007; Antman et al., 2004; Bassand et al., 2007); were hemodynamically unstable (defined as any one of: systolic blood pressure less than 90 mmHg for more than 30 minutes; signs or symptoms of ischemia, as above; or change in   43 level of consciousness); or had a total occlusion of the first coronary artery being treated. Justification for these criteria follows. Nineteen years is considered the age of majority in British Columbia, although it has been argued that competence and capacity to consent to research should not be linked to age (Bessant, 2006). However, non-inclusion of participants under the age of 19 years was not considered a significant limitation because coronary artery disease most commonly affects adults of middle age, and does not usually become clinically evident until after the age of 50 years in men and 60 years in women (Heart and Stroke Foundation of Canada, 2000). The first exclusion criterion is in keeping with both the Canadian Tri-Council policy statement and the Declaration of Helsinki (Canadian Institutes of Health Research, Natural Sciences and Engineering Research Council of Canada, & Social Sciences and Humanities Research Council of Canada, 1998 (with 2000, 2002 and 2005 amendments); World Medical Association, 2008). As for the language ability criterion, notwithstanding that the language one speaks may reflect ethnic and cultural background which, in turn, may play some role in perception and expression of symptoms (Campbell, Edwards, & Fillingim, 2005; Kern, 1987; Morris, 1991), the financial resources available to conduct this study did not permit translation of materials or conduct of the interviews in languages other than English. This is a recognised limitation. Signs and symptoms of myocardial ischemia before the research protocol could be implemented would cause problems on several levels. First, there would be a clinical imperative to treat the ischemia, which would affect subsequent experience of ischemia. Second, it would possibly be unsafe to subject such patients to prolonged balloon inflation since this would prolong the duration of the ischemic episode. And finally, pre-existing symptoms would obviously confound the interpretation of newly-induced-symptom data. Presence of   44 hemodynamic instability could present a similar clinical risk to a participant who underwent prolonged balloon inflation. The original protocol called for patients with diabetes to be excluded because of the widely-held belief that they experience less pain during MI, possibly because of autonomic dysfunction ((Acharya, Shekhar, Aggarwal, & Annand, 1991; Canto et al., 2000; Teoh et al., 2007). Although some doubt has recently been cast on this assumption (Stone, Khunti, Squire, & Paul, 2008), we nevertheless believed that it was prudent to exclude patients with diabetes, to avoid this potential confounding variable. In addition, due to the relatively small sample size, there would have been limited statistical power to examine this subset. During the conduct of the study however, there were irregularities in adherence to this criterion. This will be discussed in more detail in the following chapter (Findings). Finally, if the artery being intervened upon was found to have a total occlusion, the patient was excluded. This is because of the very different techniques required to open such lesions, meaning that the study protocol could not be implemented faithfully, and also the fact that ischemia could not have been induced in such participants. Screening, Recruitment and Consent Recruitment for the study began in December 2003 and finished in February 2008. Site 1 began recruitment first and continued until February 2006, while Site 2 began in October 2004 and continued until February 2008. In collaboration with charge nurses, potential participants were identified from the daily “slate” for the heart catheterisation laboratories. The consent process then proceeded as outlined above. Data Collection Procedures Data collection took place in the cardiac short-stay unit, where patients are admitted and prepared for their procedure, and in the catheterisation laboratory, during the PCI itself. It took   45 approximately 30 minutes to obtain consent, to answer some preparatory questions, and to record the baseline data. Patients were not required to spend any additional time in the laboratory beyond that required for their usual care. Research staff members were trained to collect data using standardised instruments and approaches. Before the procedure, participants were asked to describe the symptoms they had experienced previously that had resulted in their referral for PCI using a standardised rating scheme (see Appendix C). Similar questioning was repeated during and after balloon inflation, pertaining to any current symptoms. During PCI, the onset and duration of ischemia elicited by catheter balloon inflation is well controlled. This provided a unique opportunity to compare men and women in their reporting of sensations experienced as blood flow was occluded through the coronary artery. However, certain issues related to individual patient characteristics and the maturation of PCI technology and expertise have the potential to interfere with the application of rigid research protocols.  First, the number and location of lesions being treated vary widely from patient to patient. Two studies have examined the effect of the particular coronary artery being dilated during PCI, and the location of the lesion (i.e., proximal, mid- or distal vessels) on electrocardiographic changes (Aldrich et al., 1987; Berry, Zalewski, Kovach, Savage, & Goldberg, 1989). Their findings were that ST deviation is greatest when ischemia is caused by occlusion of either the left anterior descending or right coronary artery, as compared with the circumflex artery, which typically shows minimal, or no ST changes. Therefore, it was necessary to consider the coronary artery dilated and the location of the lesion because these may have affected not only the amount of ST deviation, but also the symptoms experienced during PCI- induced ischemia. Second, through a process termed “ischemic pre-conditioning”, the severity of myocardial ischemia during balloon inflation decreases with subsequent inflations, so that   46 symptom severity may be minimised (Deutsch et al., 1990; Kloner & Yellon, 1994; Tomai et al., 1994). Third, certain drugs (e.g., nitroglycerin) may be administered to provide myocardial protection and to alleviate anginal pain during the PCI (Kato & Yoshimoto, 2003). Finally, because of improved balloon catheter technology, comparable angiographic results and better tolerance by patients, the average balloon inflation duration ranges from 10 to 20 seconds, as opposed to inflation durations in excess of two minutes, which were common 10 years ago (Garrahy et al., 1991).  Therefore, due to the potential logistic complexity of having several different protocols for the number of vessels dilated and lesions treated, duration of balloon inflations, and medications administered, and also to avoid compromising individualisation of care for patients, we determined that the protocol during the observation period should allow the attending interventional cardiologist to implement such care as deemed appropriate to the patient’s particular clinical situation. For the purpose of this study, data were collected only for the first balloon inflation in the first treated lesion. Research coordinators were instructed to record the vessel and location of the lesion for each recording of symptoms, the medications administered during the procedure, and duration of balloon inflation.  During subsequent data analyses, these variations in procedure from participant to participant were controlled statistically, where possible. Measures Demographic Data Symptoms of ACS vary from individual to individual: subjective experiences such as these may be influenced by a person’s education, or social and cultural background (O'Rourke, 1994). Therefore, demographic data were extracted from the participants’ health records and through survey questions, to permit exploration of symptom perception and recognition as a   47 potential function of age, educational attainment, occupation, first language spoken, years of Canadian residency, and ethnicity. In particular, the question pertaining to ethnicity was: “Most people in Canada describe themselves as Canadian first, but also identify themselves based on their background or the nationality of their ancestors. What would you say is your main ethnic background?” This clearly emphasizes their ancestral background, as opposed to how they may currently identify themselves. The limitations of this question are discussed later. Clinical and Procedural Data Data regarding participants’ pre-procedural clinical status and also intra-procedural data were obtained from the British Columbia Cardiac Registry (BCCR) record that is routinely completed during the procedure. This standardised tool is used to collect data that are subsequently included in the province-wide, procedure-based registry of all patients who have undergone any of several cardiac procedures (coronary angiogram, PCI, cardiac surgery, or pacemaker or internal cardioverter-defibrillator [ICD] implantation). The tool is completed both manually and electronically by nurses and interventional cardiologists involved in the procedure. Pre-procedure clinical status is characterised by the following variables: urgency at the time of the procedure; severity of anginal symptoms, if present, as measured by the globally-used Canadian Cardiovascular Society (CCS) grading system (Campeau, 1976) (a scale for describing the degree to which angina interferes with patients’ functioning); indication for the procedure; presence of known outcome determinants, such as renal insufficiency, hypertension, smoking or other co-morbid conditions; and medications taken in the 24-hour period before the procedure. Procedural data include the actual procedure performed; technical details of the procedure (e.g., amount and type of contrast agent used, duration of fluoroscopy); an inventory of the precise location and specific treatment employed (e.g., type of stent) for each lesion treated. Coronary blood flow, both before and after the intervention, is characterised by a widely-used grading   48 system known as “TIMI flow”, which derives its name from a long series of influential studies (Thrombolysis in Myocardial Infarction), the early ones of which were designed to examine the how efficacy of thrombolytic agents. Efficacy was determined definitively by analysing patency of an infarct-related coronary artery via an angiogram soon after a thrombolytic agent was administered, which gave rise to the need for a standardised system for classifying patency (The TIMI Study Group, 1984). The TIMI flow system grades coronary blood flow using visual assessment of the speed of contrast opacification of the infarct-related artery, and is now used almost universally in clinical practice as a standard measure of coronary artery patency (Kern et al., 1996). The outcome of the procedure as a whole is also captured in the BCCR record, that is, whether total revascularization was achieved, and if not, why. Some clinical variables from the BC Cardiac Registry record had a high degree of missingness (greater than 90%) in this sample. This was due to several factors at both the system and protocol level. First, if data are not captured during the procedure, they are largely unattainable, since patients are often outpatients or patients from other hospitals. Unfortunately, there are no clinical or other consequences for members of the clinical staff, who are responsible for initial data entry, if data are incomplete. Second, the final report is not available until a data entry clerk finalises the record, often days or weeks after the procedure. Finally, the research protocol directed study staff to simply print the registry report as soon as it was available, but not to check for missing elements. However, even if they had checked at that stage, since the participant had usually been discharged, the information would not have been available to them. The variables with high missingness were New York Heart Association (NYHA) Classification of Functional Status (The Criteria Committee of the New York Heart Association, Inc. (Kossman, C. E., chairman), 1964), another widely-used scale to grade the functional status of patients with cardiac disease; cerebrovascular disease; prior ECG findings and other prior   49 diagnostic tests. Since other variables captured pre-existing and concurrent heart failure, the NYHA Classification variable was not used. Unfortunately, cerebrovascular disease, although of interest as a related condition and possible confounding variable, could not be reliably used in the analyses. Symptom and Ischemia Monitoring The research assistants were specifically trained for consistency in explaining and rehearsing the study procedure with participants, and in asking the study questions. The measure of symptoms consisted of four questions, designed to elicit verbal responses regarding sensations that participants attributed to their heart, before, during and after balloon inflation. The questions were formatted to allow participants to describe their chief symptom through an initial, open-ended question: “We would like to know about any discomfort you have been experiencing that has led to your being referred for this procedure. Thinking back over the last several months, what is the main location of any discomfort you have had?”  Descriptions of quality and intensity were sought by asking: “How severe was this discomfort, on a scale of 1 to 10, with 1 being very minimal, and 10 being the most intense?” “In one or two words, how would you describe the discomfort?” “Did you feel anything else?” If the participant was unable to provide descriptors or hesitated, the interviewer provided examples by reading some predetermined response options, such as “squeezing,” “pressing,” “burning,” or “a weight.” The last three questions were repeated until the participant stated there were no more sensations to report. These four questions, initially asked before the patient entered the catheterisation laboratory, provided baseline data about pre-existing symptoms, but also served as a rehearsal for the type of questioning that would occur during the   50 PCI. The same battery of questions was repeated (omitting reference to the past several months) during balloon inflation and immediately following balloon inflation, until the participant offered no further descriptions. As discussed earlier, to minimise the power of suggestion, the specific aims of the questioning (i.e., examining sex/gender differences, during ischemia in particular) were not revealed fully, and the questions were asked at four instances throughout the PCI procedure to distract from the emphasis on symptoms of ischemia. However, if patients had any questions about the research purpose or protocol after the procedure, research assistants were instructed to answer these freely and completely. In fact, no participant asked any further questions after the study protocol was complete. Participants’ responses were recorded on the symptom monitoring tool (see Appendix C). As well as data about sensations, data regarding the first vessel treated, the duration of balloon inflation in seconds, and medications administered during the procedure were collected and recorded on the main data collection tool. The exact timing of the balloon inflations was recorded to ensure concurrency of symptom and ST segment data collection. ST segment deviation across all monitored leads (usually seven [I, II, III, aVL, aVR, aVF and a single precordial lead, usually V2 or V3]) was measured electronically using General Electric® ST-segment monitoring software, which is used routinely during PCI procedures. This software has been shown to be highly reliable and sensitive (Goodman et al., 2000; Klootwijk et al., 1996). As well, multi-lead ECG rhythm strips were obtained for manual measurement of ST deviation. Unfortunately, there was lack of quality control during the electronic measurements, so ultimately only manual measurements, completed after all data had been collected, were used in the analyses. This is explained in detail in a following section. Deviation across all available leads was measured, to confirm the presence and extent of myocardial ischemia.   51 Analysis Data collection instruments and forms were reviewed for completeness and entered into an electronic file. Before conducting the primary analyses, univariate descriptive statistics were generated to examine data for accuracy, to determine the extent of missing data and outliers, and to assess the distributional properties of the variables. Characterization of the Sample Descriptive statistics were used to profile the sample’s demographic and baseline clinical characteristics, as well as details of the PCI procedure. Frequencies for the whole sample and for men and women were generated and examined for each variable. Demographic Variables Raw data were regrouped where necessary to yield usable classifications. Occupational status was classified using the National Occupational Classification system (Human Resources and Social Development Canada, 2008). Language first spoken was classified using the taxonomy used by Statistics Canada (2008b), which showed that the participants’ first languages belonged to 10 of the relevant categories. These categories were subsequently collapsed into English; European languages; Indo-Iranian languages; Sino-Tibetan, Malaysian or Polynesian languages; and other languages. Responses to the question of ethnic origin were initially classified using categories employed by Statistics Canada (2008a), which showed the participants belonging to 13 of the relevant ethnic categories. These categories were further collapsed into six categories: Canadian, Aboriginal, British, European, Asian and “other.” Data regarding highest educational attainment, originally collected in 16 categories, were eventually transformed into four categories, again in keeping with Census Canada classifications. To determine the extent to which this sample was an accurate representation of the population from which it was drawn (British Columbians with ischemic heart disease), socio-   52 demographic characteristics of the sample were compared with available socio-demographic characteristics of BC participants of the 2006 Canadian Community Health Survey (CCHS). The CCHS is a cross-sectional survey that collects information about health status, utilization of health care services, and health determinants for the Canadian population, on a two-year cycle. The CCHS covers approximately 98% of the Canadian population aged 12 years and over (excluding aboriginals, members of the RCMP, those living on military bases or incarcerated persons) (Statistics Canada, 2006). The most recent data available, collected in 2005 (Cycle 3.1), were used. The reported CCHS variables relating to ethnicity and language were sufficiently different from the variables used in this study to make comparisons invalid. Because of such differences between the CCHS data and the data from this study, only age, education, immigration status, and years since immigration could be compared reliably. As well, some minor recoding of these variables from this study was required to make valid comparisons. Baseline Clinical Characteristics Some raw data were transformed to create more useful clinical variables. Estimated glomerular filtration rate (eGFR) was derived from the clinical variables creatinine, age, weight, and sex, using the Cockcroft-Gault formula (Cockcroft & Gault, 1976). This formula has been shown to be preferable to the Modification of Diet in Renal Disease (MDRD) formula (Peterson et al., 1995), when dealing with ACS patients (Melloni et al., 2008). Having this estimate, renal dysfunction was then defined as eGFR of less than 90 mL/min/1.73 m2 (Kidney Disease Outcomes Quality Initiative Workgroup of the National Kidney Foundation, 2002). For later multivariate analyses, the trichotomous classification of diabetes as Type I; Type II, insulin- dependent; and Type II, non-insulin-dependent was collapsed into a dichotomous variable “any diabetes”.   53 Procedural Details As mentioned previously, data from the BC Cardiac Registry (BCCR) were used to describe all facets of the PCI procedure.  Frequencies and distributions as well as descriptive statistics were generated and examined for all of these variables. Frequency and Nature of Symptoms To proceed with answering research question one (What is the subjective experience of myocardial ischemia during a period of known, controlled myocardial ischemia, with respect to a) perception of any symptoms; b) number of symptoms; c) intensity of symptoms; and d) location and type of symptoms?), it was necessary to derive certain new variables. First, since a substantial number of participants had no discomfort, the dichotomous variable “presence of any discomfort” (yes/no) was created, to permit comparisons of these two groups. Qualitative descriptors of the participants’ main symptoms at baseline (i.e., the symptoms that had led to their referral for PCI) and during inflation had initially been recorded nearly verbatim and entered into SPSS® as a string (text) variable. After the main symptom was recorded, a checklist in the data collection instrument containing common symptoms of ischemia was used to record subsequent symptoms, thus precluding capture of the participants’ exact words for these symptoms. However, several additional fields allowed entry of any “other” discomforts or symptoms not covered by the checklist, resulting in more text. The total number of descriptors in the raw data was thus considerable, at approximately 150. Therefore, some collapsing was required before these data could be recoded into nominal numeric variables, so that further statistical analyses would be possible. This involved simple clinical or etymological interpretation: for example “middle back” and “back” were collapsed into the single category of “back”, and all descriptions containing any reference to the chest but lacking any reference to the right side were collapsed into “chest, not right”. Collapsing was strictly limited to cases in   54 which the descriptors clearly referred to only non-right chest discomfort of some sort. Similarly, for arm discomfort, “left arm” was created as a category for instances in which there were no mention of the right side. This process was applied to all string (text) variables. Again, if there was any ambiguity, the descriptor was given a unique code. A list of the raw descriptor data and resulting codes is included as Appendix D. Frequencies of each reported type of discomfort were generated for the whole sample, for men and women, and for the sub-sample of participants with ECG-confirmed ischemia. Symptom variables (from both baseline and balloon inflation questioning) were further refined to create two new dichotomous variables: the presence of chest discomfort and the presence of ‘typical’ symptoms. Chest Discomfort The operational definition adopted for chest discomfort was any description that included any of the terms chest, heart, sternum/sternal or breast/breastbone. Further descriptors of location that were allowed within this definition were left, centre/central, upper, middle/mid, all over, across, and under (breast). The term “right” on its own was not allowed in this definition, unless one of the other location descriptors (e.g., left or mid) was also included in the same phrase, so that specifically right-sided chest discomfort (which is considered atypical of ischemic discomfort) could be differentiated from either left, mid, or all-over chest discomfort (which is considered typical) (Anderson et al., 2007; Bassand et al., 2007). Therefore, right-sided chest discomfort was coded as a separate symptom. Typical Symptoms To enable comparison with certain previous studies in which symptoms were characterised as either “typical” or “atypical”, the dichotomous variable “typical” was created. The operational definition of “typical” symptoms was derived from the American College of   55 Cardiology/American Heart Association practice guidelines for the management of unstable angina or non-ST-elevation myocardial infarction (NSTEMI) patients (Anderson et al., 2007), and from two European Society of Cardiology’s practice guidelines: managing unstable angina or NSTEMI patients (Bassand et al., 2007), and managing patients with stable angina (The Task Force on the Management of Stable Angina Pectoris of the European Society of Cardiology, 2006). The resulting definition included the following symptoms: chest, shoulder, back, arm (one or both), jaw or epigastric pain or discomfort; unexplained weakness; nausea or vomiting; shortness of breath; light-headedness; diaphoresis; and fear or restlessness. Number and Intensity of Symptoms The number of symptoms reported by each participant was tallied, and the distribution of data in this variable, both for the sample as a whole and for men and women, was examined. Scores for intensity of the main reported discomfort were examined for their distributional properties, for the whole sample and for men and women. Concordance of Baseline and Induced (Inflation) Symptom To test whether balloon inflation was a reliable model of spontaneous myocardial ischemia (i.e., had acceptable test-retest reliability), symptom data from both initial questioning and questioning during balloon inflation were further transformed into nine dichotomous variables corresponding to the eight most commonly reported symptoms and also the report of “no discomfort”. These nine variables yielded a comprehensive profile of symptoms, in that they comprised 88.9% of reported previously-occurring symptoms and 88.0% of reported balloon- inflation symptoms. Cross-tabulation then allowed for evaluation of the proportion of observed agreement between the two observations. The kappa statistic was also calculated for each of these pairs of measures, to test for the degree of reliability that was not due to chance alone (Hunt, 1986). These two statistical tests may yield a paradoxical result, however: high agreement   56 but low kappa, due to asymmetry in the marginal totals of the two-by-two table (Feinstein & Cicchetti, 1990). To resolve this issue, the observed proportions of positive and negative agreement (Ppos and Pneg) were also calculated (Cicchetti & Feinstein, 1990). Identification of Myocardial Ischemia As stated previously, there were missing ECG data that were irretrievable. This resulted from either malfunction of the clinical information system or failure to record a sufficient number of leads at the time of the procedure. If ECG tracings were not captured at the time of the procedure, then this information could never be obtained, since there was obviously no opportunity to re-create the circumstances that had occurred during the study observation period. To ensure that there was assessment of each main myocardial territory (anterior, inferior, lateral and posterior walls) only cases in which there were six or more leads available for analysis, one of which was a precordial lead, were included (Wagner & Marriot, 2008). This resulted in the exclusion of 13 cases. Although the General Electric® ST measurement software is considered to be reliable and accurate (Goodman et al., 2000; Klootwijk et al., 1996), unfortunately many of the electronic measurements of ST deviation were not evaluable because care and attention had not been taken during acquisition. Therefore, all the ST segments were measured manually, using the following procedure: the rhythm strip was considered non-evaluable if the underlying cardiac rhythm was not sinus in origin, or if a bundle-branch block pattern was evident. For those rhythm strips that were deemed evaluable: the isoelectric line was identified using the PR segment; the J-point was identified; and, the height of the ST segment, in relation to the isoelectric line, was measured at a point 60 milliseconds after the J-point (Drew et al., 2005). Research assistants were given written guidelines to identify sinus rhythm and bundle branch block, and were instructed to verify their assessment with the investigator if there was any uncertainty. Three research assistants were   57 trained in this procedure by the investigator: two were experienced critical care nurses, who were very familiar with rhythm recognition, and therefore received only review of the correct procedure for ST measurement. The other was a non-nursing graduate student who was taught how to measure all necessary intervals and how to recognize sinus rhythm. The measurements performed by the research assistants were cross-checked until 100% agreement was achieved: for the two nurses, this took 15 cases, for those measurements performed by the graduate student, this was achieved in 25 cases. As well, the investigator verified some ECG rhythms with staff cardiologists at study site 2, when interpretation was unclear. Once individual ST measurements were completed, an operational definition for myocardial ischemia was needed to differentiate those participants for whom there was ECG evidence of ischemia from those for whom there was not. Again using the ACC/AHA Guidelines for Management of Patients with Unstable Angina/NSTEMI (Anderson et al., 2007), ischemia was defined as ST deviation of 1 mm or more in any lead. Thus, a subset of the sample was confirmed as those who experienced ischemia, and most subsequent analyses were undertaken on the ischemic subsample. Henceforth, therefore, the term ischemia should be understood to mean ECG evidence of ischemia, as defined above. Sex/Gender Differences in Symptom Profiles After completing univariate analyses of the variables outlined above to answer the research question, “What is the subjective experience of myocardial ischemia during a period of known, controlled myocardial ischemia, with respect to: (a) perception of any symptoms, (b) number of symptoms, (c) intensity of symptoms, and (d) location and type of symptoms?”, I proceeded with answering the research question, “Are there any sex/gender differences?” in each of these components of the experience. This entailed bivariate analyses using sex/gender as the covariate. Chi-square tests were used to test for sex/gender differences in the presence of   58 any symptoms, in location or type (e.g., shortness of breath, nausea) of symptom, and in the reporting of “typical” symptoms or chest pain. The large number of symptoms reported meant that some symptoms had expected frequency scores of less than five in 25% or more of the table cells. For these variables, Fisher’s Exact test was used. Intensity scores pertaining to inflation were somewhat positively skewed (skewness 0.18; SE 0.16) and platykurtotic (-1.06; SE 0.31), owing primarily to the high frequency of zero scores (no discomfort) obtained during balloon inflation. The t-test was used for sex/gender comparison on this variable (both including and excluding the zero scores) because the skewness was not extreme, the rest of the distribution was reasonably normal, and the t-test is robust to some violation of assumptions about distribution (Stocks, 2001). However, for completeness, the Kolmogorov-Smirnov test was also used to test for sex/gender differences in intensity. The Kolmogorov-Smirnov is a non-parametric test based on the assumptions that (a) all observations in the two samples are randomly selected and independent and (b) the measurement scale is at least ordinal. This test for two independent samples is more powerful than either the χ2 test for contingency tables or the median test for independent samples (Sheskin, 2007). Results from all tests were compared. In addition, a χ2 test was conducted to test for an association between sex/gender and the reporting of no discomfort during balloon inflation. The distribution of the number of reported symptoms in the whole sample was also non-normal (extremely leptokurtotic), with skewness of 1.61 (SE 0.14) and kurtosis of 5.61 (SE 0.28). This distribution also persisted in the subsamples of men and women. Therefore, the Mann-Whitney U-test was used for this comparison. In addition, to further knowledge of how sex or gender may be related to symptoms of myocardial ischemia, comparing findings from this study to previous researchers’ findings is desirable. However, as has been lamented by Canto (2007), until now there has been lack of   59 consistency in defining various symptoms or constellations of symptoms, making comparison difficult. Towards this end, an attempt was made to invoke some common typologies that have previously been used for defining symptoms (e.g., differing operational definitions of “typical” or “chest” pain or discomfort), but this was dependent on the availability of precise operational definitions. Regrettably, only one such definition was retrieved (Milner, Funk, Arnold, & Vaccarino, 2002); bivariate analysis, as described above, was repeated with this new variable as the outcome variable. Next, to explore other predictors of reporting certain symptoms, multivariate analyses were undertaken. As has been stressed before, the initial intent was to limit detailed analyses of reported symptoms to those reported primarily during balloon inflation by the sub-sample of participants with documented ischemia. However, for reasons that will be described more fully later, both the baseline and balloon-inflation symptom data were reviewed for the presence of latent classes. In light of this, it was useful to also include an analysis of the association of sex/gender with selected symptoms reported at baseline. For both the baseline and balloon inflation data analyses (the latter including only the sub-sample of participants with documented ischemia), multivariate analysis proceeded as follows. Symptom variables were explored using bivariate analysis with sex/gender; any symptom for which there were differences with statistical significance of less than 0.25 (Hosmer & Lemeshow, 2004) was then subjected to further bivariate analyses using all available clinical and demographic variables that were deemed to potentially predict the reporting of symptoms. Those clinical and demographic variables that reached significance of 0.25 or less were then entered en bloc into the model as predictors, using logistic regression, after which the model was trimmed of variables that were non-contributory (not reaching a significance level of  </= 0.05) in a step-wise fashion, to arrive at the most parsimonious model.   60 Analysis for Latent Classes Finally, latent class analysis was employed to explore whether there is an unobserved (latent) structure underlying symptom expression (research question three). Latent class analysis (LCA) is a data reduction technique used to discover and enumerate latent groups or classes that may exist, based on observed data from several manifest variables (UCLA Academic Technology Services; Uebersax, 2006). The term latent is used because such a variable is “…a random variable whose realizations are hidden from us…” (Skrondal & Rabe-Hesketh, 2004). Accordingly, somewhat like “factors” discovered in factor analysis, one can make inferences regarding unobserved constructs; that is, subclasses within a particular diagnostic class (e.g., patients with ACS) can potentially be identified. The assumption is made that each participant falls into only one class (Ungvari, Goggins, Leung, Lee, & Gerevich, 2009). Either categorical or ordinal data are appropriate for LCA. The following is an outline of the steps that were taken in the LCA. The number of latent classes was first estimated through an iterative process, using MPlus® statistical software (Muthén & Muthén, 2007). The most appropriate number of classes (goodness of model fit) was determined by examining both Aikaike’s information criterion (AIC) and the Bayesian information criterion (BIC). Since the data were relatively sparse (i.e., a large number of variables compared with the number of observations) the chi-square distribution was not considered to be appropriate to determine the p-value, so bootstrap p-values were obtained (Langeheine, Pannekoek, & Van de Pol, 1996), using 100 bootstrap draws. A non-significant p-value (greater than 0.05) was sought to determine the optimal number of clusters. To further support the validity of the final model, the Vuong-Lo-Mendell-Rubin likelihood ratio test of the null hypothesis (that there is only one class) was conducted. Once the model parameters were specified, class membership probabilities were calculated from estimated conditional response   61 probabilities. These probabilities, along with the estimated prevalence of each class and Bayesian theory, were used to determine the probability of membership in each class. Based on these probabilities, class membership was assigned to each case. It became apparent after this step that only one class was evident when considering symptoms reported during balloon inflation by the subset of participants with ischemia. Therefore, to take full advantage of the data available and the opportunity to characterise the ACS population further, the criteria for inclusion were expanded, and baseline symptoms of the whole sample were subjected to latent class analysis. After assignment of class membership, bivariate analyses of the relationships between class membership and each of the available socio-demographic and clinical covariates were conducted. In cases where the bivariate analyses uncovered differences with statistical significance of less than 0.25 (Hosmer & Lemeshow, 2004), multivariate analyses were undertaken to address research question four: “What are the characteristics of members of a particular class?” Potential predictors were entered en bloc, and tests of model fit were examined. Predictor variables that were deemed to have a low contribution to the model (significance of greater than 0.5) were then removed in a step-wise fashion, and the log-likelihood differences were examined for significance using chi-square distribution tables. If the log-likelihood difference reached significance by this metric, the last variable removed was returned to the model, and that model was deemed the best possible.   62 CHAPTER 4: FINDINGS The primary intent of the study was to examine symptoms in patients who are having current, ECG-confirmed myocardial ischemia. Therefore, detailed analyses related to reported symptoms were mostly limited to data from the sub-sample of participants with ECG- documented ischemia, focusing primarily on those symptoms reported during balloon inflation. The entire sample was considered in the following analyses: description of demographic, clinical and procedural characteristics, and frequency of reporting any symptoms. Subsequent analyses (i.e., number and intensity of symptoms and detailed analyses of location and type of symptoms) were divided into those relating to the baseline symptoms (included the whole sample, n = 305) and those relating to the inflation symptoms (included the subsample of participants with ECG- confirmed ischemia, n = 245). To permit the reader to easily compare these different analyses, baseline and inflation findings usually appear in the same table, or in sequential tables in some cases. All tables are appropriately labelled with information regarding the pertinent event and sample size. Study Sample The target population for the study was patients experiencing an ACS. The sample frame included patients who were undergoing elective, semi-urgent or urgent PCI.  Study recruitment occurred at two study centres between October 2004 and February 2008. Patients were sequentially screened for eligibility. Three hundred five men and women who underwent elective, semi-urgent or urgent percutaneous coronary intervention (PCI) comprised the final sample. Figure 1 outlines the disposition of potential participants.   63 Screened 820 (328+492)   Eligible 775 (94.5%)  (310+465)   Ineligible (e.g. non-English-speaking, emergency procedure, clinically unstable) 45 (5.5%) (18+27)   Consented 560 (72.2%)  (224+336)    Refused 213 (27.5%)  (85 + 128)  Missed 2 (0.02%) (1+1)   Remained Eligible after Angiogram 325 (58.0%) (140+185)   Ineligible after Angiogram (e.g. total occlusion of artery, PCI not performed) 235 (41.9%) (94+141)  PCI & interview complete 318 (97.8%) (136+182)   PCI or interview incomplete 7 (2.2%) (4+3)  Complete ECG data 305 (95.5%) (131+174) (FINAL SAMPLE) Incomplete ECG data 13 (4.5%) (2+11)  Figure 1. Details of all phases of participant screening and enrolment. Note: Numbers in parentheses indicate Site 1 & 2 statistics, respectively   64 The exact number of potential participants screened was not reliably recorded at Site 2. Therefore, this site’s quantities, for some phases of the recruitment process, were estimated, based on the percentage of participants that this site contributed to the enrolled sample (60%). It was assumed that the sites’ rates of initial and post-angiogram eligibility were similar. Reasons for exclusion of participants after initial inclusion (usually after a diagnostic angiogram was obtained) fell into three main categories: clinical (208 [88.5%]); for example, the presence of a chronic total occlusion of the coronary artery; the patient received analgesics for a non-cardiac reason, which could confound symptom perception; or a PCI was not performed because of clinical indications to refer for cardiac surgery or anatomic reasons to recommend medical management; technical (7 [3.0%]); for example,  equipment malfunction; and staffing (15 [6.4%]); for example, the procedure was completed outside of the regular hours of work for the research staff. The reason for post-angiogram exclusion was missing in 5 (2.1%) cases. Margin of Error for the Sample To quantify the precision of the estimates made from this sample, Table 3 lists the margins of error that are expected for several ranges of frequency point estimates from both the whole sample (n = 305) and the sub-sample of participants who had ECG-confirmed ischemia (n = 245).   65 Table 3.  Margins of Error for Point Estimates of Frequency, by Sample Size 95% CI1 for Margin of Error Point Estimate of Frequency (%) N = 305 n = 245 <1-10 -0.10 – 13.37 -0.12 - 13.76 11-20 7.49 – 24.49 7.08 – 25.01 21-30 16.43 – 35.14 15.90 – 35.73 31-40 25.81 – 45.50 25.21 – 46.13 41-50 35.48 – 55.61 34.84 – 56.26 51-60 45.39 – 65.50 44.74 – 66.13 61-70 55.53 – 75.14 54.89 – 75.74 71-80 65.91 – 84.49 65.32 – 85.01 81-90 76.60 – 93.37 76.09 – 93.76 91-100 87.79 – 100.10 87.42 – 100.12 1 CI, confidence interval Demographic Characteristics of the Sample Table 4 outlines the demographic characteristics of the sample. Women made up roughly 40% of the total, with the overall mean age being approximately 64 years. The age of the female participants, on average, was 3.0 years older than that of the male participants. Approximately two thirds of the participants (63.6%) reported their ethnicity, that is, their ancestral background, as “Canadian”, with British being the second most common ethnicity (14.4%). More men reported having a Canadian ethnic background than did women (67.4% vs. 57.9%). In keeping with the predominantly “Canadian ethnicity” of the sample, more than three quarters (78.7%) claimed English as their first language. Roughly the same percentage was evident when the female and male groups were compared (77.7% vs. 79.3%, respectively). The educational attainment of the sample was as follows: 21.6% had not completed high school, 30.5% had completed high school only, and 47.8% had completed at least some post-secondary education. The women in the sample, however, had lower levels of educational attainment than did the   66 men, with only 72.0% of them having completed high school or more, whereas 82.0% of men had achieved this level.  Table 4  Demographic Characteristics of Sample Gender Characteristic All (N = 305) Women (n = 121) Men (n = 184) Age, mean (SD) 63.9 (10.7) 65.8 (11.6) 62.8 (9.9) Gender (n (%)) Female Male  121 (39.7) 184 (60.3)  -- --  -- -- Ethnicity (n (%)) Canadian British European Asian Other (e.g., other North American, Latin, Central & South American)  194 (63.6) 44 (14.4) 24 (7.9) 13 (4.3) 30 (9.8)  70 (57.9) 25 (20.7) 9 (7.4) 4 (3.3) 13 (10.7)  124 (67.4) 19 (10.3) 15 (8.2) 9 (4.9) 17 (9.2) First language spoken (n (%)) English European language Indo-Iranian language Sino-Tibetan or Malayo-Polynesian language Other Language (e.g. Aboriginal)  240 (78.7) 42 (13.8) 12 (3.9) 9 (3.0) 2 (0.7)  94 (77.7) 20 (16.5) 3 (2.5) 4 (3.3)   146 (79.3) 22 (12.0) 9 (4.9) 5 (2.7) 2 (1.1) Highest education (%) Less than high school Completed high school Some post-secondary Completed post-secondary  66 (21.6) 94 (30.8) 78 (25.6) 67 (22.0)  34 (28.1) 36 (29.8) 31 (25.6) 20 (16.5)  32 (17.4) 58 (31.5) 47 (25.5) 47 (25.5)    67 Table 4 (cont’d) Demographic Characteristics of Sample Gender Characteristic All (N = 305) Women (n = 121) Men (n = 184) Occupation (n (%)) Management Business/finance/administration Natural & applied sciences Healthcare Social sciences/education/government/religion Arts & culture/sports & recreation Sales & service Trades, transportation, equipment operators Unique to resource-based industries Retired, previous occupation not specifieda Self-employed Unpaid homemaker/housewife Disabled Unemployed Missing  21 (6.9) 35 (11.5) 12 (3.9) 14 (4.6) 18 (5.9) 12 (3.9) 37 (12.1) 38 (12.5) 10 (3.3) 84 (27.5) 2 (0.7) 13 (4.3) 2 (0.7) 2 (0.7) 5 (1.6)  7 (5.8) 18 (14.9)  5 (4.1) 7 (5.8) 2 (1.7) 18 (14.9) 6 (5.0)  38 (31.4) 1 (0.8) 13 (10.7) 1 (0.8) 1 (0.8) 4 (3.3)  14 (7.6) 17 (9.2) 12 (6.5) 9 (4.9) 11 (6.0) 10 (5.4) 19 (10.3) 32 (17.4) 10 (5.4) 46 (25.0) 1 (0.5)  1 (0.5) 1 (0.5) 1 (0.5) Immigrant to Canada (n (%)) 90 (29.5) 28 (23.1) 62 (33.7) Years Since Immigration to Canada, mean (SD) 42.5 (17.1) 48.6 (17.5) 39.7 (16.3) a Retirees specifying a pre-retirement occupation (25 [8.1%]) were categorized within their respective occupation; therefore, only those not specifying an occupation were  categorized as “retired”.  The occupational group with the largest representation was the retirees, with more women than men reporting retired status (31.4% vs. 25.0%, respectively). Among those in the specifying an occupation (either current or prior to retirement), all nine occupational categories of the National Occupational Classification (NOCS) (Human Resources and Social Development Canada, 2008) were represented, with trades, transportation and equipment   68 operation; business, finance and administration; and sales and service being the occupations reported most often. However, the most common occupations for those in the workforce differed by gender;  employed women reported business, finance and administration (which includes administrative assistance such as secretarial positions) and sales and service as their first- and second-most common occupations, and “unpaid homemaker or housewife” was the third- most common occupation reported. Employed men, on the other hand, reported trades, transportation and equipment and sales and service as the first- and second-most common, while business, finance and administration was reported third. There were more men than women employed in the natural and applied sciences; arts and culture/sports and recreation; and occupations unique to resource-based industries (such as jobs related to mining or the oil industry). The comparisons that were possible between this sample and the British Columbia population who self-reported having heart disease were limited to age, gender, immigrant status, years in Canada since immigration, and educational attainment (see Table 5). This study’s sample was younger (53.8% under 65 years old versus 36.6%), had a higher proportion of men (60.3% vs. 56.4%), and had slightly less education (48% vs. 53% having started or completed post- secondary education) than the weighted sample of people from British Columbia who participated in the 2005 Canadian Community Health Survey (Statistics Canada, 2006). Finally, there was a somewhat lower percentage of people who had immigrated to Canada in this study (29.5% vs. 34.0%), but the percentage of those who had been in Canada greater than nine years since immigration was similar between the samples (94.1% vs. 94.4%).   69 Table 5  Comparison of Sample Demographic Characteristics with British Columbia Population who Self- Reported Having Heart Disease Data Source Characteristic Study Sample (N = 305) CCHSa (N = 145,933) Age (n (%)) < 65 years >/= 65  years  164 (53.8) 141 (46.2)  52,367 (36.6) 93,566 (63.4) Gender (n (%)) Female Male  121 (39.7) 184 (60.3)  64,338 (43.6) 83,127 (56.4) Immigrant (n (%()  Missing 90 (29.5) 9 (3.0) 50,150 (34.0) 1531 (1.0) Length of time in Canada since immigration (n (%)) 0-9 years > 9 years Not applicable Missing  5 (1.6) 85 (27.9) 206 (67.5) 9 (3.0)  2955 (2.0) 47,196 (32.0) 95,784 (65.0) 1531 (1.0) Highest Education (n (%)) Less than high school Completed high school Some post-secondary Completed post-secondary Missing  66 (21.6) 93 (30.5) 77 (25.3) 66 (21.6) 3 (1.0)  42.137 (28.6) 21,322 (14.5) 10,433 (7.1) 68,380 (46.4) 5193 (3.5) aCCHS, Canadian Community Health Survey, 2006 cycle, limited to BC residents who self-reported having heart disease.  Baseline Clinical Characteristics Table 6 outlines the baseline (pre-procedure) clinical attributes of the study sample. Over one half of the sample (53.4%) had significant functional limitations due to angina, as measured by the Canadian Cardiovascular Society classification of angina (Campeau, 1976), which was   70 comparable between men and women. Similarly, the procedure was classified as urgent or semi- urgent for 54.0% of the sample, which did not differ by sex/gender. The main indication for the procedure was almost exclusively ACS or stable coronary artery disease (97.4%), and men and women did not differ in this characteristic. About one third of the entire sample had had a previous MI, but the rate was lower in women (31.4% vs. 38.6% in men). As might be expected given the prevalence of previous MI in the sample, nearly 30% of the participants had had a prior revascularisation procedure, but this was less frequent in women (27.2%) than in men (31.0%). Rates of some co-morbid conditions (heart failure and renal dysfunction) were higher in the women than in the men, but the rate of diabetes (all types) was slightly higher in the men than in the women (21.8% vs. 16.5%). These rates are in keeping with prevalence rates reported for Canadians between 50 and 80 years of age (Public Health Agency of Canada, 2008). The medications received or taken within the previous 24 hours by the participants included common cardiac medications: over 90% had had ASA, 77% had had beta-blocking agents; 73% had had lipid-lowering agents, and 67.9% had had clopidogrel. Of note, the women were prescribed these drugs less frequently than were men: ASA (86.8% vs. 95.1%), beta blockers (71.9% vs. 80.4%), and lipid-lowering agents (69.4% vs. 75.5%). However, more than twice as many women as men (29.8% vs. 12.5%) were prescribed calcium channel antagonists.   71 Table 6  Baseline Clinical Characteristics of Sample Gender Characteristic All (N = 305) Women (n = 121) Men (n = 184) CCSa angina class (n (%)) 0 I II III IV Atypical Missing  1 (0.3) 17 (5.6) 75 (24.6) 54 (17.7) 109 (35.7) 13 (4.3) 36 (11.8)  1 (0.8) 6 (5.0) 27 (22.3) 22 (18.2) 41 (33.9) 6 (5.0) 18 (14.9)   11 (6.0) 48 (26.1) 32 (17.4 ) 68 (36.9) 7 (3.8) 18 (9.8) Urgency of procedure (n (%)) Elective Semi-Urgent Urgent Missing  138 (45.2) 19 (6.2) 147 (48.2) 1 (0.3)  54 (44.6) 3 (2.5) 63 (52.1) 1 (0.8)  84 (45.7) 16 (8.7) 84 (45.7)  Main indication for procedure (n (%)) ACSb Stable angina/CADc Atypical angina Positive exercise test/perfusion imaging Other, not specified  131 (43.0) 166 (54.4) 1 (0.3) 2 (0.6) 6 (1.9)  53 (43.8) 65 (53.7) 1 (0.8)  2 (1.7)  78 (42.4) 101 (54.9)  2 (1.0) 4 (2.1) Prior MId (n (%)) 109 (35.7) 38 (31.4) 71 (38.6) Prior revascularisation (n (%))  PCIe  CABGf  72 (23.6) 18 (5.9)  28 (23.1) 5 (4.1)  44 (23.9) 13 (7.1) aCCS, Canadian Cardiovascular Society; bACS, acute coronary syndrome; cCAD, coronary artery disease; dMI, myocardial infarction; ePCI, percutaneous coronary intervention; fCABG, coronary artery bypass graft surgery; grenal dysfunction was defined as estimated glomerular filtration rate (eGFR) (Cockroft-Gault) < 90 ml/min per1.73 m2; hACE, angiotensin-converting enzyme.   72 Table 6 (cont’d) Baseline Clinical Characteristics of Sample Gender Characteristic All (N = 305) Women (n = 121) Men (n = 184) Heart failure (n (%))  Pre-existing  Missing  At time of procedure  14 (4.6) 2 (0.7) 1 (0.3)  9 (7.4) 1 (0.8)   5 (2.7) 1 (0.5) 1 (0.5) Renal dysfunctiong (n (%))  Missing 180 (59.0) 10 (3.3) 85 (70.2) 4 (3.3) 95 (51.6) 6 (3.3) Diabetes (%) Type I Type II (insulin-dependent) Type II (non-insulin-dependent)  2 (0.7) 25 (8.2) 33 (10.8)   7 (5.8) 13 (10.8)  2 (1.1) 18 (9.8) 20 (10.9) Smoking status (n (%)) Never Current Former  103 (33.8) 42 (13.8) 160 (52.5)  51 (42.1) 18 (14.9) 52 (43.0)  52 (28.3) 24 (13.0) 108 (58.7) aCCS, Canadian Cardiovascular Society; bACS, acute coronary syndrome; cCAD, coronary artery disease; dMI, myocardial infarction; ePCI, percutaneous coronary intervention; fCABG, coronary artery bypass graft surgery; grenal dysfunction was defined as estimated glomerular filtration rate (eGFR) (Cockroft-Gault) < 90 ml/min per1.73 m2; hACE, angiotensin-converting enzyme.   73 Table 6 (cont’d) Baseline Clinical Characteristics of Sample Gender Characteristic All (N = 305) Women (n = 121) Men (n = 184) Pre-procedure medications (n (%)) IV nitroglycerin Long-acting nitrate ASA Beta blocker ACEh inhibitor Calcium-channel antagonist Lipid-lowering agent Clopidogrel Oral hypoglycemic Hormone therapy Anticoagulant Anti-depressant No medication  2 (0.7) 78 (25.6) 280 (91.8) 235 (77.0) 169 (55.4) 59 (19.3) 23 (73.1) 207 (67.9) 9 (3.0) 9 (3.0) 54 (17.7) 13(4.3) 5 (1.6)  1 (0.8) 32 (26.4) 105 (86.8) 87 (71.9) 65 (53.7) 36 (29.8) 84 (69.4) 79 (65.3) 3 (2.5) 8 (6.6) 22 (18.2) 6 (5.0) 3 (2.5)  1 (0.5) 46 (25.0) 175 (95.1) 148 (80.4) 104 (56.5) 23 (12.5) 139 (75.5) 128 (69.6) 6 (3.3) 1 (0.5) 32 (17.4) 7 (3.8) 2 (1.1) aCCS, Canadian Cardiovascular Society; bACS, acute coronary syndrome; cCAD, coronary artery disease; dMI, myocardial infarction; ePCI, percutaneous coronary intervention; fCABG, coronary artery bypass graft surgery; grenal dysfunction was defined as eGFR (Cockroft-Gault) < 90 ml/min/1.73 m2; hACE, angiotensin-converting enzyme.  Procedural Characteristics Table 7 summarises the sample characteristics with respect to procedural details.   74 Table 7  Procedural Characteristics of Sample Gender Characteristic All (N = 305) Women (n = 121) Men (n = 184) Access site (n (%)) Femoral Radial Missing  294 (96.4) 10 (3.3) 1 (0.3)  117 (96.7) 4 (3.3)   177 (96.2) 6 (3.3) 1 (0.5) First vessel/segment treated (n (%)) Left anterior descending Proximal Mid Distal Diagonal 1st 2nd Ramus Circumflex Proximal Mid Obtuse marginal 1st 2nd 3rd Right coronary Ostium Proximal Mid Distal Posterior descending  115 (37.6) 44 (14.4) 69 (22.6) 1 (0.3) 8 (2.6) 4 (1.3) 4 (1.3) 8 (2.6) 44 (14.4) 11 (3.6) 33 (10.8) 27 (9.0) 15 (4.9) 10 (3.3) 2 (0.7) 100 (32.9) 3 (1.0) 38 (12.5) 38 (12.5) 21 (6.9) 3 (1.0)   44 (36.3) 12 (9.9) 31 (25.6) 1 (8) 4 (3.4) 2 (1.7) 2 (1.7) 4 (3.3) 13 (10.8) 3 (2.5) 10 (8.3) 10 (8.3) 7 (5.8) 2 (1.7) 1 (0.8) 45 (37.1) 1 (0.8) 20 (16.5) 15 (12.4) 9 (7.4) 1 (0.8)  71 (38.6) 32 (17.4) 39 (21.2)  4 (2.2) 2 (1.1) 2 (1.1) 4 (2.2) 31 (16.8) 23 (12.5) 8 (4.3) 17 (9.1) 8 (4.3) 8 (4.3) 1 (0.5) 55 (29.8) 2 (1.0) 18 (9.8) 23 (12.5) 12 (6.5) 2 (1.1)   75 Table 7 (cont’d) Procedural Characteristics of Sample Gender Characteristic All (N = 305) Women (n = 121) Men (n = 184) % Stenosis Pre-PCI, mean (SD) (n = 299) Post-PCI, mean (SD) (n = 292)  83.0 (15.8) 4.0 (10.5)  82.0 (16.5) 4.6 (9.7)  83.7 (15.3) 3.5 (11.1) TIMIa flow grade (n (%)) (n = 275) Pre-PCI <3 3 Missing Post-PCI <3 3 Missing   27 (8.9) 248 (81.3) 30 (9.8)  5 (1.7) 270 (88.5) 30 (9.8)   7 (5.8) 106 (87.6) 8 (6.6)  1 (0.8) 112 (92.6) 8 (6.6)   20 (10.8) 142 (77.2) 22 (12.0)  4 (2.1) 158 (85.9) 22 (12.0) Stent deployed (n (%)) 292 (95.7) 115 (95.0) 177 (96.2) Fluoro Time (mins), mean (SD) (n = 302) 12.7 (10.1) 11.3 (7.5) 13.5 (11.4) Balloon Inflation Duration (secs), mean (SD) 71.9 (31.8) 63.7 (37.1) 77.6 (40.8) Complications (n (%)) (n = 302) None VTb/VFc, converted Access site complication Vagal reaction Other (e.g., dissection, angina, abrupt closure) Missing  292 (95.7) 2 (0.7) 1 (0.3) 3 (1.0) 4 (1.4) 3 (1.0)  116 (95.9)  1 (0.8) 1 (0.8) 2 (1.6) 1 (0.8)  177 (96.2) 2 (1.1)  2 (1.1) 2 (0.5) 2 (1.1) aTIMI, Thrombolysis in Myocardial Infarction; b VT, ventricular tachycardia; c VF, ventricular fibrillation.   76 Table 7 (cont’d) Procedural Characteristics of Sample Gender Characteristic All (N = 305) Women (n = 121) Men (n = 184) Complete Revascularisation Achieved (n (%)) (n = 298) Missing Reasons for incomplete revascularisation (n (%)) By intent PCI failure Missing  256 (83.9) 7 (2.3)   38 (90.5) 2 (4.7) 2 (4.7)  104 (86.0) 2 (1.7)   13 (86.6) 1 (6.6) 1 (6.6)  152 (82.6) 5 (2.7)   25 (92.6) 1 (3.7) 1 (3.7) aTIMI, Thrombolysis in Myocardial Infarction; b VT, ventricular tachycardia; c VF, ventricular fibrillation.  The vascular access site used for almost all of the participants (more than 95%) was the femoral artery. The first coronary artery treated was either the left anterior descending (LAD) coronary artery or the right coronary artery (RCA) in approximately two thirds of the sample (37.6% and 32.9%, respectively). The other one third of the participants had a mixture of circumflex, diagonal, ramus and posterior descending coronary arteries treated. When examining men and women separately, a difference in the location of the first lesion treated was evident: women more often received intervention on the RCA than men did (37.1% vs. 29.8%), and less often on the circumflex artery (10.8% vs. 16.8%). The reported visual estimate of stenoses, both pre- (83.0%) and post-procedure (4.0%), was approximately the same for men as for women, as was the deployment of stents, which was about 95%. Examination of  the “TIMI flow” values, the standard grading system for describing intra-coronary blood flow (The TIMI Study Group, 1984), revealed that the vast majority of patients had TIMI 3 flow (indicating normal flow down the artery) before and after the intervention (81.3% and 88.5%, respectively), but a higher   77 proportion of men displayed TIMI flow of less than 3. The women, compared with the men, had, on average, shorter fluouroscopy time (11.3 vs. 13.5 minutes, respectively) and shorter balloon inflation time (63.7 vs. 77.6 seconds, respectively). Rates of procedural complications (e.g., arrhythmias, access site complications or coronary artery dissection) were less than 5% overall, and were similar for men and women. Revascularisation was deemed complete for the vast majority of the participants (83.9%). Finally, more women had complete revascularisation than did men (86.0% vs. 82.6%). Subjective Experience of Myocardial Ischemia during Balloon Inflation  Findings addressing research question number 1, “What is the subjective experience of myocardial ischemia during a period of known, controlled myocardial ischemia?” are now presented, beginning with (a) the frequency of reporting any symptoms. Next, analyses of the remaining components of the subjective experience are presented: (b) the number of symptoms reported; (c) the intensity of symptoms reported, and (d) the location and type of symptom reported. For each of these components, data from both baseline and balloon inflation are presented. Findings related to sex/gender differences are presented in the same table, to permit easy comparison. Finally, an examination of the validity of balloon inflation as a model for spontaneous ischemia will be presented. Frequency of Reporting Any Symptoms Though not presented in tabular form, the data revealed that all but four of the participants had had symptoms before their referral for PCI. Data regarding the frequency of reporting any symptoms during balloon inflation are presented in Table 8, both for those with and without ECG-confirmed ischemia. Although virtually all of the participants (99%) reported having had symptoms at baseline, during balloon inflation, 233 participants (76.4%) reported   78 one symptom or more. This rate was the same for those with confirmed ischemia (76.3%) during balloon inflation, and for those without (76.7%).  Table 8  Presence of Any Reported Symptoms during Balloon Inflation, by Gender All (N = 305)  Documented Ischemiaa (n = 245)  No Documented Ischemia (n = 60) Gender Yes n (%) χ2LR (df) p  Yes n (%) Χ2LR (df) p  Yes n (%) χ2LR (df) p Women (%) 97 (80.2)    80 (79.2)    17 (85.0) Men (%) 136 (73.9)    107 (74.3)    29 (72.5)  1.61 (1) 0.21   0.80 (1) 0.37   1.23 (1) 0.27 a Criterion for ischemia: ≥ 1mm ST-segment deviation in any ECG lead.  Sex/Gender Differences in Participants’ Reporting of Any Symptoms When the frequency of reporting any symptoms was stratified by sex/gender, for those with ischemia, and for those without ischemia, no statistically significant differences between men and women were found (see Table 8). Number of Symptoms Reported Tables 9 and 10 show the number of symptoms reported by all participants at baseline and those with confirmed ischemia during balloon inflation, respectively. The mean number of symptoms reported by this sub-sample was greater for the baseline event than for the balloon inflation experience (3.8 vs. 1.3). Sex/Gender Differences in the Number of Symptoms Reported Comparison of the median number of symptoms reported by the women and men revealed that the women reported more symptoms at baseline (4 vs. 3, respectively, p = 0.01) but not during balloon inflation (Mdn for both = 1, p = >0.05). This is shown in Tables 9 and 10.   79 Table 9  Number of Reported Symptoms at Baseline, by Gender Number of Symptoms Gender M SD Mdn Mann-Whitney U p All (N = 305) 4 2 4 Women (n = 121) 4 2 4 Men (n = 184) 4 2 3 9221.00 0.01 Note. Includes only participants with documented ischemia. Criterion for ischemia: ≥ 1 mm. ST-segment deviation in any ECG lead.  Table 10  Number of Reported Symptoms during Balloon Inflation, by Gender Number of Symptoms Gender M SD Mdn Mann-Whitney U p All (N = 245) 1 1 1 Women (n = 101) 1 1 1 Men (n = 144) 1 1 1 6396.50 0.09 Note. Includes only participants with documented ischemia. Criterion for ischemia: ≥ 1 mm. ST-segment deviation in any ECG lead.  Intensity of Reported Symptoms Tables 11 and 12 and Figures 2 and 3 summarise the symptom intensity data for baseline and balloon inflation, both including and excluding those who reported “no discomfort” during balloon inflation (i.e., scored zero in intensity).   80 Table 11  Intensity of Main Symptom, by Gender (including those reporting no inflation discomfort) Baseline (n = 304)  Inflation (n = 243) Gender Mean (SD) ta (df) p  Mean (SD) ta (df) P All 6.4 (2.5)    3.8 (2.9) Women (n = 120 baseline; 101 inflation) 6.8 (2.3)  4.6 (3.2) Men (n = 184 baseline; 142 inflation) 6.1 (2.6) 2.14 (229.56) 0.03  3.2 (2.6) -3.88 (185.80) <0.001 Note. Inflation subsample includes only participants with ischemia. Criterion for ischemia: ≥ 1mm ST-segment deviation in any ECG lead. Number analysed varies due to missing intensity data. a Equal variances not assumed   Table 12  Intensity of Main Symptom, by Gender (excluding those reporting no discomfort at inflation) Note. Inflation subsample includes only participants with ischemia; criterion for ischemia: ≥ 1mm ST-segment deviation in any ECG lead a Equal variances not assumed Baseline (n = 300)  Inflation (n = 185) Gender Mean (SD) ta (df) p  Mean (SD) ta (df) p All 6.5 (2.4) - -  5.0 (2.3) - - Women (n = 119 baseline, 80 inflation) 6.8 (2.2)  5.9 (2.4) Men (n = 181,105) 6.2 (2.5) -2.188 (274.3) 0.03  4.3 (2.0) 4.82 (155.7) <0.001   81    Figure 2  Intensity of main reported symptom at baseline, by gender (with and without those reporting no discomfort).   82  Figure 3  Intensity of main reported symptom during balloon inflation, by gender (with and without those reporting no discomfort).   83 The overall mean intensity score reported was approximately 6.5 on a scale of 0 to 10, at baseline, both when those reporting no discomfort were included and excluded (only three participants reported that they had had no discomfort at baseline). During inflation, the mean intensity reported was 3.8 (SD = 2.9) on a scale of 0 to 10 when “no discomfort” scores were included, and 5.0 (SD = 2.3) on a scale of 1 to 10 when they were excluded. The greater intensity of the reported baseline symptoms compared with the balloon inflation symptoms was statistically significant, regardless of whether participants with no discomfort were included. Sex/Gender Differences in the Intensity of Reported Symptoms Tables 11 and 12 also display gender-specific symptom intensity scores, and tests for differences between those scores. Women reported statistically significantly higher intensity of symptoms both at baseline and during inflation, whether or not the analysis included those with no discomfort. As mentioned previously, the inclusion of the “no discomfort” scores rendered the distribution skewed (skewness 0.186; SE 0.16). Therefore, a Kolmogorov-Smirnov test was also conducted to ascertain whether the scores differed between men and women. This test was also statistically significant (Z = 1.93, p = < 0.001). Location and Type of Reported Symptoms The participants gave a total of 202 unique descriptions of the symptoms they experienced either at baseline (before their referral for PCI) or during the balloon inflation period (139 and 63, respectively). As outlined in Chapter 3, because many of these descriptors overlapped considerably (e.g., “across front of chest”, “all over chest”, and “across chest”; “across top of chest” and “across upper chest”; or “sweating” and “sweaty”), they were grouped, using clinical and etymological knowledge, resulting in 49 symptoms overall (see Appendix D). Very little interpretation was imposed. Table 13 lists the percentages of participants who reported each of the derived symptom categories as the main symptoms that they had   84 experienced both before referral for PCI (that is, the responses they gave to the question, “Thinking back over the last several months, tell me where the main location has been for any discomfort you have had”), and at balloon inflation (using the same question, altered to pertain to the balloon inflation situation). There were originally 97 unique descriptors of main discomfort at baseline and 51 during inflation. These yielded 23 categories of main discomfort.   85 Table 13  Location/Type and Frequency of Reported Main Symptoms, by Gender Baseline (N = 305)  Balloon Inflation (n = 245)  Discomfort or Symptom All (n (%)) Women (n (%)) (n = 121) Men (n (%)) (n = 184)  All (n (%)) Women (n (%)) (n = 101) Men (n (%)) (n = 144) Chest, not right-sided 237 (77.7) 93 (76.9) 144 (78.3)  132 (53.9) 51 (50.5) 81 (56.2) Right chest 2 (0.7)  2 (1.1)  5 (2.0) 1 (1.0) 4 (2.8) Left arm 5 (1.6) 2 (1.7) 3 (1.6)  3 (1.2) 1 (1.0) 2 (1.4) Right arm 3 (1.0)  3 (1.6) Both arms 4 (1.3) 2 (1.7) 2 (1.1)  2 (0.8)  2 (1.4) Left shoulder 11 (3.6) 5 (4.1) 6 (3.3)  1 (0.4) 1 (1.0) Intra-scapular 4 (1.3) 2 (1.7) 2 (1.1)  2 (0.8)  2 (1.4) Both shoulders     1 (0.4)  1 (0.7) Axilla(e) 1 (0.3)  1 (0.5)  1 (0.4)  1 (0.7) Jaw / teeth 3 (1.0) 1 (0.8) 2 (1.1)  7 (2.8) 4 (4.0) 3 (2.1) Throat 6 (2.0) 4 (3.3) 2 (1.1)  16 (6.5) 10 (9.9) 6 (4.2) Neck 3 (1.0) 3 (2.5)   5 (2.0) 5 (5.0) Back of head     2 (0.8) 1 (1.0) 1 (0.7) Back 4 (1.3) 2 (1.7) 2 (1.1)  4 (1.6) 2 (2.0) 2 (1.4) Ribs 3 (1.0)  3 (1.6) Epigastrium/indigest’n/heartburn 6 (2.0) 3 (2.5) 3 (1.6)  1 (0.4) 1 (1.0) Note. “Balloon inflation” includes only participants with ischemia.  Criterion for ischemia: ≥1 mm ST-segment deviation in any ECG lead.  86 Table 13 (cont’d) Location/Type and Frequency of Reported Main Symptoms, by Gender Baseline (N = 305)  Balloon Inflation (n = 245) Discomfort or Symptom (%) All (n (%)) Women (n (%)) (n = 121) Men (n (%)) (n = 184)  All Women (n (%)) (n = 121) Men (n (%)) (n = 184) Abdomen 1 (0.3)  1 (0.5)  1 (0.4)  1 (0.7) Right Leg     1 (1.4) 1 (1.0) Aneurysm     1 (0.4)  1 (0.7) Shortness of breath 8 (2.6) 3 (2.5) 5 (2.7)  1 (0.4) 1 (1.0) Fatigue     1 (0.4) 1 (1.0) Vomiting 1 (0.3)  1 (0.5) No discomfort 3 (1.0) 1 (0.8) 2 (1.1)  58 (23.7) 21 (20.8) 37 (25.7) Note. “Balloon inflation” includes only participants with ischemia.  Criterion for ischemia: ≥1 mm ST-segment deviation in any ECG lead.   87 Main Symptoms: Baseline With respect to the baseline symptoms, 18 unique main symptoms were reported by at least one person. Non-right-sided chest pain was the most common symptom (77.7%), and of the remaining 17 categories, only discomfort in the left shoulder and shortness of breath exceeded a rate of 2% (3.6% and 2.6%, respectively). No participant described discomfort in both shoulders, the back of the head, abdomen, right leg or “aneurysm”, or fatigue as a main symptom at baseline (though these were reported as main symptoms during balloon inflation). Three participants reported having experienced “no discomfort” before their referral for PCI. Main Symptoms: Balloon Inflation Examination of the main symptoms reported during the balloon inflation revealed that the unique descriptors were fewer for this observation than for those reported as baseline symptoms (approximately 50). However, after refining the raw data, symptoms occurred in all 23 of the final symptom categories. Chest pain (non-right-sided) was again the most commonly reported symptom, and was reported by the majority of participants (53.9%), followed by no discomfort (23.7%), throat (6.5%), jaw or teeth (2.8%), and right-sided chest discomfort (2.0%). During this observation period, no participant reported discomfort in the right arm or the ribs, or vomiting as a main symptom (though these were reported as main symptoms that had been experienced at baseline). The bivariate analyses of these symptoms and sex/gender are described in a later section. All Symptoms: Baseline As outlined in Chapter 3, the research assistants continued to question the participants about their discomfort after they answered the first question pertaining to their main discomfort. Questioning continued until the participant indicated there were no more symptoms to report.   88 An inclusive listing of all symptoms reported by the participants both at baseline and during balloon inflation, and their frequency, is presented as Table 14.  89 Table 14  Location/Type and Frequency of All Reported Symptoms, by Gender Baseline (N = 305)  Balloon Inflation (n = 245) Discomfort or Symptom All (n (%)) Women (n (%)) Men (n (%))  All (n (%)) Women (n (%)) Men (n (%)) Chest (not right) 265 (86.9) 98 (81.0) 167 (90.8)  142 (58.0) 55 (54.5) 87 (60.4) Shortness of breath 153 (50.2) 65 (53.7) 88 (47.8)  2 (0.8) 2 (2.0) Fatigue 105 (34.4) 45 (37.2) 60 (32.6)  4 (1.6) 2 (2.0) 2 (1.4) Diaphoresis 101 (33.1) 37 (30.6) 64 (34.8)  6 (2.4) 4 (4.0) 2 (1.4) Left arm 52 (17.0) 27 (22.3) 25 (13.6)  15 (6.1) 7 (6.9) 8 (5.6) Both arms 59 (19.3) 20 (16.5) 39 (21.2)  6 (2.4) 1 (1.0) 5 (3.5) Throat 41 (13.4) 25 (20.7) 16 (8.7)  40 (16.3) 24 (23.8) 16 (11.1) Left shoulder 26 (8.5) 12 (9.9) 14 (7.6)  3 (1.2) 1 (1.0) 2 (1.4) Both shoulders 13 (4.3) 7 (5.8) 6 (3.3)  5 (2.0) 1 (1.0) 4 (2.8) Jaw/teeth 39 (12.8) 24 (19.8) 15 (8.2)  15 (6.1) 10 (9.9) 5 (3.5) Neck 18 (5.9) 12 (9.9) 6 (3.3)  12 (4.9) 9 (8.9) 3 (2.1) Intrascapular 9 (3.0) 5 (4.1) 4 (2.2)  2 (0.8)  2 (1.4) Epigastrium/indigest’n/heartburn 68 (22.3) 25 (20.7) 43 (23.4)  3 (1.2) 1 (1.0) 2 (1.4) Back 32 (10.5) 21 (17.4) 11 (6.0)  9 (3.7) 5 (5.0) 4 (2.8) Nausea 50 (16.4) 26 (21.5) 24 (13.0)  7 (2.9) 2 (2.0) 5 (3.5) No discomfort 3 (1.0) 1 (0.8) 2 (1.1)  58 (23.7) 21 (20.8) 37 (25.7) Note. “Balloon inflation” includes only participants with ischemia. Criterion for ischemia: ≥ 1mm ST-segment deviation in any ECG lead.  90 Table 14 (cont’d) Location/Type and Frequency of All Reported Symptoms, by Gender Baseline (N = 305)  Balloon Inflation (n = 245) Discomfort or Symptom All (n (%)) Women (n (%)) Men (n (%))  All (n (%)) Women (n (%)) Men (n (%)) Vomiting 1 (0.3)  1 (0.5) Abdomen 1 (0.3)  1 (0.5)  2 (0.8)  2 (1.4) Axilla 3 (1.0) 1 (0.8) 2 (1.1)  1 (0.4)  1 (0.7) Right arm 9 (3.0) 4 (3.3) 5 (2.7)  10 (4.1) 4 (4.0) 6 (4.2) Right side chest 2 (0.7)  2 (1.1)  5 (2.0) 1 (1.0) 4 (2.8) Dizziness 81 (26.6) 39 (32.2) 42 (22.8)  3 (1.2) 2 (2.0) 1 (0.7) Hands 3 (1.0)  3 (1.6)  1 (0.4)  1 (0.7) Ears 3 (1.0) 2 (1.7) 1 (0.5) Lungs 1 (0.3) 1 (0.8) Palpitations 1 (0.3) 1 (0.8) Left wrist 1 (0.3)  1 (0.5) Salivation 2 (0.7)  2 (1.1) Hot flashes 1 (0.3) 1 (0.8) Cramp 1 (0.3)  1 (0.5) Diarrhea 4 (1.3) 2 (1.7) 2 (1.1) Anxiety 2 (0.7) 1 (0.8) 1 (0.5)  1 (0.4) 1 (1.0) Note. “Balloon inflation” includes only participants with ischemia. Criterion for ischemia: ≥ 1mm ST-segment deviation in any ECG lead.  91 Table 14 (cont’d) Location/Type and Frequency of All Reported Symptoms, by Gender Baseline (N = 305)  Balloon Inflation (n = 245) Discomfort or Symptom All (n (%)) Women (n (%)) Men (n (%))  All (n (%)) Women (n (%)) Men (n (%)) Left leg 2 (0.7) 1 (0.8) 1 (0.5)    0 Right leg 3 (1.0) 2 (1.7) 1 (0.5)  3 (1.2) 2 (2.0) 1 (0.7) Left clavicle 1 (0.3) 1 (0.8) Tongue 1 (0.3) 1 (0.8) Lips 1 (0.3) 1 (0.8) Dry mouth 1 (0.3)  1 (0.5)  1 (0.4)  1 (0.7) Cough 1 (0.3)  1 (0.5) Restlessness 1 (0.3) 1 (0.8) Cold feet 1 (0.3)  1 (0.5) Shaking 1 (0.3) 1 (0.8) Back of head     2 (0.8) 1 (1.0) 1 (0.8) Aneurysm     1 (0.4)  1 (0.7) Headache 5 (1.6) 4 (3.3) 1 (0.5)  4 (1.6) 1 (1.0) 3 (2.1) Cheeks     1 (0.4) 1 (1.0) Note. “Balloon inflation” includes only participants with ischemia. Criterion for ischemia: ≥ 1mm ST-segment deviation in any ECG lead.   92 The entire array of descriptors used to describe the sensations (202) fell into 46 symptom categories after refinement. Each of these was used by at least one participant to describe their baseline symptoms, whereas 29 were used by at least one participant to describe the sensations felt during balloon inflation. At baseline, the most commonly reported symptom was non-right-sided chest discomfort (86.9%), followed by shortness of breath, fatigue, diaphoresis, dizziness, indigestion or epigastric discomfort, discomfort in both arms, nausea, throat and jaw discomfort. All Symptoms: Balloon Inflation During the balloon inflation, non-right-sided chest pain dominated the list of descriptors (58.0%), followed by no discomfort, then throat, left-arm, and neck discomfort. Although a further 24 symptoms were reported during inflation, none of them was reported by greater than 5.0% of the sample. Sex/Gender Differences in the Location or Type of Symptoms Reported Main Symptoms Some symptoms were reported by only one gender: no woman reported discomfort in the right side of the chest, right arm, underarms, ribs or abdomen, or vomiting as her main baseline symptom, whereas no man reported neck discomfort as his main symptom at baseline. For the main symptoms during the balloon inflation, only men reported bilateral-arm, intra- scapular, bilateral-shoulder, underarm, abdominal or “aneurysm” discomfort, and only women reported left shoulder, neck, epigastric or right-leg discomfort, or shortness of breath or fatigue. All symptoms With respect to the list of all symptoms reported, no woman reported right-sided chest, hand, left wrist or abdominal discomfort; or vomiting, salivation, dry mouth, cough, cramping or cold feet at baseline. No man reported discomfort in the lungs, left clavicle, tongue, or lips, or   93 shaking, restlessness, hot flashes, or palpitations, at that time. During the balloon inflation, no woman reported intrascapular, axillary, hand, or “aneurysm” discomfort, or dry mouth. No man reported shortness of breath, anxiety, or cheek discomfort. Chest Discomfort Sex/gender differences in the location and type of symptoms reported were further explored with bivariate analyses. The symptom of chest discomfort was explored first because there is great reliance placed on the meaning of this symptom by health professionals and potential patients alike (Canto et al., 2007; Pope et al., 2000). Tables 15 to 18 summarise the frequency of non-right-sided chest discomfort as the main reported symptom and as a symptom ever reported, both at baseline and balloon inflation. No statistically significant sex/gender differences were apparent when examining it as the main symptom, at either observation. However, when considering all symptoms reported, there were significantly fewer women than men who reported chest discomfort pertaining to the baseline event (p = 0.02), but this statistical significance did not occur in the analyses related to the balloon inflation period, when the presence of ischemia was verified.  Table 15  Frequency of Chest Discomfort Reported as Main Symptom at Baseline, by Gender Gender Chest Discomfort Reported as Main Symptom (n (%)) χ2LR (df) p All (N = 305) 166 (54.4) Women (n = 121) 64 (52.9) Men (n = 184) 102 (55.4) 0.19 (1) 0.66    94 Table 16  Frequency of Ever-Reported Chest Discomfort at Baseline, by Gender Gender Chest Discomfort Reported as Main Symptom (n (%)) χ2LR (df) p All (N = 305) 265 (86.9) Women (n = 121) 98 (81.0) Men (n = 184) 167 (90.8) 5.97 (1) 0.02 Note. Includes only participants with ischemia only. Criterion for ischemia 1 mm ST deviation in any ECG lead.  Table 17  Frequency of Chest Discomfort Reported as Main Symptom during Balloon Inflation, by Gender Gender Chest Discomfort Reported as Main Symptom (n (%)) χ2LR (df) p All (N = 245) 131(53.5) Women (n = 101) 51 (50.5) Men (n = 144) 80 (55.6) 0.61 (1) 0.43 Note. Includes participants with ischemia only. ECG criteria for ischemia 1 mm ST deviation in any lead.  Table 18  Frequency of Ever-Reported Chest Discomfort during Balloon Inflation, by Gender Gender Chest Discomfort Ever- Reported (n (%)) χ2LR (df) p All (N = 245) 142 (58.0) Women (n = 101) 55 (54.5) Men (n = 144) 87 (60.4) 0.86 (1) 0.35 Note. Includes participants with ischemia only. ECG criteria for ischemia 1 mm ST deviation in any lead.  Typical Symptoms The reporting of any “typical” symptoms was examined for possible sex/gender differences (see Tables 19 to 22). As described in Chapter 3, the definition of typical was drawn from widely-accepted clinical practice guidelines for managing patients with stable and unstable angina, and MI (Anderson et al., 2007; Antman et al., 2004; Bassand et al., 2007; The Task Force   95 on the Management of Stable Angina Pectoris of the European Society of Cardiology, 2006). This analysis revealed no statistically significant differences between men and women: this was true when considering either only the main symptoms reported or all symptoms reported, at either baseline or balloon inflation.  Table 19  Frequency of Any “Typical” a Symptom Reported as Main Symptom at Baseline, by Gender Gender “Typical”a Symptoms Present (n (%)) χ2LR (df) p All (N = 305) 288 (94.4) Women (n = 121) 113 (93.4) Men (n = 184) 175 (95.1) 0.40 (1) 0.53 Note. Includes participants with ischemia only. ECG criteria for ischemia 1 mm ST deviation in any lead. a “Typical” symptoms include: chest, arm(s), shoulder, jaw, back or epigastric discomfort; shortness of breath; diaphoresis; weakness; indigestion; nausea &/or vomiting; dizziness &/or lightheadedness; fear or restlessness  Table 20  Frequency of “Typical” Ever-Reported Symptoms at Baseline, by Gender Gender “Typical”a Symptoms Present (n (%)) χ2LR (df) p All (N =  305) 300 (98.4) Women (n=121) 119 (98.3) Men (n=184) 181 (98.4) 0.00 (1) 0.99 Note. Includes participants with ischemia only. ECG criteria for ischemia 1 mm ST deviation in any lead. a “Typical” symptoms include: chest, arm(s), shoulder, jaw, back or epigastric discomfort; shortness of breath; diaphoresis; weakness; indigestion; nausea&/or vomiting; dizziness &/or lightheadedness; fear or restlessness    96 Table 21  Frequency of Any “Typical” a Symptom Reported as Main Symptom during Balloon Inflation, by Gender  Gender “Typical”a Symptoms Present (n (%)) χ2LR (df) p Both (n = 245) 157 (64.1) Women (n = 101) 66 (65.3) Men (n = 144) 91 (63.2) 0.12 (1) 0.73 Note. Includes participants with ischemia only. ECG criteria for ischemia 1 mm ST deviation in any lead. a “Typical” symptoms include: chest, arm(s), shoulder, jaw, back or epigastric discomfort; shortness of breath; diaphoresis; weakness; indigestion; nausea&/or vomiting; dizziness &/or lightheadedness; fear or restlessness.  Table 22 Frequency of “Typical” Ever-Reported Symptoms during Balloon Inflation, by Gender Gender “Typical”a Symptoms Present (n (%)) χ2LR (df) p Both (n = 245) 171 (69.8) Women (n = 101) 71 (70.3) Men (n = 144) 100 (69.4) 0.02 (1) 0.89 Note. Includes participants with ischemia only. ECG criteria for ischemia 1 mm ST deviation in any lead. a “Typical” symptoms include: chest, arm(s), shoulder, jaw, back or epigastric discomfort; shortness of breath; diaphoresis; weakness; indigestion; nausea&/or vomiting; dizziness &/or lightheadedness; fear or restlessness.  Milner’s Typical Symptoms The association between reports of typical symptoms and gender was also examined using the operational definition employed by Milner et al. (2002). Tables 23 to 25 outline these findings. No measures of association were calculated for ever-reported symptoms at baseline because all participants with ischemia reported having had at least one “Milner-typical” symptom at baseline. For the remainder of the events examined in these analyses, no statistically significant differences were found between men and women.   97 Table 23  Frequency of Any of “Milner-Typical”  Symptoms Reported as Main Symptom at Baseline, by Gender Gender “Milner-Typical”a Symptoms Present (n (%)) χ2LR (df) p All (N = 305) 269 (88.2) Women (n = 121) 108 (89.3) Men (n = 184) 161 (87.5) 0.22 (1) 0.64 Note. Includes participants with ischemia only. ECG criteria for ischemia 1 mm ST deviation in any lead. a Milner-typical” symptoms: chest, arm(s), jaw, throat or neck discomfort; shortness of breath; diaphoresis.  Table 24  Frequency of Any “Milner-Typical” Symptoms Reported as Main Symptom at Balloon Inflation, by Gender Gender “Milner-Typical” a Symptoms Present (n (%)) χ2LR (df) p All (n = 245) 166 (67.8) Women (n = 101) 72 (71.3) Men (n = 144) 94 (65.3) 0.99 (1) 0.32 Note. Includes participants with ischemia only. ECG criteria for ischemia 1 mm ST deviation in any lead. a Milner-typical” symptoms: chest, arm(s), jaw, throat or neck discomfort; shortness of breath; diaphoresis.  Table 25  Frequency of Any “Milner-Typical” Ever-Reported Symptoms during Balloon Inflation, by Gender Gender “Milner-Typical” a Symptoms Present (%) χ2LR (df) p All (n = 245) 177 (72.0) Women (n = 101) 74 (30.2) Men (n = 144) 103 (42.0) 0.09 (1) 0.76 Note. Includes participants with ischemia only. ECG criteria for ischemia 1 mm ST deviation in any lead. a” Milner-typical” symptoms: chest, arm(s), jaw, throat or neck discomfort; shortness of breath; diaphoresis.   98 Only Non-Chest Discomfort Examining the types of reported symptoms from a different vantage point, the frequency of reporting only non-chest discomfort symptoms is displayed in Tables 26 and 27.  Table 26  Frequency of Reporting Only Non-Chest-Discomfort Symptoms at Baseline, by Gender Gender Only Non-Chest Discomfort Symptom(s) Present (%) χ2LR (df) p All (n = 305) 37 (12.1) Women (n = 121) 22 (18.2) Men (n = 184) 15 (8.2) 6.72 (1) 0.01 Note. Includes participants with ischemia only. ECG criteria for ischemia 1 mm ST deviation in any lead.  Table 27  Frequency of Reporting Only Non-Chest-Discomfort Symptoms during Balloon Inflation, by Gender Gender Only Non-Chest Discomfort Symptom(s) Present (%) χ2LR (df) p Both (n = 245) 45 (18.4) Women (n = 101) 25 (24.8) Men (n = 144) 20 (13.9) 4.604 (1) 0.03 Note. Includes participants with ischemia only. ECG criteria for ischemia 1 mm ST deviation in any lead.   This type of symptom report had lower frequency at baseline than during balloon inflation (12% vs. 18%, respectively), and women reported only non-chest symptoms statistically significantly more than did men at both times (p = < 0.05 for both comparisons). All Remaining Symptoms: Approach to Analysis The next analysis involved examining all remaining ever-reported symptoms for possible sex/gender differences. The entire symptoms list resulted in 92 more analyses: each of 46 symptoms at both baseline and balloon inflation. As has been stressed before, the initial intent   99 was to limit detailed analyses of reported symptoms to data from the sub-sample of participants with documented ischemia, focusing primarily on the balloon inflation phase. However, for reasons that will be described in more detail later, I examined the baseline symptom data also, for latent classes. In light of this, it was useful to also include an analysis of the association of sex/gender with selected symptoms reported at baseline. This is presented in Table 28. These particular variables were selected for presentation because they were of known clinical importance, previous investigators had found sex/gender differences in the reporting of these symptoms, or because the bivariate analysis produced a p-value of 0.25 or less, and therefore warranted further exploration using multivariate analyses (Hosmer & Lemeshow, 2004). The symptoms for which the relationship with sex/gender reached statistical significance of 0.25 or less were explored using logistic regression analysis. Covariates included in the models were determined by first examining the bivariate relationships between each such symptom and all available demographic and clinical variables for which there was less than two percent missingness. Covariates were considered appropriate for inclusion in each particular symptom prediction model if they had reached a significance of 0.25 or less in the bivariate analysis of that symptom; these covariates varied for each symptom. These adjusted analyses are also displayed in Table 28, and covariates other than gender that reached significance are noted. Next, focusing on symptoms during balloon inflation only, for which there could be ECG confirmation of ischemia, bivariate analyses of selected symptoms are presented in Table 29. These particular symptoms were selected in the same fashion as were those for baseline symptoms, and selection of covariates for multivariate analysis also followed the procedure described above.   100 Table 28  Association of Gender with Selected Symptoms Reported at Baseline (unadjusted and adjusted; men as referent) Gender Discomfort or Symptom Women (n (%)) (n = 121) Men (n (%)) (n = 184) Unadjusted OR (95% CI) Adjusted OR (95% CI) No discomfort 1 (0.8) 2 (1.1) 0.76 (0.01-14.72)* - Localised Discomfort Chest, not right-sided 98 (81.0) 167 (90.8) 0.43 (0.22-0.85)  0.33 (0.16-0.67) U, BMI, V Left arm 27 (22.3) 25 (13.6) 1.83 (1.00-3.33)  1.71 (0.93-3.14) E, ANTI-D Right arm 4 (3.3) 5 (2.7) 1.22 (0.24-5.81 * - Both arms 20 (16.5) 39 (21.2) 0.74 (0.41-1.34) 0.77 (0.42-1.42) Left shoulder 12 (9.9) 14 (7.6) 1.34 (0.60-3.00)  1.95 (0.77-4.97) V, CHOL, NTG Both shoulders 7 (5.8) 6 (3.3) 1.82 (0.60-5.56)  2.49 (0.76-8.19) ACS, ANTI-D SOB = shortness of breath; OR = odds ratio; CI = confidence interval; *Exact Fisher CI (expected cell frequency less than 5); ‡may have reached significance of p </ =  0.25, but insufficient responses to permit analysis; Significant co-variates: Aage; ACEreceiving angiotensin-converting enzyme inhibitor; ACSacute coronary syndrome main indication for procedure; ANTI-C receiving anticoagulant agent; ANTI-D receiving antidepressant agent; BMIbody mass index; CCSCanadian Cardiovascular Society grade of angina; CHOLblood cholesterol level; CLPreceiving clopidogrel; Ddiabetes;  Eeducation level; HRTreceiving hormone replacement therapy; LLreceiving lipid-lowering agent; MIprior MI; NTG receiving IV nitroglycerin; PCIprior percutaneous coronary intervention; PVDperipheral vascular disease; SMKsmoking status; Uurgency;  Vvessel   101 Table 28 (cont’d) Association of Gender with Selected Symptoms Reported at Baseline (adjusted and unadjusted, men as referent) Gender Discomfort or Symptom Women (%) (n = 121) Men (%) (n = 184) Unadjusted OR (95% CI) Adjusted OR (95% CI) Localised Discomfort Throat 25 (20.7) 16 (8.7) 2.73 (1.40-5.37) 3.11 (1.52-6.35)ACE Jaw/Teeth 24 (18.8) 15 (8.2) 2.79 (1.40-5.57) 2.93 (1.44-5.95)V Neck 12 (9.9) 6 (3.3) 3.27 (1.19-8.95)  2.47 (0.86-7.13) HRT, SMK Throat, Jaw/Teeth or Neck 46 (38.0) 32 (17.4) 2.91 (1.72-4.95) 3.24 (1.86-5.64)NTG, V Back 21 (17.4) 11 (6.0) 3.30 (1.53-7.13)  3.12 (1.42-6.87) SMK, ANTI-C Epigastrium, Indigest’n, Heartburn 25 (20.7) 43 (23.4) 0.85 (0.49-1.49)  0.99 (0.55-1.81) A, ANTI-C Between Shoulders 5 (4.1) 4 (2.2) 1.94 (0.41-9.96)* - Abdomen 0 1 (0.5) - - Axilla 1 (0.8) 2 (1.1) -‡ - Hands 0 3 (1.6) - - Cheeks 0 0 - - Headache 4 (3.3) 1 (0.5) 6.22 (0.61-309.25)* - SOB = shortness of breath; OR = odds ratio; CI = confidence interval; *Exact Fisher CI (expected cell frequency less than 5); ‡may have reached significance of p </= 0.25, but insufficient responses to permit analysis; Significant co-variates: Aage; ACEreceiving angiotensin-converting enzyme inhibitor; ACSacute coronary syndrome main indication for procedure; ANGstable angina main indication for procedure; ANTI-C receiving anticoagulant agent; ANTI-D receiving antidepressant agent; BBreceiving beta blockers; CABGprior CABG; CLPreceiving clopidogrel; Ddiabetes; Eeducation level; HFcomorbid heart failure; HRTreceiving hormone replacement therapy; LLreceiving lipid-lowering agent; MIprior MI; N receiving IV nitroglycerin; PCIprior percutaneous coronary intervention; PVDperipheral vascular disease; SMKsmoking status; Uurgency;  Vvessel   102 Table 28 (cont’d) Association of Gender with Selected Symptoms Reported at Baseline (adjusted and unadjusted, men as referent) Gender Discomfort or Symptom Women (%) (n = 121) Men (%) (n = 184) Unadjusted OR (95% CI) Adjusted OR (95% CI) Non-Localised Symptoms SOB 65 (53.7) 88 (47.8) 1.27 (0.80-2.01) 1.05 (0.63-1.74)ANG Fatigue 45 (37.2) 60 (32.6) 1.22 (0.76-1.98) 1.35 (0.82-2.24) Cough 0 1 (0.5) -‡ - Diaphoresis 37 (30.6) 64 (34.8) 0.83 (0.51-1.35) 0.84 (0.49-1.41) U, V Dizziness 39 (32.2) 42 (22.8) 1.61 (0.96-2.69) 1.61 (0.96-2.69) Nausea 26 (21.5) 24 (13.0) 1.83 (0.99-3.36) 1.73 (0.87-3.42) ACS, SMK Vomiting 0 1 (0.5) -‡ - Palpitations 1 (0.8) 0 -‡ - Anxiety 1 (0.8) 1 (0.5) -‡ - Grouped Symptoms Only non-chest symptoms 22 (18.2) 15 (8.2) 2.50 (1.24-5.05)  3.28 (1.52-7.08) D, PCI, CABG, BB, HF Any “typical” symptoms 119 (98.3) 181 (98.4) 0.99 (0.16-5.99) 0.97 (0.16-5.94) SOB = shortness of breath; OR = odds ratio; CI = confidence interval; *Exact Fisher CI (expected cell frequency less than 5); ‡may have reached significance of p </= 0.25, but insufficient responses to permit analysis; Significant co-variates: Aage; ACEreceiving angiotensin-converting enzyme inhibitor; ACSacute coronary syndrome main indication for procedure; ANTI-C receiving anticoagulant agent; ANTI-D receiving antidepressant agent; BMIbody mass index; CHOLblood cholesterol level; CLPreceiving clopidogrel; Ddiabetes;  Eeducation level; HRTreceiving hormone replacement therapy;  LLreceiving lipid-lowering agent; MIprior MI; NTG receiving IV nitroglycerin; PCIprior percutaneous coronary intervention; PVDperipheral vascular disease; SMKsmoking status; Uurgency;  Vvessel   All Remaining Symptoms: Baseline Bivariate analyses of the baseline data revealed statistically significant sex/gender differences in six symptom reports: discomfort in the chest, left arm, throat, jaw, neck, back, and   103 the compilation of throat, jaw and neck; and reporting only non-chest discomfort. When adjusted for important socio-demographic and clinical covariates as described above, sex/gender remained a predictor in all symptoms except left arm and neck discomfort. The magnitude of the effect of sex/gender was increased for throat (OR 3.11; 95% CI: 1.52-6.38); jaw (OR 2.93; 95% CI: 1.44-5.95); combined throat, jaw and neck discomfort (OR 3.24; 95% CI: 1.86-5.64), and for reporting only non-chest discomfort (OR 3.28; 95% CI: 1.52-7.08) when these covariates were controlled. The sex/gender effect was diminished (though still statistically significant) for chest (OR 0.33; 95% CI: 0.16-0.67) and back (OR 3.12; 95% CI: 1.42-6.87) discomfort when these covariates were controlled. The predictors that attenuated the effect of sex/gender in the models were, for chest discomfort: urgency of the procedure, body mass index, and target vessel; and, for back discomfort: smoking status and use of anti-coagulants within the previous 24 hours. Other predictors of statistical significance in the models were target vessel; use of IV nitroglycerin or beta blockers in the previous 24 hours; having had a prior revascularisation procedure; presence of any type of diabetes; and pre-existing heart failure.   104 Table 29  Association of Gender with Selected Symptoms Reported during Balloon Inflation (unadjusted and adjusted; men as referent) Gender Discomfort or Symptom Women (%) (n = 101) Men (%) (n = 144) Unadjusted OR (95% CI) Adjusted OR (95% CI) No discomfort 21 (20.8) 37 (25.7) 0.76 (0.41-1.40) - Localised Discomfort Chest, not right- sided 55 (54.5) 87 (60.4) 0.78 (0.47-1.31) 0.95 (0.52-1.74) A, PVD, CLP Left arm 7 (6.9) 8 (5.6) 1.27 (0.44-3.61) - Right arm 4 (4.0) 6 (4.2) 1.07 (0.15-6.48)* - Both arms 1 (1.0) 5 (3.5) 0.40 (0.04-2.14)* 0.40 (0.08-2.03) Left shoulder 1 (1.0) 2 (1.4) 0.71 (0.01-13.83)* - Both shoulders 1 (1.0) 4 (2.8) 0.35 (0.01-3.62)* - Throat 24 (23.8) 16 (11.1) 2.49 (1.25-4.99) 2.94 (1.44-6.03)A Jaw/Teeth 10 (9.9) 5 (3.5) 3.06 (1.01-9.23) 3.03 (0.99-9.31)PCI Neck 9 (8.9) 3 (2.1) 4.57 (1.10-26.93)* 7.80 (1.93-31.58)MI Throat, Jaw/Teeth or Neck 37 (36.6) 22 (15.3) 3.21 (1.75-5.89)  4.55 (2.31-8.98) A, PCI, ACE, CLP Back 5 (5.0) 4 (2.8) 1.82 (0.38-9.41)* - Epigastrium, Indigest’n, Heartburn 1 (1.0) 2 (1.4) 0.71 (0.01-13.83)* - Between Shoulders 0 2 (1.4) -‡ -‡ Note: Includes participants with documented ischemia only. SOB = shortness of breath; OR = odds ratio; CI = confidence interval; *Exact Fisher CI (expected cell frequency <5); ‡reached significance of p </= 0.25, but insufficient responses to permit analysis; significant co-variates: Aage; ACEreceiving an angiotensin-converting enzyme inhibitor; CLPreceiving clopidogrel; Ddiabetes; Eeducation level; LLreceiving lipid-lowering therapy; MIprior MI; PCIprior percutaneous coronary intervention; PVDperipheral vascular disease; STENpre-PCI stenosis; Uurgency   105 Table 29 (cont’d)  Association of Gender with Selected Symptoms Reported during Balloon Inflation (adjusted and unadjusted, men as referent) Gender Discomfort or Symptom Women (%) (n = 101) Men (%) (n = 144) Unadjusted OR (95% CI) Adjusted OR (95% CI) Localised Discomfort Abdomen 0 2 (1.4) -‡ -‡ Axilla 0 1 (0.7) - - Hands 0 1 (0.7) - - Cheeks 1 (1.0) 0 -‡ -‡ Headache 1 (1.0) 3 (2.1) 0.47 (0.01-5.97)* - Non-Localised Symptoms SOB 2 (2.0) 0 -‡ -‡ Fatigue 2 (2.0) 2 (1.4) 1.43 (0.10-20.06)* - Cough 0 0 - - Diaphoresis 4 (4.0) 2 (1.4) 2.91 (0.41-32.82)* 2.44 (0.34-17.61) SMK, STEN Dizziness 2 (2.0) 1 (0.7) 2.88 (0.15-171.43)* - Nausea 2 (2.0) 5 (3.5) 0.56 (0.05-3.52)* - Vomiting 0 0 - - Palpitations 0 0 - - Anxiety 1 (0.4) 0 -‡ -‡ Grouped Symptoms Only non-chest discomfort 25 (24.8) 20 (13.9) 2.04 (1.06-3.92) 2.11 (1.09-4.09)PVD Any “typical” symptoms 71 (70.3) 100 (69.4) 1.04 (0.60-1.81) 1.29 (0.70-2.40)A, PVD, PCI, LL Note: Includes participants with documented ischemia only. SOB = shortness of breath; OR = odds ratio; CI = confidence interval; *Exact Fisher CI (expected cell frequency <5); ‡reached significance of p </= 0.25, but insufficient responses to permit analysis; significant co-variates: Aage; ACEreceiving an angiotensin-converting enzyme inhibitor; CLPreceiving clopidogrel; Ddiabetes; Eeducation level; LLreceiving lipid-lowering therapy; MIprior MI; PCIprior percutaneous coronary intervention; PVDperipheral vascular disease; STENpre-PCI stenosis; Uurgency   106 All Remaining Symptoms: Balloon Inflation During balloon inflation, four reported symptoms demonstrated statistically significant sex/gender differences: throat, neck, jaw discomfort, and reporting only non-chest-pain discomfort.  When adjusted for potentially important covariates, sex/gender emerged as an even stronger predictor of the following reported symptoms: only non-chest pain discomfort (OR 2.11; 95% CI: 1.09-4.09); throat discomfort (OR 2.94; 95% CI: 1.44-6.03); and neck discomfort (OR 8.59; 95% CI: 1.87-39.43), but was slightly weaker, and losing statistical significance, for jaw or teeth discomfort, (OR 3.03; 95% CI: 0.99-9.31). Combining throat, jaw or neck discomfort indicated that women were more likely to experience any of this combination of symptoms than were men (OR 4.55; 95% CI: 2.31 - 8.98), the strength of this relationship again increasing with adjustment for covariates. In the model, age emerged as a significant predictor of having typical symptoms, chest, throat or neck discomfort, and the throat-jaw-neck discomfort compilation, with the odds of reporting these symptoms increasing with each decade of age. Increasing urgency of the PCI procedure (from elective to semi-urgent to urgent) predicted participants’ reporting of jaw discomfort, as did having had a prior PCI. Finally, not only age but also having had a prior MI predicted reporting of neck discomfort. Concordance of Reported Baseline Symptoms with Symptoms Reported During Balloon Inflation As a final step in the analysis of the symptom data, the validity of balloon inflation as a model of spontaneous ischemia was evaluated. For this analysis, only those participants in whom ischemia during balloon inflation was confirmed via ECG were included. Symptom dyads were cross-tabulated: those symptoms that participants reported having had at baseline (before referral for PCI), and those they reported during balloon inflation. From these contingency tables, percent agreement, kappa, and proportion of observed agreement (Ppos and Pneg) statistics were derived (see Table 30).   107 Table 30  Concordance of Reported Baseline and Balloon Inflation Symptoms Symptom % Agreement Kappa Ppos Pneg No discomfort 0 0 0 0 Localised Discomfort Chest 61 0.13 0.74 0.35 Left arm 13 0.16 0.22 0.89 Both arms 9 0.09 0.14 0.92 Throat 37 0.23 0.35 0.88 Left shoulder 0 -0.02 0.00 0.95 Jaw/teeth 47 0.23 0.29 0.92 Neck 50 0.45 0.48 0.97 Non-Localised Symptoms Diaphoresis 4 0.03 0.07 0.81 Fatigue 3 0.02 0.05 0.81 Shortness of breath 1 0.00 0.02 0.68 Note. Includes only participants with ischemia. Criterion for ischemia: ≥1 mm ST-segment deviation in any ECG lead.   The symptom with the highest agreement and proportion of observed positive agreement was chest discomfort, followed by neck and then throat discomfort. The proportion of positive agreement (Ppos) (i.e., a symptom that had been reported as being present at baseline was also reported during inflation) for other symptoms was low, with no symptom reaching greater than 35% positive agreement. The proportion of negative agreement (Pneg) (i.e., a symptom that was not reported as being present at baseline was also not reported during inflation) was greater than 80% for most symptoms, but only 35% for chest discomfort. Kappas were consistently low for all symptoms. With the presentation of these descriptive findings regarding the subjective experience of myocardial ischemia complete, analysis for the presence of latent classes is now presented.   108 Exploration for Latent Classes To answer the question, “Do certain symptoms occur in clusters or latent classes?”, latent class analysis was undertaken. The original intent was to focus this exploratory analysis on symptoms reported by the ischemic sub-sample during balloon inflation. However, this analysis demonstrated that, with respect to the balloon inflation responses, participants belonged to a single class. Therefore, for thoroughness, I also examined the symptoms reported at baseline. Further, given this decision, it seemed unnecessary to focus only on the sub-sample who experienced ischemia during balloon inflation, because verified ischemia only applied to the balloon inflation experience, not the baseline symptom experience that the participants recounted. So, the whole sample (N = 305) was included. All symptoms were subjected to LCA except those that had very low frequency (i.e., less than 0.5%, e.g., “itchiness”, “cheeks”, “cold feet”) and were deemed likely to be clinically trivial. Latent Classes in Baseline Symptom Reporting For the baseline symptom data, a first 2-class model had an acceptable Aikaike’s information criterion (AIC) (4424.433 for the 2-class model versus 4506.569 for the 1-class model) and sample-adjusted Bayesian information criterion (BIC) (4454.616 for the 2-class model versus 4521.387 for the 1-class model), but the quality of the classification was lower than desired, as measured by the entropy score of 0.66 (Celeux & Soromenho, 1996). Therefore, three additional 2-class models that included one, seven and eight fewer symptoms, respectively, were evaluated. In these models, however, AIC and sample-adjusted BIC values worsened and entropy values did not improve. Therefore, the low entropy value notwithstanding, the initial 2- class model was accepted as the best possible model for these data. Further supporting the validity of the 2-class model, the Vuong-Lo-Mendell-Rubin likelihood ratio test of the null hypothesis that there is only one class (versus two) was significant at 0.008. Approximately one   109 third (95 [31.1%]) of the participants were estimated to belong to Class 1, and the remaining two thirds (210 [68.9%]) to Class 2, based on the estimated posterior probabilities. Table 31 outlines the percentage of each class that reported each of 27 selected symptoms, and Figure 4 depicts this graphically.  Significant differences between the classes in symptom frequency were: chest, throat, both shoulders, jaw, intrascapular and back discomfort; and shortness of breath, diaphoresis, fatigue, dizziness, indigestion, headache and nausea, Class 1 having higher frequency of reporting all of these symptoms. Table 31  Percentage of Participants Reporting Selected Symptoms at Baseline, by Class Class, % p </= 0.05 Discomfort or Symptom 1 (n = 95) 2 (n = 210)  Localised Symptoms Chest 96.0 82.5 * Left arm 15.0 18.0 Right arm 4.8 2.1 Both arms 26.7 15.8 Left shoulder 9.5 8.0 Both shoulders 7.2 2.8 * Throat 20.6 10.0 * Neck 9.2 4.3 Jaw 19.1 9.7 * Back 16.2 7.8 * Intrascapular 6.8 1.1 * Abdomen 1.0 0.0 Left clavicle 0.0 0.5 Headache 4.0 0.5 * Note: Group with higher percentage is bolded, for each symptom   110 Table 31 (cont’d)  Percentage of Participants Reporting Selected Symptoms at Baseline, by Class Class, % p </= 0.05 Discomfort or Symptom 1 (n = 95) 2 (n = 210)  Non-Localised Symptoms Shortness of breath 72.6 39.4 * Diaphoresis 63.2 18.7 * Fatigue 70.1 17.3 * Nausea 29.9 9.9 * Vomiting 0.0 0.5 Indigestion 39.6 1.4 * Dizziness 63.1 9.0 * Palpitations 0.0 0.5 Diarrhea 2.9 0.5 Anxiety 0.0 1.0 Cough 1.0 0.0 Restlessness 0.0 0.5 Note: Group with higher percentage is bolded, for each symptom   111 Proportion of Participants Reporting Selected Symptoms, by Class 0.00 0.10 0.20 0.30 0.40 0.50 0.60 0.70 0.80 0.90 1.00 Ch es t Le ft A rm Rig ht Arm Bo th Arm s Le ft S ho uld er Bo th Sh ou lde rs SO B Dia ph or es is Fa tig ue Th ro at Ne ckJa w Ba ck Na us ea Ind ige stio n Vo m itin g Int ra sc ap ula r Ab do m ina l Rig ht Ch es t Diz zin es s Pa lpit ati on s Dia rrh ea An xie ty Le ft C lav icle Co ug h Re stle ns sn es s He ad ac he Symptom Pr o po rti o n Class 1 Class 2  Figure 4  Proportion of participants reporting selected symptoms, by class.  Power Analysis A post-hoc analysis of the power afforded by this sample for the latent class analysis was conducted using a Monte Carlo simulation, with 1000 replications. This revealed a power of greater than 80% for most of the symptoms. Only three symptoms fell below the 80% level of power for predicting membership in class one: diaphoresis (71.0%), indigestion (61.0%), and dizziness (62.0%); only shortness of breath fell below 80% power for class two (57.0%). Since the best model from the LCA resulted in two classes, binary logistic regression analysis was the most suitable approach to examine predictors of class membership. This was employed to explore the 12 socio-demographic and clinical variables that had reached significance of 0.25 or less in the bivariate analysis as covariates. The final model included   112 sex/gender (which was non-predictive of class membership) and also two predictors of class membership: (a) age (with each increasing decade of age, the likelihood of belonging to class 1 decreased [OR 0.77; 95% CI: 0.61-0.97]); and (b) grade of angina (participants with a more severe grade of angina, that is occurring even with exercise intensity and duration less than strenuous, rapid or prolonged had a higher likelihood of belonging to Class 1 [OR 2.24 CI: 1.25-4.02]). Table 32 displays the details of this analysis.  113 Table 32  Unadjusted and Adjusted Association of Class Membership with Selected Socio-Demographic and Clinical Variables (Class 2 as referent) All Class 1 n = 95 Class 2 n = 210 Variable M (SD) Unadjusted OR (95% CI) (Univariate) Adjusted† OR1 (95% CI) (Multivariate) Age (decades) 6.4 (1.1) 6.2 (1.2) 6.5 (1.0) 0.80 (0.64-1.01) 0.77 (0.61-0.97)  % Female 39.7 45.3 37.1 1.39 (0.86-2.29) 1.49 (0.90-2.48) Immigrant 29.5 23.2 32.4 0.60 (0.34-1.05) -- CCS2 angina grade =/> 3 69.5 80.0 64.8 2.18 (1.22-3.88) 2.24 (1.25-4.02) Indication for procedure: ACS3 43.0 48.4 40.5 1.38 (0.85-2.25) -- Pre-existing heart failure 4.6 7.4 3.4 2.28 (0.78-6.71) -- Renal dysfunction4 61.0 52.2 65.0 0.59 (0.36-0.97) -- Smoking Never (referent) Former Current       33.8      13.8      52.5      35.8     18.9     45.3       32.9      11.4      55.7  1.00 075 (0.44-1.28) 1.52 (0.73-3.18) -- Receiving ACE5 inhibitor 55.4 63.2 51.9 1.59 (0.97-2.61) -- Anti-depressants 4.3 6.3 3.3 1.96 (0.64-5.98) -- † Adjusted for potential confounders that had reached significance </= 0.25 in bivariate analysis. To preserve statistical power the model was trimmed of non-contributory variables until the best model was obtained; only variables in final model presented in adjusted odds ratios; 1OR, odds ratio for belonging to Class 2; 2CCS, Canadian Cardiovascular Society; 3 ACS, acute coronary syndrome; 5ACE, angiotensin-converting enzyme; 4renal dysfunction defined as eGFR (Cockroft-Gault) < 90 ml/min/1.73 m2   114 Summary In summary, findings from this prospective exploratory study of sex and gender differences in the reported symptoms of 305 participants undergoing elective, semi-urgent or urgent PCI have been presented. The sample was described with respect to socio-demographic characteristics, pre-procedure clinical status and procedural characteristics. Findings relating to the subjective experience of myocardial ischemia, as indicated by reported symptoms, were then presented, including perception of any symptoms, number and intensity of symptoms, and a thorough examination of the location and type of symptoms. Exploration of sex/gender differences in each of these components of symptom expression was presented, as well as multivariate analysis of the association of sex/gender and other variables with selected symptoms. Finally, findings from an exploratory latent class analysis were presented. A discussion of these findings and their implications for both practice and research follows.   115 CHAPTER 5: DISCUSSION  The main objective of this study was to further knowledge about whether there are sex/gender differences in the experience of myocardial ischemia during an ACS. In this final chapter, I summarise the findings of the study, the study design’s strengths and limitations, discuss several salient findings in relation to previously reported research, suggest what implications for practice the current findings may have, and conclude with recommendations for future research in this area. Summary of Findings  Although therapy for myocardial infarction (MI) has made great strides in the past 25 years, the best available treatments for ST-elevation MI (STEMI), potentially the most serious type, are exquisitely dependent on timely initiation. Great attention has been focused on developing systems of care to improve the consistency of rapid delivery of these treatments to patients with STEMI.  However, initiating this sequence of events depends almost entirely on patients’ recognition of their symptoms as being serious, possibly related to their heart, and requiring prompt medical attention. In spite of significant improvements in the health-care system segment of the requisite chain of events, patient delays in treatment seeking remain troublesome. Many studies (De Luca, Suryapranata, Ottervanger, & Antman, 2004; Finnegan et al., 2000; Gibler et al., 2002; Goldberg, Yarzebski, Lessard, & Gore, 2000; Goldberg et al., 2002; McGinn et al., 2005; Meischke et al., 1993; Murphy, Chen, Cannon, Antman, & Gibson, 2002; Sheifer et al., 2000; Steg et al., 2003), though not all (for example Zerwic et al., 2003), have pointed to longer delays amongst women. But even in a study that did not demonstrate a longer delay amongst women (Zerwic et al.), sex/gender differences in the factors that influence treatment-seeking decisions were found. Research attention has focused on whether women in fact make treatment- seeking decisions from a different starting point: symptoms that are different from men’s. By employing balloon inflation during percutaneous coronary intervention (PCI) as a model of   116 spontaneous ischemia, this study sought to determine: (a) whether there are sex/gender differences in the subjective experience of myocardial ischemia and (b) whether classes of people could be identified that might explain the pattern(s) or profiles of reported symptoms.  The rate of perceiving any symptoms was similar for women and men, at both baseline and balloon inflation. Women reported a significantly greater number of symptoms at baseline, but not during balloon inflation, and significantly greater intensity of symptoms, both at baseline and during balloon inflation. Although significantly fewer women than men reported having had chest discomfort during the baseline event, no differences were found between women’s and men’s reports of chest discomfort during balloon inflation, when only those with ECG-confirmed ischemia were compared. Other than sex/gender, statistically significant predictors of reporting having had chest discomfort at baseline were urgency of the procedure, BMI and first vessel treated. Predictors of reporting chest discomfort during balloon inflation were age, having peripheral vascular disease and receiving clopidogrel. As for the rates of reporting so-called typical symptoms, no sex/gender differences were found at either measurement time. However, having peripheral vascular disease, having had a previous PCI, and receiving lipid-lowering medication were predictive of reporting typical symptoms. There were sex/gender differences in the rates of reporting other symptoms, however. Women were particularly more likely to report having had throat, jaw, back and only non-chest discomfort at baseline, while during balloon inflation, women were again more likely to report having throat, jaw, neck and only non-chest discomfort. Other predictors of these symptoms were first vessel treated and age (jaw discomfort); age and receiving an ACE-inhibitor (throat discomfort); and smoking, receiving hormone replacement therapy, or having had a previous MI (neck discomfort).   117 The pattern of baseline symptom reports formed two distinct classes: by probability estimates, approximately one third of the participants belonged to one class, and the remaining two thirds belonged to a second class. Those belonging to the first class, termed the “many/non-localised” symptoms class, had higher rates of reporting predominantly non- localised symptoms, such as dizziness, diaphoresis and shortness of breath, and a higher probability of reporting more symptoms compared with the second class, the “few/localised” symptom class. Logistic regression modelling demonstrated that only age and Canadian Cardiovascular Society (CCS) grade of angina were predictors of class membership: participants who were younger and more incapacitated by angina were more likely to belong to the smaller “many, non-localised” symptom class. Sex/gender was not associated with class membership. Strengths and Limitations Strengths Objective Confirmation of Ischemia  This study has several strengths. Perhaps the most significant strength is that by carrying out data collection during a PCI procedure, ischemic events could be confirmed (or disconfirmed) using ST-segment measurement from ECG rhythm strips recorded during balloon inflation. This removed any doubt as to whether the symptoms being described related to an actual ischemic event, a feature which strengthens the inferences being made about the symptom data. No published study of sex/gender differences in symptoms of ACS could be found that incorporated this specific design strength. Prospective Design  A second strength is that this was a truly prospective study. One half of the symptom data were collected during the ischemic event for which those symptoms were being reported (balloon inflation). This resulted in an important advantage: the well-known pitfalls of poor recall on the part of the participants (Canto et al., 2007; Chen et al., 2005) were avoided, which   118 increases confidence in the quality of the data. And, although the baseline symptom data were drawn from the participants’ recall of their prior symptoms, this was offset somewhat by the fact that there was no reliance on secondary sources such as health records. Thus, the limitations of the health record as a data source for symptoms (DeVon et al., 2004) did not plague this study, which also increases confidence in the validity of the data. Sample Size  The reasonably large sample size (N = 305) was another strength of this study. Unlike several published studies, the relatively large sample provided adequate precision for point estimates and odds ratios, as outlined in Table 1 and other tables describing the multivariate analyses. Only four studies of sex/gender differences in symptoms of ACS with larger samples were found (Kosuge et al., 2006; Lovlien, Schei, & Hole et al., 2006; Milner et al., 2002; Thuresson et al., 2005), other than reports from registries. One other study that employed a cluster analysis of symptoms (Ryan et al., 2007) had a larger sample, but this was a secondary analysis using pooled data. Limitations Selection Bias  There were some limitations to this study that must be acknowledged. The aforementioned strength obtained by collecting data primarily during balloon inflation paradoxically introduced a selection bias. By including only those who had been referred for angiogram, important exclusions may have occurred. It has been well-documented (Anand et al., 2005; Sedlak et al., in press) that, although early angiography is recommended for patients with ACS and non-ST-elevation MI, lower rates of referral for angiography persist for women. One obvious reason for a patient not being referred for an angiogram is not having a clinical picture suggestive of coronary artery disease, and symptom profile is one feature on which clinicians rely heavily when evaluating the clinical picture. Therefore, drawing the sample from only those who   119 were referred for angiogram could have excluded people with atypical symptoms, thereby skewing the symptom profile of female participants towards one that is more “typical.” This would favour the finding of no sex/gender differences in either chest discomfort or “typical” symptoms that was obtained in this study. The extent to which this has affected the findings is unknown. Duration of Balloon Inflation  The research setting, again, perhaps paradoxically, can be seen as a limitation because of two factors: the duration of ischemia and the atmosphere of the cardiac catheterisation laboratory. To ensure participants’ safety by not exposing them to unduly long episodes of ischemia, the research protocol limited the duration of balloon inflation to a maximum of two minutes. In fact, several participants were not exposed to the full two minutes. Nevertheless, one could argue that even two minutes does not adequately represent the “real-life” experience of ACS, which usually lasts more than five minutes, and may last for hours (Anderson et al., 2007). Longer-lasting ischemia might have produced different and possibly a greater number of symptoms. Reasons for this include worsening ventricular function as ischemia continues (Braunwald, 2001), leading to shortness of breath, dizziness or palpitations, for example; or increased psychological distress as the discomfort does not abate or as the patient assigns meaning to the symptoms (Unruh, 1996), leading to anxiety, fear, nausea or sweating, for example. Short ischemic time during the study may therefore have accounted for the fewer number of symptoms reported during inflation than at baseline, when the duration of participants’ ischemia had presumably been longer (mean 4 symptoms versus 1 symptom, respectively). It could also explain fewer symptoms of worsening systolic function (e.g., shortness of breath, fatigue, vomiting or dizziness) or anxiety (e.g., shortness of breath or anxiety) reported during inflation, although the number of participants reporting some of these symptoms was small at both times. However, this factor (comparatively short ischemic time)   120 would presumably have affected men and women equally, so there is no reason to believe that it was an important influence on the differences found between men and women in the number or specific type of symptoms reported. Moreover, the only symptoms for which significant sex/gender differences were found at baseline but not at balloon inflation were chest and back discomfort, which are not usually specifically associated with systolic dysfunction or anxiety.  One could also expect that greater intensity of symptoms might have occurred if there had been longer-lasting ischemia, because of temporal summation of pain, which has been shown to be more pronounced in women than in men (Fillingim et al., 2009). This may partly explain the lower intensity scores reported by participants on the whole, during balloon inflation. It is also possible that the sex/gender difference in intensity would have been more pronounced had there been longer inflation times, since this would presumably have allowed for greater temporal summation. However, a difference that has statistical significance does not equate with clinical importance, and in this case it is not likely of much clinical importance.  A further limitation emerged in the conduct of this study: the duration of balloon inflation during the PCI was statistically significantly shorter for women than for men. This was unintentional, and speculation as to why this may have happened would be just that: speculation. However, the possible effects of this occurrence deserve exploration. Shorter inflation times effectively meant that women had less opportunity for ischemia to occur, though differences of seconds, as opposed to minutes or hours, are probably not clinically significant. Indeed, there were no statistically significant differences in the proportion of women and men that did, in fact, experience ECG-confirmed ischemia. But even though the prevalence of ischemia was not different, one might wonder if shorter inflation times for women would result in fewer symptoms, lower symptom intensity, or fewer symptoms related to systolic dysfunction or anxiety being reported by women, based on the arguments above. However, this did not occur.   121 Clinical Environment  The other aspect of this setting that could have been a limitation is the clinical atmosphere itself. Again, this does not accurately mimic the reality of the early stages of an ACS, when patients are usually in a non-healthcare setting. Participants may have felt safe and secure surrounded by many skilled cardiac health professionals, or some may have felt reluctant to report all sensations, for fear of embarrassment. On the other hand, the highly technical environment, coupled with a likeliness of patients to perceive the procedure as serious and at least somewhat frightening, may have caused more anxiety than would be present in other settings where ACS might occur, at least in the early stages. If participants did feel reassured by the skilled team around them, this would account for the smaller number of symptoms reported and for the almost complete absence of anxiety-related symptoms, though, as previously discussed, there were actually relatively few anxiety-related symptoms reported at baseline. It could also explain the decreased intensity of discomfort reported. Embarrassment would similarly account for fewer and less intense symptoms. Anxiety related to the procedure and the technology would most likely have resulted in a greater number and intensity of symptoms being reported, especially those related to sympathetic stimulation (e.g., palpitations, shortness of breath, diaphoresis). Since the opposite occurred (fewer symptoms of less intensity during balloon inflation), this particular limitation was probably not operative during this study. In addition, the first two effects (feeling of safety and feeling reluctant to report symptoms) may have compounded the effect discussed above - shorter duration of ischemia - to produce a smaller number of symptoms, of lesser intensity. In terms of gender differences, the limitations caused by the clinical atmosphere present in the cardiac catheterisation laboratory would probably have applied equally to men and women. This, coupled with the fact that a sex/gender difference in the number of symptoms had also been found at baseline, indicates these   122 limitations probably did not influence the finding of a sex/gender difference in the number of reported symptoms during balloon inflation. Concordance of Baseline and Inflation Symptoms Another limitation was the extent to which balloon inflation provided a valid model of spontaneous myocardial ischemia. As demonstrated by the analysis of concordance between the participants’ reported baseline symptoms and their symptoms reported during balloon inflation, there was high positive agreement for chest discomfort, and since this was the most frequently reported symptom by both men and women, this is both important and encouraging. Other than for neck discomfort, concordance was low for many of the other most frequently-reported symptoms. Low concordance could simply be a different index of the cumulative effect of the foregoing two limitations, that is, the short duration of ischemia and the artificial, non-‘real life’ nature of the environment. At least this could explain the low concordance scores for diaphoresis, fatigue and shortness of breath, since they, more than localised symptoms, may be associated with longer duration of ischemia, as discussed. However, it is difficult to explain the low concordance between the other localised symptoms on this basis. It may indicate that some of the localised symptoms reported to have occurred at baseline (e.g., left arm, both arms, left shoulder) had not actually been related to ischemia. Unfortunately, whether this is true cannot be known with certainty, since the presence of ischemia could not be verified for these reports. If one takes that view though, one can argue the research protocol accomplished what it was meant to accomplish: it separated symptoms attributable to ischemia from those not attributable to ischemia. However, since many of the symptoms in question are widely considered typical (Anderson et al., 2007; Antman et al., 2004; Bassand et al., 2007; Van de Werf et al., 2008), it is unlikely that reports of these baseline symptoms were completely unrelated to ischemia. Other explanations are elusive.   123 Measures of Gender, Depression and Anxiety  A measure of gender was not formally incorporated into the study protocol, and this is a limitation. Instead, the participants’ gender was inferred by the research assistants from a number of sources, including usual social cues (e.g., outward appearance, name, etc.). It is recognised that this method may have failed in some cases, because, since gender is not a dichotomous phenomenon, judging gender by outward appearance would not have had adequate sensitivity to the entire continuum of gender expression. As Knaak (2004) stressed, by using this “gender = male/female” approach, “… we invoke assumptions of biological foundationalism and ‘natural’ sex/gender congruency, seriously oversimplifying gender’s complexity and variability” (p. 313). She emphasized “…the need to resist framing gender as a single, biologically based, dichotomous term that is stable and unchanging” (p. 313). Although this method of collecting data about this crucial variable certainly has limitations, it is not entirely without foundation. After all, people express their gender identity through gender roles (which includes things such as body image, dress, and interactions with others) and may modify this enactment based on their acceptance or renouncement of those roles (Johnson et al., 2007). Thus, how a person’s gender is perceived by others is influenced to a significant extent by the conscious decisions they make about their own outward appearance. Nevertheless, as mentioned, this blunt tool may have missed nuances within this variable, which could have led to misattributing certain symptoms or symptom profiles to men or women. Accordingly, the finding of no difference in frequency of reporting chest discomfort, for example, might not have held true if a re-analysis demonstrated that only the women with more “masculine” attributes, as measured by a gender scale (e.g., the Bem Sex Role Inventory (Bem, 1981)), who would therefore have been classified as belonging to a more “masculine” gender group, had had this complaint. The extent to which this occurred is, unfortunately, not known.   124 Finally, neither anxiety nor depression was measured in this study. The potential benefit of measuring these psychological variables is that both have been shown to be related to pain (depression and pain are strongly co-morbid; anxiety can heighten pain) (Fillingim et al., 2009). Therefore, if there were an association of depression or anxiety with symptom differences it would not have been evident, and these differences may have been misattributed or over- attributed to sex/gender (or other measured covariates). This limitation is somewhat offset because data regarding medications taken in the previous 24 hours were collected, and antidepressants were included in that list of medication classes. Accordingly, this variable was examined for an association with symptom reports. Current Findings in Relation to Other Evidence No Sex/Gender Difference in Chest Discomfort or Other Typical Symptoms  Chest pain or discomfort is the symptom that is perhaps the most recognizable and relied upon by both the lay public (Fang, Keenan, Dai, & Denny, 2008) and health professionals (Canto et al., 2007; McCarthy, Beshansky, D'Agostino, & Selker, 1993; Pope et al., 2000), when determining if an MI is occurring. Therefore, findings related to sex/gender differences in the occurrence of this symptom arguably have the largest potential impact. Although data from the baseline symptom reports in the current study revealed a sex/gender difference, no difference was found between men and women in the reports of chest discomfort during a period of known ischemia. This is in contrast with several earlier studies. Canto et al.’s (2007) comprehensive review indicates that in the vast majority (80% of those including participants with ACS and 100% of those including participants with MI) of studies reporting on sex/gender differences, more women presented with no chest discomfort than did men. However, six of the ten studies conducted after the Canto review (and reviewed for this study) (DeVon, Ryan et al., 2008; Dey et al., 2009; Kosuge et al., 2006; Lovlien, Schei, & Gjengedal, 2006; Omran & Al- Hassan, 2006; Thuresson et al., 2005) did not find any sex/gender differences in this symptom.   125 Those finding no difference were equally split between studies including ACS participants, and those including MI participants. Many of these investigators adjusted their analyses for age and other co-morbidities, which could account for the differences in findings. Some earlier studies (Bayer, Chadha, Farag, & Pathy, 1986; Brieger et al., 2004; Canto et al., 2000; Goldberg et al., 1998; Milner, Vaccarino, Arnold, Funk, & Goldberg, 2004; Stern et al., 2004; Then, Rankin, & Fofonoff, 2001) did find that age was an important predictor of having chest or typical symptoms, though the effect of sex/gender was not evaluated in all cases, and age did not attenuate the effect of sex/gender in all of studies in which it was evaluated. Indeed, controlling for clinical and demographic variables added rigour and depth to the analysis in the current study also: age, peripheral vascular disease, and receiving clopidogrel emerged as significant predictors of reporting chest discomfort, but sex/gender still did not attain significance.  Another explanation for this finding is that the research design itself (prospectively collecting symptom data during a period of known, controlled ischemia) contributed in two ways: (a) eliminating recall bias and (b) eliminating from the analysis any reports of chest discomfort that were not due to ischemia. Many authors, especially those who have published systematic literature reviews (Canto et al., 2007; Chen et al., 2005; DeVon & Zerwic, 2002), have cited the retrospective design of most previous studies (i.e., participants recalled an event in the past) as a serious limitation, since there could be no confirmation of ischemia. Although this study has overcome this limitation, to accept this design factor as being the main cause of these findings differing from past findings, one would have to believe that either the majority of the cases of false positive chest pain (i.e., chest pain in the absence of ischemia) occur in men, or that only women fail to recall chest discomfort in their reports. Both of these seem highly unlikely, since there is no evidence to support either proposition, and there is no logical reason to suppose either is true. Nonetheless, the sex/gender difference that was present at baseline was attenuated in the sub-sample for whom ischemia could be confirmed.   126  Another possible explanation of the finding that there is no difference in chest or typical symptoms between men and women is that, as suggested by Shin et al. (2009), methods of data collection (e.g., checklists, open-ended questions, combination approaches) affect what symptoms are reported. Similarly, this explanation would only fit these findings if open-ended questioning (as employed in this study) prompted more reports of chest discomfort among women than would other methods. On the contrary, Shin et al. found that open-ended questioning did not produce gender differences in chest pain reports, whereas adding checklists of symptoms did. However, Shin et al. did not compare open-ended questioning alone or in combination, to the use of checklists alone, which have been used extensively by investigators. Closed checklists arguably favour reporting of more classical symptoms and might discourage reporting of symptoms that are not perceived (by either the patient or the health provider) to be typical. To test this notion in a cursory way, the findings of 14 recent studies were cross- tabulated to explore whether there is a gender difference in reported chest discomfort (yes versus no) and type of data collection method employed (closed versus open). Six of six studies (DeVon & Zerwic, 2003; DeVon, Ryan et al., 2008; Kosuge et al., 2006; Milner et al., 1999; Omran & Al-Hassan, 2006; Thuresson et al., 2005) that used open methods found no gender differences, whereas only two out of eight studies (Dey et al., 2009; Lovlien, Schei, & Gjengedal, 2006) using closed methods found no differences (Fisher’s Exact test = <0.01). In the current study, having ECG confirmation of ischemia during symptom monitoring, and employing a data collection technique that perhaps elicited a more complete expression of symptoms, perhaps led to a symptom profile that is closer to the “true” profile of those experiencing ACS. Notwithstanding the limitations in this model of spontaneous ischemia, as discussed earlier, the fact remains that many participants did report chest discomfort, and the concordance between baseline and balloon inflation reports for this symptom was high, which adds confidence in the chest discomfort findings.   127  Finally, the size of the whole sample (N = 305) provided an acceptable margin of error for the measured point prevalence rates (95% confidence limits for prevalence rates of chest discomfort at baseline, which fell between 71% and 80%, ranged from +/- 4.5% to 5.2%). There was a very slight reduction in precision with the smaller sample (n = 245) used in the analysis that was limited to participants with ischemia (95% confidence limits for the prevalence rates of chest discomfort during balloon inflation, which fell between 51% and 60%, ranged from +/- 6.1% to 6.3%). Due to the prevalence rates being so close to 50%, the power of this ischemic sub-sample size to detect a difference in rates of chest discomfort during balloon inflation was disappointingly low at 14.2%, so the possibility of a Type II error, when comparing the subgroups’ rates, must be acknowledged.  In terms of typical symptoms, even defining the term was problematic, as discussed in the literature review, which makes comparison with previous findings difficult and, likely meaningless. However, the operational definition used in this study incorporated items for which there is little disagreement across various clinical practice guidelines, and no differences in symptoms were found between men and women. Few researchers have specifically analysed data for differences in the frequency of reporting the composite “typical symptoms”. Rather, many have stipulated the presence of typical symptoms as an inclusion criterion or simply made reference to the term in the discussion, without defining it. However, Milner et al. (2002) found that typical symptoms were a common component of women’s clinical presentation, but they did not test for sex/gender differences in the frequency of reporting typical symptoms. In another study (Milner et al., 1999), it was found that, although women were more likely than men to have some atypical symptoms, they were as likely or more likely to have many symptoms that would be considered typical. Applying a broad definition of typical, it seems that women and men do not differ in the frequency of reporting this constellation of symptoms.   128  In this study, the relationship of chest pain and other typical ACS symptoms with a diagnosis of diabetes was not significant. Diabetes was only statistically significant as a predictor of having only non-chest symptoms, and only at baseline. Findings related to diabetes and ACS symptoms have been inconsistent in past studies, and not all of the reviewed studies have examined sex/gender and diabetes as covariates within the same model, making comparison with the current findings difficult. Some studies stratified participants by duration of diabetes and by age (but not necessarily by sex/gender), but, with only 60 participants with diabetes in this sample, such sub-analyses would likely have lacked statistical power. Thus, the current findings cannot be seen as adding important new knowledge about a potential relationship between diabetes and ACS symptoms. Sex/Gender Differences in the Number and Intensity of Symptoms Women Report More Symptoms  In many studies of sex/gender differences in symptoms of MI or ACS, researchers have not reported the number or intensity of symptoms experienced. The finding in this study of a greater number of symptoms reported by women is consistent with roughly half of the studies that have reported it. In one (Canto et al., 2007) of three published reviews (Canto et al., 2007; Chen et al., 2005; DeVon & Zerwic, 2002) there is mention of the number of symptoms experienced, and this appears to be based on one study alone (Milner et al., 1999) (women were reported to experience more). Furthermore, of 10 studies with specific sex/gender analyses (see Table 2) that were not included in the published reviews of the literature, and 8 studies that were included in at least one of the reviews, but I have also highlighted here because they are so frequently cited (Brieger et al., 2004; Culic, Eterovic, Miric, & Silic, 2002; DeVon & Zerwic, 2003; Goldberg et al., 1998; Kudenchuk, Maynard, Martin, Wirkis, & Weaver, 1996; Milner et al., 1999; Milner et al., 2002; Milner et al., 2004), only 6 measured the number of reported symptoms and tested for sex/gender differences. Three of these (DeVon & Zerwic, 2003; Granot et al.,   129 2004; Milner et al., 1999) found that women report statistically significantly more symptoms than do men; three did not. Since these few studies result in an equivocal conclusion, it would not be legitimate to describe this finding as either supporting or refuting previous findings.  There are a few possible reasons for the current finding that there was a sex/gender difference in the number of symptoms reported. First, it may be related to the method of data collection, as discussed previously. However, only one of the three studies that did find a difference used open-ended questioning. So, the notion that the method of eliciting symptoms accounts for different findings between the studies, and that open-ended questioning is a more sensitive tool for capturing women’s ways of relaying symptom information is not borne out in this particular instance. Nevertheless, it is important to acknowledge that checklists, by their very nature, limit the number of symptoms the participant can choose from (although men are similarly constrained with this method), but also may completely exclude some symptoms that women (and men) experience, thus potentially resulting in an incomplete clinical picture. One can only speculate whether this finding might have been replicated in more studies, had open- ended questioning been used.  Second, it has been argued that psychosocial factors (e.g., socially-prescribed gender roles [including the gender of the person to whom the symptom is reported, though this has only been demonstrated in men, in experimental studies], different coping mechanisms, higher levels of anxiety, and higher prevalence of depression) influence women’s tendency to report higher pain intensity (Fillingim et al., 2009). These factors could arguably influence the number of symptoms reported in a similar way, and could have been operational in this study. Unfortunately, no measures of either anxiety or depression (except whether participants were taking anti-depressant medication) were included. It is interesting to note though that the research assistants were almost all women, which may have affected the responses of the men.   130  The above speculation addresses the possible differences in findings about the number of reported symptoms. However, it should be noted that it is not known whether men actually have fewer symptoms or merely report fewer. Indeed, differentiating the presence of a symptom from the willingness to report it is not an easy task (Ellermeier, 1997). An elegant laboratory experiment has been reported that differentiates the discrimination of pain (i.e., sensitivity) from the response to it (i.e., reporting it), and what sex/gender differences exist in this phenomenon (Ellermeier), but it is not known whether such a model is applicable to other, non-pain symptoms. Women Report Higher Intensity of Discomfort  Similarly, many of the previously reviewed studies did not report the intensity of symptoms experienced. Of the same 10 studies with specific sex/gender analyses that were not included in the published reviews of the literature (see Table 2), and 8 frequently-cited others (Brieger et al., 2004; Culic et al., 2002; DeVon & Zerwic, 2003; Goldberg et al., 1998; Kudenchuk et al., 1996; Milner et al., 1999; Milner et al., 2002; Milner et al., 2004), only 6 reported intensity differences. Two thirds of those found that women reported significantly higher intensity than men did (DeVon & Zerwic, 2003; DeVon, Ryan et al., 2008; Granot et al., 2004; Thuresson et al., 2005). Not only is the current finding supported by previous ACS- symptom research, it is also consistent with what is reported in the literature concerning several types of pain (Fillingim et al., 2009).  One of the possible explanations for this finding is shared with the explanations for the higher number of symptoms, discussed above. Specifically, the psychosocial factors thought to influence the tendency for higher reported pain intensity (Fillingim et al., 2009) have gained widespread acceptance in the pain literature. Since I have invoked concepts from the pain literature, two points should be noted. First, the intensity scores relate to the first reported symptom only, and it is possible that this was not considered to be the most severe by the   131 participant. One study protocol directed that participants be specifically instructed to rate the severity of the worst symptom they had experienced. This was necessary in that particular study though, because symptom data were collected with a checklist-based measure, which would not have otherwise captured the participants’ perceived severity ranking. Common sense would suggest that the possibility of the first-reported symptom in this study not being the symptom that was perceived as the most severe is likely not very great. Second, not all first-reported symptoms were pain-related, though the vast majority were (97.1% at baseline; 74.9% during balloon inflation). So, although the psychosocial factors listed above have been studied only in relation to pain, it seems plausible that such factors would operate similarly with respect to other, non-pain (or discomfort) symptoms, such as shortness of breath or fatigue, resulting in higher intensity of these symptoms.  Physiologic reasons for increased pain intensity in women also have been postulated. Most of these relate to the multiple, and, as yet, poorly-understood effects of sex hormones on pain (e.g., the hormone-dependent prevalence of certain types of pain; effects on the inflammatory response and thereby nociception; and effects on endogenous opioids) (Fillingim et al., 2009; Unruh, 1996). Nonetheless, since more than 80% of the women in the sample were over the age of 55 years, and only 6% were on hormone replacement therapy, it is unlikely that these particular suggested mechanisms exerted a significant effect in this study. Sex/Gender Differences in Other Symptoms  The statistically significant sex/gender differences in symptom reports for this study were in the reports of throat, jaw, and neck discomfort. This finding is consistent with many previous studies with respect to jaw (Dey et al., 2009; Granot et al., 2004; Kosuge et al., 2006; Lovlien, Schei, & Gjengedal, 2006; Lovlien, Schei, & Hole et al., 2006; Thuresson et al., 2005), neck (Kosuge et al., 2006; Thuresson et al., 2005), and throat discomfort (Kosuge et al., 2006; Lovlien, Schei, & Gjengedal, 2006). As well, all three of the published literature reviews (Canto   132 et al., 2007; Chen et al., 2005; DeVon & Zerwic, 2002) noted significant sex/gender differences in jaw and neck discomfort in several of the studies included in their reviews.  It is noteworthy that some co-variates were also significant predictors of reporting throat discomfort (age), jaw discomfort (having had a previous PCI), and neck discomfort (having had a previous MI). However, although these were significant in the model, the effect of sex/gender on throat and neck discomfort was actually amplified by controlling for these covariates, indicating that an imbalance in the prevalence of these variables existed between the men and women in this sample, such that the sex/gender effect was diluted in the bivariate analysis.  A review of the pain literature indicates that several investigators have shown that women have a higher prevalence of temporomandibular joint pain and tooth, jaw and other types of orofacial conditions (Fillingim et al., 2009). One proposed mechanism is the influence of sex hormones, since these differences are not apparent until after girls reach puberty. However, since these conditions are primarily musculoskeletal in origin, this may not necessarily explain ischemic pain being referred to this area. Neuro-anatomical studies and interventions, though, have revealed that the neck and jaw region is innervated primarily by the vagus nerve (Foreman, 2007; Mork et al., 2004), and there is also evidence to suggest that women have higher cardiac vagal activity than do men (Chambers & Allen, 2007). So, not only do the findings in this study of increased jaw, throat, and neck pain support past findings, but furthermore, plausible anatomical and physiological explanations exist. Two Classes of Symptom Expression  As reported at the beginning of this chapter, exploration for an unseen (latent) pattern in the data revealed two fairly distinct classes. Class 1 (which comprised only about one third of the sample) was characterised by a higher probability of reporting several symptoms, and also a tendency to report more non-localised symptoms. Predictors of belonging to this class were younger age and more severe functional disability from ischemic symptoms (that is, CCS class).   133 These findings are in contrast with the two other studies that have been published that similarly explored ACS symptom data for latent classes. One of these studies (Rosenfeld, Perrin, & Darney, 2005) was small and included women only, so comparison is not informative. The other study (Ryan et al., 2007) was a secondary analysis of data from nine studies, and had a sample of 1,073 participants. Five distinct classes (or clusters) were found in this study (which was also the case in the other smaller study, interestingly), unlike the current study. One of the classes identified by Ryan resembled the “Many/Non-Localising” symptom class identified in this study, with members having a high probability of having chest discomfort; arm, shoulder or hand discomfort; nausea and vomiting; shortness of breath; sweating; dizziness; weakness and fatigue. Likewise, there was one class that had a similar pattern to this study’s “Few/ Localising” symptom class, in that its members had a high probability of having only chest or shoulder, arm and hand discomfort. However, since these investigators identified five classes in all, it is probably not valid to compare only two of them to the classes identified in this study. Although age was reported as a significant predictor of class membership in Ryan’s study, the direction of its effect on membership in specific classes was not explicated. Unlike the current study, Ryan found sex/gender to be predictive of class membership, but again, the specific direction of this effect on class membership was not described. So, although the findings of the two studies are not comparable, given the different number of classes identified, it is possible that more classes would have been identified in the current study had there been a larger and perhaps more heterogeneous sample.  Predictors of membership in the “Many/Non-Localising” symptom class are interesting to ponder. There is abundant evidence that one effect of age is decreased nociception (Gibson & Farrell, 2004), so this could account for the tendency of older persons to report relatively fewer symptoms (even though they are likely to have a higher co-morbid burden). One could also speculate that advancing age, again, in spite of the fact that it is likely to introduce more co-   134 morbidity, may also introduce a certain philosophical acceptance of one’s failing health and mortality. Similarly, this generational cohort may subscribe to stronger notions of stoicism (Yong, 2006), or be more reserved with health professionals. Both of these factors could affect the willingness to report multiple symptoms.  The effect of the other predictor of symptom profile class, CCS class of angina, seems logical: having more severe functional limitation from symptoms of ischemia is probably distressing and frustrating to patients, at least, and frightening at worst. It is not difficult to see that this could prompt patients to report symptoms and not “hold back”. It was argued in a previous section that gender roles influence the number of symptoms one reports; I propose that perhaps the impact of severity of ischemic symptoms increasing beyond a certain point may be enough to overcome the influence of socially-prescribed roles, so that this gender difference is no longer present. Implications for Practice  The findings from this study have implications for professional practice, in the areas of patient and public education, clinical assessment, and professional education. Each of these domains is discussed in turn. Patient and Public Education  The trigger for initiating the chain of events that may culminate in effective, timely treatment of ST-elevation MI is the patient recognising his or her symptoms as serious and warranting professional attention. Therefore, it is crucially important that all members of the public understand what symptoms a person will likely experience when they are having ACS, which in turn requires evaluation to rule out (or in) MI.  Some investigators began exploring differences between men’s and women’s clinical presentation of ACS forty years ago (Reimann & Jahrmarker, 1969). As awareness increased that the model for typical symptoms of chronic and acute myocardial ischemia (Rose, 1962) is based   135 on a few dozen men, and evidence of the benefits of early reperfusion mounted, researchers paid increasing attention to the issue of sex/gender differences in the presentation of ACS. In the mid-1990s, researchers produced provocative findings, suggesting that women may have a different symptom profile from that which is considered typical (Goldberg et al., 1998; Kudenchuk et al., 1996; Maynard & Weaver, 1992; Milner et al., 1999). Even though early studies may have had methodological flaws, and some subsequent studies have not found differences, the popular press nonetheless espoused this issue, and the lay public listened. Although it is important to recognise that there may be some differences between men and women in this regard, it appears that the pendulum may have swung too far, as evidenced by this statement on the Oprah Winfrey website (2009): “Many of the symptoms of heart disease are often ignored, unrecognised or misdiagnosed, because women’s symptoms are completely [emphasis added] different than men’s”.  The potential impact of such a message and similar messages found in the media is important. The worst effect would be that people would surmise that if they (or a woman they know) are having “men’s” symptoms such as chest pain or pressure, they must not be having a heart attack. This could obviously lead to very detrimental delays in treatment-seeking. The corollary could also occur: people may misattribute a host of non-cardiac symptoms in women as cardiac, resulting in unnecessary worry and usage of emergency services. These misconceptions need to be addressed with clear and broadly delivered public health messages.  The findings from this study and others support the message that women are just as likely as men to have chest discomfort, and they should respond to it in exactly the same way as men should. Current messages about heart attack warning signs found on the Heart and Stroke Foundation of Canada (2008) website support this, as does their “Heart Truth” campaign (2008) website, a large-scale effort to deliver messages to the public, especially women, about women’s risks for and experiences of heart disease (including MI). However, the over-reaction of the   136 popular press and other media needs to be countered with increased efforts in all types of popular media, to dispel these inaccuracies. Other significant findings from this study, which are supported by previous research, namely that women report more symptoms, and are considerably more likely to experience throat, jaw, and neck pain, should also be incorporated into public messages, since having many symptoms and symptoms that are not classically associated with heart conditions may confuse women and deter them from seeking help. These two points were not evident on either the Heart and Stroke Foundation of Canada website or the Heart Truth website. Finally, it is important that such messages be targeted not only at women, but also at men, since it is often family, friends, or co-workers who are present and influence a person’s decision to seek care. I am currently beginning discussions with the Heart and Stroke Foundation of Canada about these findings and their possible uses. Clinical Assessment  Once ACS patients enter the healthcare system, it is imperative that they be assessed quickly and effectively. This study points to women’s symptoms being more similar to men’s than different, so there is no indication to recommend searching for a completely different symptom profile. However, women may not report chest discomfort first, and as discussed in a previous section, their willingness to report symptoms may be more responsive to open-ended, as opposed to rigid, pre-determined questioning. And, by the same token, since women seem to experience more symptoms (inferred, but not proven, by the observation that they report more), it may take more probing for the clinician to obtain a complete symptom profile. This would require patience and time, but may avert missed diagnoses. Although the rate of missed diagnosis in emergency departments for all forms of ACS is low (about 2%), the consequences of inappropriate discharge are serious, and may lead to as much as a 25% increase in mortality (McCarthy et al., 1993; Pope et al., 2000). Women under 55 years of age, non-white patients, and   137 those who reported shortness of breath as their primary symptom have been found to be significantly more likely have the ACS diagnosis missed (Pope et al.). Above all, the first responder, emergency triage nurse or physician should not conclude that the patient who says “No” when questioned about chest discomfort, or who does not mention it spontaneously as their first or even second symptom, is not experiencing the ACS. In the fast-paced environment of today’s emergency departments and emergency medical services, this may be a difficult pattern of practice to break. A concrete strategy for promoting this assessment approach would be to develop and implement an assessment tool for patients with suspected ACS that guides the clinician through a series of open-ended questions. Professional Education  Separation of recommendations for actual assessment practices from the education required to ensure they occur is somewhat artificial, but necessary for this discussion. The research of the last 15 years on the issue of sex/gender differences in ACS symptoms has been confusing for health professionals as well as the lay public. Only the most scrupulous review of the literature reveals the ambiguities and flaws in the research to date. Therefore it is probable that the average physician or nurse in active clinical practice has either been misled, depending on what research findings they have been exposed to, or they have been perplexed by the contradictory findings, and left not knowing which parts of this knowledge should be incorporated into their practice. Current clinical practice guidelines from some of the prominent professional associations (the American Heart Association, the American College of Cardiology, the Canadian Cardiovascular Society, the European Society of Cardiology) incorporate some content about sex/gender differences in ACS symptoms (Anderson et al., 2007; Antman et al., 2004; Bassand et al., 2007; Van de Werf et al., 2008), but they are somewhat inconsistent and limited. In a promising initiative, the National Heart, Lung and Blood Institute convened the women’s ischemic syndromes evaluation (WISE) panel in 2002, to evaluate available evidence   138 and make recommendations for practice and research. Several publications came of this, and the one addressing clinical practice (Hayes, Long, Hand, Finnegan, & Selker, 2004) makes several recommendations about “…communicating messages about strategies proven to aid identification and treatment of women with ACS…” (p. e61). However, it falls short of specifying just what those strategies are. Another publication from this initiative summarised important messages about ACS in women, including a recommendation to educate healthcare professionals about avoiding (presumably male) stereotypes, among other things. However, it is certainly not clear that this has happened in a systematic way in Canada, nor is it clear that “proven” strategies for identification of women with ACS exist.  Gaps in our knowledge do exist, and are addressed below. It is prudent to recommend that some educational programs are instituted now, however. The focus must be on clarifying the evidence and giving concrete direction for practice. The current findings would support the delivery of three main messages: (a) women experience chest discomfort with the same frequency as men and a small proportion of men and women do not report chest discomfort in the presence of ischemia; (b) women experience throat, neck and jaw discomfort more commonly than do men; (3) women may not report chest discomfort as their first or even second symptom (or at all, but this is not different from men), and it requires skilled and patient interviewing to elicit all symptoms.  These three simple messages could be disseminated in many ways: professional websites; creation of a clinical practice guideline on this particular subject; workshops at national and local conferences; campaigns of mass emailing to professional association members; advertisements in journals, and so forth. Of particular importance is that these messages reach all three pivotal professional groups: nurses, physicians, and emergency paramedics. It would be highly desirable if the message could be standardised across these groups (undertaking it as a joint venture), and   139 even more effective if some face-to-face education sessions could be inter-professional, to promote inter-professional exchange on this subject among clinicians. Implications for Research  Although this study has added to the body of knowledge regarding sex/gender and ACS symptoms, it also has generated some questions for future research. First, the framework for this study stipulated that one reason women may delay seeking treatment for ACS is that they experience a different constellation of symptoms from men, which they do not recognise as cardiac in origin. In light of the current findings, it should be stated that we still lack a clear understanding of the particular reasons for women’s treatment-seeking delays, which have been shown in some studies to be longer than men’s.  Research to elucidate these reasons is still needed: how do women’s relationships with family, physicians, or the healthcare system in general contribute to delays? Do women have a persistent knowledge gap in relation to their susceptibility to heart disease and MI? If findings from future studies reveal an explicit target for interventions to mitigate treatment-seeking delays, then such interventions should be developed and tested.  In view of the relatively small sample and the possibility that there was a Type II error, a larger study, using the same protocol with some modifications, could be contemplated. One issue that would need to be rectified would be more stringent staff training to ensure consistent application of the research protocol (questioning of participants, duration of balloon inflation). It is possible that with adequate funding, a larger sample could be recruited in a shorter period of time, which would contribute to less deviation from the protocol. However, some of the acknowledged limitations of the study relate to the study design itself. There is good reason to doubt that balloon inflation is a valid model of spontaneous myocardial ischemia.  The very strong association of sex/gender with jaw, throat, and neck discomfort is provocative. Although some anatomic and physiologic mechanisms have been suggested, one   140 cannot say they are understood clearly. Therefore, a neuro-anatomic exploration of the distribution and function of the vagus nerve in particular, and other central nervous system structures that innervate both the myocardium and the throat, jaw, or neck, would be informative, and might improve our understanding of different manifestations of myocardial ischemia.  As mentioned several times, the method of assessing ACS symptoms may be an important variable in explaining previously-reported gender differences. It would be interesting to first test this hypothesis by conducting a randomised trial of two or more assessment methods (e.g., open-ended versus closed questioning), and second, if men’s or women’s symptom reporting was found to be sensitive to a particular method of assessment, a clinical trial could be conducted to test a tool or protocol for assessment.  Another area for future exploration is testing of public messages about symptoms of heart attacks. Although the content of existing messages seems accurate (Heart and Stroke Foundation of Canada, 2008; Heart and Stroke Foundation of Canada: The Heart Truth, 2008), the belief that women’s symptoms are different seems to persist. So perhaps the messages are not reaching the intended audience. Messages that employ different media, different words or phrasing, and adaptations for different age, ethnic and cultural groups should be designed and tested.  Finally, it may be worthwhile to pursue further latent class analyses with larger samples, since this analysis was subject to the possibility of a Type II error. This would complement the rationale for the larger study contemplated above, because there may be factors, or combinations of factors that account for the heterogeneity of profiles in the sample that we cannot know a priori, and will only uncover through latent class analysis. If sex/gender differences were still not evident with a larger sample, that would constitute even more reason to explore subclasses within the population of ACS patients.   141 Conclusions  Treatment for ACS has improved steadily over the past 25 years. The time-sensitive nature of available treatments has focused attention on improving the speed and accuracy of diagnosis, which in turn has shone light on differences between men and women in terms of timeliness of treatment-seeking and their clinical presentation. For a variety of reasons, including under-representation of female patients in the development of clinical tools to assess typical symptoms and in the clinical trials that have contributed to our understanding of ACS, considerable research attention has been paid to ascertaining if there are sex/gender differences in ACS symptoms. However, these studies have had limitations, so the current study aimed to overcome some of those shortfalls by examining the issue in a more rigorous fashion.  In this sample of patients undergoing elective or urgent PCI, no significant differences were found in the frequency of reported chest and other typical symptoms between men and women. However, the findings revealed that women reported non-chest symptoms significantly more often than did men, and reported a significantly higher number of symptoms and intensity of discomfort, compared with men. Two distinct patterns of symptom reporting were identified in the sample: one that was characterised by a higher number and more non-localised symptoms, and the other characterised by relatively fewer and predominantly localised symptoms. These findings add to the evidence that women do, in fact, report the typical symptoms of ACS with similar frequency to men. Efforts to highlight differences between men’s and women’s symptoms of ACS to the lay public and professionals alike, although well-intentioned, have perhaps overshot their target, to the point where some women seem to believe that their symptoms of heart attack will be nothing like those of men’s. This misconception must be rectified. Cardiac health professionals must receive this information too, along with sensitisation to the fact that women are likely to report symptoms that have been considered “atypical” more often, and to report a greater number of symptoms. Due to the complex interrelationship   142 between sex-based and gender-based factors that could explain these differences, future research must take a comprehensive view and an interdisciplinary approach to disentangle and uncover these reasons, and to design interventions that are aimed at improving the efficiency and accuracy of diagnosis for women experiencing ACS.   143 REFERENCES  Acharya, D. U., Shekhar, Y. C., Aggarwal, A., & Annand, I. S. (1991). Lack of pain during myocardial infarction in diabetes: Is autonomic dysfunction responsible? American Journal of Cardiology, 68, 793-796. Airaksinen, K. E. J., Ikaheimo, M. J., Linnaluoto, M., Tahvanainen, K. U. O., & Huikuri, H. V. (1998). Gender differences in autonomic and hemodynamic reactions to abrupt coronary occlusion. Journal of the American College of Cardiology, 31, 301-306. Aldrich, H. R., Hindman, N. B., Hinohara, T., Jones, M. G., Boswick, J., Lee, K. L. et al. (1987). Identification of the optimal electrocardiographic leads for detecting acute epicardial injury in acute myocardial infarction. American Journal of Cardiology, 59(1), 20-23. Anand, S. S., Xie, C. C., Mehta, S., Franzosi, M. G., Joyner, C., Chrolavicius, S. et al., and CURE Investigators. (2005). Differences in the management and prognosis of women and men who suffer from acute coronary syndromes. Journal of the American College of Cardiology, 46, 1845-1851. Anderson, J. L., Adams, C. D., Antman, E. M., Bridges, C. R., Califf, R. M., Casey, D. E. et al. (2007). ACC/AHA 2007 guidelines for the management of patients with unstable angina/non–ST-elevation myocardial infarction: A report of the American College of Cardiology/American Heart Association Task Force on Practice Guidelines [Electronic version] Circulation, 116. Antman, E. M., Anbe, D. T., Armstrong, P. W., Bates, E. R., Green, L. A., & Hand, M. (2004). ACC/AHA guidelines for the management of patients with ST- elevation myocardial infarction [Electronic version]. Circulation, 110, e82-e293.   144 Arslanian-Engoren, C., Patel, A., Fang, J., Armstrong, D., Kline-Rogers, E., Duvernoy, C. S. et al. (2006). Symptoms of men and women presenting with acute coronary syndromes. American Journal of Cardiology, 98(9), 1177-1181. Barker, D. J. (2000). In utero programming of cardiovascular disease. Theriogenology, 53, 555-574. Bassand, J. P., Hamm, C. W., Ardissino, D., Boersma, E., Budaj, A., Fernandez-Aviles, F., Fox, K. A. A., Hasdai, D., Ohman, E. M., Wallentin, L., & Wijins, W. (2007). Guidelines for the diagnosis and treatment of non-ST-segment acute coronary syndromes [Electronic version]. European Heart Journal, 28, 1598-1660. Bayer, A. J., Chadha, J. S., Farag, R. R., & Pathy, M. S. (1986). Changing presentation of myocardial infarction with increasing old age. Journal of the American Geriatric Socitey, 34, 263- 266. Becklake, M. R., & Kauffmann, F. (1999). Gender differences in airway behaviour over the human life span. Thorax, 54, 1119-1138. Bem, S. L. (1981). Bem sex role inventory: Professional manual. Palo Alto, CA: Consulting Psychologists Press. Berry, C., Zalewski, A., Kovach, R., Savage, M., & Goldberg, S. (1989). Surface electrocardiogram in the detection of transmural myocardial ischemia during coronary artery occlusion. American Journal of Cardiology, 63, 21-26. Bessant, J. (2006). The fixed age rule: Young people, consent and research ethics. Youth Studies Australia, 25, 50-57. Bhopal, R., & Donaldson, L. (1998). White, European, Western, Caucasian, or What? Inappropriate labeling in research on race, ethnicity, and health. American Journal of Public Health, 88, 1303-1307.   145 Braunwald, E. (2001).  Coronary blood flow and myocardial ischemia. In E. Braunwald, D. P. Zipes & P. Libby. (Eds.), Heart disease: A textbook of cardiovascular medicine (6th ed.) (pp. 1087- 1113). Philadelphia: Saunders. Brieger, D., Eagle, K. A., Goodman, S. G., Steg, P. G., Budaj, A., White, K. et al. (2004). Acute coronary syndromes without chest pain, an underdiagnosed and undertreated high-risk group: Insights from the global registry of acute coronary events. Chest, 126, 461-469. Buchanan, F. F., Myles, P. S., & Cicuttini, F. (2009). Patient sex and its influence on general anesthesia. Anesthesia and Intensive Care, 37, 207-218. Buchthal, S. D., Den Hollander, J. A., Merz, C. N., Rogers, W. J., Pepine, C. J., Reichek, N. et al. (2000). Abnormal myocardial phosphorus-31 nuclear magnetic resonance spectroscopy in women with chest pain but normal coronary angiograms. New England Journal of Medicine, 342, 829-835. Califf, R. M., & Mark, D. B. (2005). Plagiarism notice. American Heart Journal, 149, 743. Campbell, C. M., Edwards, R. R., & Fillingim, R. B. (2005). Ethnic differences in responses to multiple experimental pain stimuli. Pain, 113, 20-26. Campeau, L. (1976). Grading of angina pectoris [Letter to the editor]. Circulation, 54, 522-523. Canadian Institutes of Health Research, Natural Sciences and Engineering Research Council of Canada & Social Sciences and Humanities Research Council of Canada. (1998 (with 2000, 2002 and 2005 amendments)). Tri-Council policy statement: Ethical conduct for research involving humans. Ottawa, ON: Interagency Panel on Research Ethics. Canto, J. G., Goldberg, R. J., Hand, M. M., Bonow, R. O., Sopko, G., Pepine, C. J. et al. (2007). Symptom presentation of women with acute coronary syndromes: Myth vs reality. Archives of Internal Medicine, 167, 2405-2413.   146 Canto, J. G., Shlipak, M. G., Rogers, W. J., Malmgren, J. A., Frederick, P. D., Lambrew, C. T. et al. (2000). Prevalence, clinical characteristics, and mortality among patients with myocardial infarction presenting without chest pain. JAMA: The Journal of the American Medical Association, 283, 3223-3229. Celeux, G., & Soromenho, G. (1996). An entropy criterion for assessing the number of clusters in a mixture model. Journal of Classification, 13, 195-212. Chambers, A. S., & Allen, J. J. B. (2007). Sex differences in cardiac vagal control in a depressed sample: Implications for differential cardiovascular mortality. Biological Psychology, 75, 32-36. Chareonthaitawee, P., Gibbons, R. J., Roberts, R. S., Christian, T. F., Burns, R., Yusuf, S., for the CORE investigators (Collaborative Organisation for RheothRx Evaluation). (2000). The impact of time to thrombolytic treatment on outcome in patients with acute myocardial infarction. Heart, 84, 142-148. Chen, W., Woods, S. L., & Puntillo, K. A. (2005). Gender differences in symptoms associated with acute myocardial infarction: A review of the research. Heart & Lung, 34, 240-247. Chico, A., Tomas, A., & Novials, A. (2005). Silent myocardial ischemia is associated with autonomic neuropathy and other cardiovascular risk factors in type 1 and type 2 diabetic subjects, especially in those with microalbuminuria. Endocrine, 27, 213-217. Cicchetti, D. V., & Feinstein, A. R. (1990). High agreement but low kappa: II. Resolving the paradoxes. Journal of Clinical Epidemiology, 43, 551-558. Cockcroft, D. W. & Gault, H. (1976). Prediction of creatinine clearance from serum creatinine. Nephron, 16, 31-41. Culic, V., Eterovic, D., Miric, D., & Silic, N. (2002). Symptom presentation of acute myocardial infarction: Influence of age, sex and risk factors. American Heart Journal, 144, 1012-1017.   147 De Luca, G., Suryapranata, H., Ottervanger, J. P., & Antman, E. M. (2004). Time delay to treatment and mortality in primary angioplasty for acute myocardial infarction: Every minute of delay counts. Circulation, 109, 1223-1225. Deutsch, E., Berger, M., Kussmaul, W. G., Hirshfeld, J. W. J., Herrmann, H. C., & Laskey, W. K. (1990). Adaptation to ischemia during percutaneous transluminal coronary angioplasty. Clinical, hemodynamic, and metabolic features. Circulation, 82, 2044-2051. DeVon, H. A., Ryan, C. J., Ochs, A. L., & Shapiro, M. (2008). Symptoms across the continuum of acute coronary syndromes: Differences between women and men. American Journal of Critical Care, 17, 14-25. DeVon, H. A., Ryan, C. J., & Zerwic, J. J. (2004). Is the medical record an accurate reflection of patients' symptoms during acute myocardial infarction? Western Journal of Nursing Research, 26, 547-560. DeVon, H. A., & Zerwic, J. J. (2002). Symptoms of acute coronary syndromes: Are there gender differences? A review of the literature. Heart & Lung, 31, 235-245. DeVon, H. A., & Zerwic, J. J. (2003). The symptoms of unstable angina: Do women and men differ? Nursing Research, 52, 108-118. DeVon, H. A., Larimer, K. & Penckofer, S. (2008). The association of diabetes and older age with the absence of chest pain during acute coronary syndrome. Western Journal of Nursing Research, 30, 130-144. Dey, S., Flather, G. D., Brieger, D., Gurfinkel, E. P., Steg, P. G., FitzGerald, G., for the GRACE Investigators. (2009). Sex-related differences in the presentation, treatment and outcomes among patients with acute coronary syndromes: The Global Registry of Acute Coronary Events. Heart, 95, 20-26. Dracup, K., & Moser, D. K. (1991). Treatment-seeking behaviour among those with signs and symptoms of acute myocardial infarction. Heart & Lung, 20, 570-575.   148 Drew, B. J., Califf, R. M., Funk, M., Kaufman, E. S., Krucoff, M. W., Laks, M. M. et al. (2005). Practice standards for electrocardiographic monitoring in hospital settings: An American Heart Association scientific statement from the councils on cardiovascular nursing, clinical cardiology, and cardiovascular disease in the young. Journal of Cardiovascular Nursing, 20, 76- 106. Ellermeier, W. (1997). On separating pain from the willingness to report it. Behavioural and Brain Sciences, 20, 448-449. Fang, J., Keenan, N., Dai, S., & Denny, C. (2008). Disparities in adult awareness of heart attack warning signs and symptoms -- 14 states, 2005 [Electronic verstion]. Morbidity and Mortality Weekly Report, 57(7), March 6, 2008-175-179. Feinstein, A. R., & Cicchetti, D. V. (1990). High agreement but low kappa: I. The problems of two paradoxes. Journal of Clinical Epidemiology, 43, 543-549. Fibrinolytic Therapy Trialists’ (FTT) Collaborative Group. (1994). Indications for fibrinolytic therapy in suspected acute myocardial infarction: Collaborative overview of early mortality and major morbidity results from all randomised trials of more than 1000 patients. Lancet, 343, 311-322. Fillingim, R. B., King, C. D., Ribeiro-Dasilva, M. C., Rahim-Williams, B., & Riley, J. L. I. (2009). Sex, gender, and pain: A review of recent clinical and experimental findings. The Journal of Pain, 10, 447-485. Finnegan, J. R., Meischke, H., Zapka, J. G., Meshack, A., Benjamin-Garner, R., Estabrook, B. et al. (2000). Patient delay in seeking care for heart attack symptoms: Findings from focus groups conducted in five U.S. regions. Preventive Medicine, 31, 205-213. Foreman, R. D. (2007). Neurological mechanisms of chest pain and cardiac disease. Cleveland Clinic Journal of Medicine, 74 (S1), S30-S33.   149 Garrahy, P. J., Dean, L. S., Bulle, T. M., Anderson, J. C., Baxley, W. A., & Nath, H. (1991). The influence of balloon inflation duration on the acute angiographic result of percutaneous transluminal coronary angioplasty. Journal of Interventional Cardiology, 4, 181-187. Gersh, B. J., & Anderson, J. L. (1993). Thrombolysis and myocardial salvage: Results of clinical trials and the animal paradigm: Paradoxic or predictable? Circulation, 88, 296-306. Gibbons, R. J., Abrams, J., Chatterjee, K., Daley, J., Deedwania, P. C., Douglas, J. S. et a. (2003). ACC/AHA 2002 guideline update for the management of patients with chronic stable angina—Summary article. Circulation, 107, 149-158. Gibler, W. B., Armstrong, P. W., Ohman, E. M., Weaver, W. D., Stebbins, A. L., & Gore, J. M. (2002). Persistence of delays in presentation and treatment for patients with acute myocardial infarction: The GUSTO-I and GUSTO-III experience. Annals of Emergency Medicine, 39, 123-30. Gibson, S. J., & Farrell, M. (2004). A review of age differences in the neurophysiology of nociception and the perceptual experience of pain. Clinical Journal of Pain, 20, 227-239. Goldberg, R. J., O'Donnell, C., Yarzebski, J., Bigelow, C., Savageau, J., & Gore, J. M. (1998). Sex differences in symptom presentation associated with acute myocardial infarction: A population-based perspective. American Heart Journal, 136, 189-195. Goldberg, R. J., Steg, P. G., Sadiq, I., Granger, C. B., Jackson, E. A., Budaj, A. et al. (2002). Extent of, and factors associated with, delay to hospital presentation in patients with acute coronary disease (the GRACE registry). American Journal of Cardiology, 89, 791-796. Goldberg, R. J., Yarzebski, J., Lessard, D., & Gore, J. M. (2000). Decade-long trends and factors associated with time to hospital presentation in patients with acute myocardial infarction: The Worchester Heart Attack Study. Archives of Internal Medicine, 160, 3217-3223.   150 Goodman, S. G., Barr, A., Sobtchouk, A., Cohen, M., Fromell, G. J., Laperrière, L. et al. (2000). Low molecular weight heparin decreases rebound ischemia in unstable angina or non-Q- wave myocardial infarction: The Canadian ESSENCE ST-segment monitoring substudy. Journal of the American College of Cardiology, 36, 1507-1513. Granot, M., Goldstein-Ferber, S., & Azzam, Z. S. (2004). Gender differences in the perception of chest pain. Journal of Pain and Symptom Management, 27, 149-155. Gruppo Italiano per lo Studio della Streptochinasi nell’Infarto Miocardico (GISSI). (1986). Effectiveness of intravenous thrombolytic treatment in acute myocardial infarction. Lancet, 1, 397-402. Hackett, T. P., & Cassem, N. H. (1969). Factors contributing to delay in responding to the signs and symptoms of acute myocardial infarction. American Journal of Cardiology, 24, 651-658. Hayes, S. N., Long, T., Hand, M. M., Finnegan, J. R. J., & Selker, H. P. (2004). Women's Ischemic Syndromes Evaluation: Current status and future research directions: Report of the National Heart, Lung and Blood Institute workshop October 2–4, 2002: Section 6: Key messages about acute ischemic heart disease in women and recommendations for practice. Circulation, 109, e59-e61. Hazelhurst, L. T., & Claassen, N. (2006). Gender differences in the sweat response during spinning exercise. Journal of Strength and Conditioning Research, 20, 723-724. Health Canada. (2003). Exploring concepts of gender and health. Ottawa, ON: Publications, Health Canada. Heart and Stroke Foundation of Canada. (2000). The changing face of heart disease and stroke in Canada. Ottawa, ON: Author. Retrieved from http://www.phac- aspc.gc.ca/publicat/changing_face-nouveau_visage/pdf/card2k-eng.pdf   151 Heart and Stroke Foundation of Canada. (2008). Heart attack warning signals and cardiac arrest warning signs. Retrieved August 19, 2009, from http://www.heartandstroke.com/site/c.ikIQLcMWJtE/b.3483917/k.171E/Heart_disease_ _Heart_Attack_Warning_Signals.htm Heart and Stroke Foundation of Canada: The Heart Truth. (2008). Warning signs. Retrieved August 19, 2009, from http://www.thehearttruth.ca/warning_signs/ Herlitz, J., Bang, A., Karlson, B. W., & Hartford, M. (1999). Is there a gender difference in aetiology of chest pain and symptoms associated with acute myocardial infarction? European Journal of Emergency Medicine, 6, 311-315. Hermens, W. T., Willems, G. M., Nijssen, K. M., & Simoons, M. L. (1992). Effect of thrombolytic treatment delay on myocardial infarct size. Lancet, 340, 1297. Hosmer, D. W., & Lemeshow, S. (2004). Applied logistic regression (2nd ed.). New York: Wiley. Human Resources and Social Development Canada. (2008). National occupation classification. Retrieved October 31, 2008, from http://www5.hrsdc.gc.ca/NOC- CNP/app/occupation_index.aspx?lc=e Hunt, R. J. (1986). Percent agreement, Pearson's correlation, and kappa as measures of inter- examiner reliability. Journal of Dental Research, 65, 128-130. Inoue, Y., Yoshiko, T., Kaori, O., Kuwahara, T., Ogura, Y., & Ueda, H. (2005). Sex- and menstrual cycle-related differences in sweating and cutaneous blood flow in response to passive heat exposure. European Journal of Applied Physiology, 94, 323-332. Jacobs, A. K., Antman, E. M., Faxon, D. P., Gregory, T., & Solis, P. (2007). Development of systems of care for ST-elevation myocardial infarction patients: Executive summary. Circulation, 116, 217-230. Johnson, J. L., Greaves, L., & Repta, R. (2007). Better science with sex and gender: A primer for health research. Vancouver: BC Women’s Health Research Network.   152 Kato, T., & Yoshimoto, N. (2003). Ischaemic preconditioning and outcomes after angioplasty: Effects of drug therapy. Drugs, 63, 133-138. Kern, E. (1987). Cultural-historical aspects of pain. Acta Neurochirurgica, 38 (Supplement), 165- 181. Kern, M. J., Moore, J. A., Aguirre, F. V., Bach, R. G., Caracciolo, E. A., Wolford, T. et al. (1996). Determination of angiographic (TIMI grade) blood flow by intracoronary doppler flow velocity during acute myocardial infarction. Circulation, 94, 1545-1552. Kidney Disease Outcomes Quality Initiative Workgroup of the National Kidney Foundation. (2002). K/DOQI Clinical practice guidelines for chronic kidney disease: Part 5, guideline 4: Estimation of GFR. Retrieved November 2, 2008, from http://www.kidney.org/professionals/KDOQI/guidelines_ckd/p5_lab_g4.htm King, K. B., & McGuire, M. A. (2007). Symptom presentation and time to seek care in women and men with acute myocardial infarction. Heart & Lung, 36, 235-243. Kloner, R. A., & Yellon, D. (1994). Does ischemic preconditioning occur in patients? Journal of the American College of Cardiology, 24, 1133-1142. Klootwijk, P., Langer, A., Meij, S., Green, C., Veldkamp, R. F., Ross, A. M. et al. (1996). Ischaemic heart disease: Non-invasive prediction of reperfusion and coronary artery patency by continuous ST segment monitoring in the GUSTO-I trial [Electronic verstion]. European Heart Journal, 17, 689-698. Knaak, S. (2004). On the reconceptualizing of gender: Implications for research design. Sociological Inquiry, 74, 302-317. Kosuge, M., Kimura, K., Ishikawa, T., Ebina, T., Hibi, K., Tsukahara, K. (2006). Differences between men and women in terms of clinical features of ST segment acute myocardial infarction. Circulation Journal, 70, 222-226.   153 Kudenchuk, P. J., Maynard, C., Martin, J. S., Wirkis, M., & Weaver, W. D. (1996). Comparison of presentation, treatment and outcomes of acute myocardial infarction in men versus women. (The Myocardial Infarction Treatment and Intervention Registry). American Journal of Cardiology, 78, 9-14. Kushner, F. G.; Hand, M.; Smith, S. C. Jr.; King, S. B. III; Anderson, J. L.; Antman, E. M.; Bailey, S. R.; Bates, E. R.; Blankenship, J. C.; Casey, D. E. Jr.; Green, L. A.; Hochman, J. S.; Jacobs, A. K.; Krumholz, H. M.; Morrison, D. A.; Ornato, J. P.; Pearle, D. L.; Peterson, E. D.; Sloan, M. A.; Whitlow, P. L.; Williams, D. O. (2009). 2009 Focused Updates: ACC/AHA Guidelines for the Management of Patients with ST-Elevation Myocardial Infarction (Updating the 2004 Guideline and the 2007 Focused Update) and ACC/AHA/SCAI Guidelines on Percutaneous Coronary Intervention (Updating the 2005 Guideline and 2007 Focused Update): A Report of the American College of Cardiology Foundation/American Heart Association Task Force on Practice Guidelines. Circulation, 120, 2271-2306. Kyker, K. A., & Limacher, M. C. (2002). Gender differences in the presentation and symptoms of coronary artery disease. Current Women's Health Reports, 2, 115-119. Langeheine, R., Pannekoek, J., & Van de Pol, F. (1996). Bootstrapping goodness-of-fit measures in categorical data analysis. Sociological Methods and Research, 24, 492-516. Langer, A., Freeman, M. R., Josse, R. G., Steiner, G., & Armstrong, P. W. (1991). Detection of silent myocardial ischemia in diabetes mellitus. American Journal of Cardiology, 67, 1073-1078. Lee, H., Bahler, R., Chung, C., Alonzo, A., & Zeller, R. A. (2000). Prehospital delay with myocardial infarction: The interactive effect of clinical symptoms and race. Applied Nursing Research, 13, 125-133.   154 Lefler, L. L., & Bondy, K. N. (2004). Women's delay in seeking treatment with myocardial infarction: A meta-synthesis. Journal of Cardiovascular Nursing, 19, 251-268. Lovlien, M., Schei, B., & Gjengedal, E. (2006). Are there gender differences related to symptoms of acute myocardial infarction? A Norwegian perspective. Progress in Cardiovascular Nursing, 21(1), 14-19. Lovlien, M., Schei, B., & Hole, T. (2006). Women with myocardial infarction are less likely than men to experience chest symptoms. Scandinavian Cardiovascular Journal, 40, 342-347. Manuel, D. G., Leung, M., Nguyen, K., Tanuseputro, P., & Johansen, H. (2006). Burden of cardiovascular disease in Canada. In J. V. Tu, W. A. Ghali, L. Pilote & S. Brien (Eds.), CCORT Canadian cardiovascular atlas (pp. 15-22). Oakville, ON: Pulsus Group and Institute for Clinical Evaluative Sciences. Manzella, D., & Paolisso, G. (2005). Cardiac autonomic activity and type II diabetes mellitus. Clinical Science, 108, 93-99. Maynard, C., & Weaver, W. D. (1992). Treatment of women with acute MI: New findings from the MITI registry. Journal of Myocardial Ischemia, 4, 27-37. McCarthy, B. D., Beshansky, J. R., D'Agostino, R. B., & Selker, H. P. (1993). Missed diagnosis of acute myocardial infarction in the ED: Results from a multicenter study. Annals of Emergency Medicine, 22, 579-582. McGinn, A. P., Rosamond, W. D., Goff, D. C. J., Taylor, H. A., Miles, J. S., & Chambless, L. (2005). Trends in prehospital delay time and use of emergency medical services for acute myocardial infarction: Experience in 4 US communities from 1987-2000. American Heart Journal, 150, 392-400. Meischke, H., Eisenberg, M. S., & Larsen, M. P. (1993). Prehospital delay interval for patients who use emergency medical services: The effect of heart-related medical conditions and demographic variables. Annals of Emergency Medicine, 22, 1597-1601.   155 Meischke, H., Ho, M. T., Eisenberg, M. S., Schaeffer, S. M., & Larsen, M. P. (1995). Reasons patients with chest pain delay or do not call 911. Annals of Emergency Medicine, 25, 193-197. Melloni, C., Peterson, E. D., Chen, A. Y., Szczech, L. A., Newby, L. K., Harrington, R. A. (2008). Cockcroft-Gault versus modification of diet in renal disease: Importance of glomerular filtration rate formula for classification of chronic kidney disease in patients with non-ST-segment elevation acute coronary syndromes. Journal of the American College of Cardiology, 51, 991-996. doi:10.1016/j.jacc.2007.11.045 Merz, C. N., Shaw, L. J., Reis, S. E., Bittner, V., Kelsey, S. F., & Olson, M. (2006). Insights from the NHLBI-sponsored women's ischemia syndrome evaluation (WISE) study: Part II: Gender differences in presentation, diagnosis, and outcome with regard to gender-based pathophysiology of atherosclerosis and macrovascular and microvascular coronary disease. Journal of the American College of Cardiology, 47(3 Suppl), S21-S29. Meshack, A. F., Goff, D. C., Chan, W., Ramsey, D., Linares, A., Reyna, R., & Pandey, D. (1998). Comparison of reported symptoms of acute myocardial infarction in Mexican Americans versus non-Hispanic whites (the Corpus Christi Heart Project). American Journal of Cardiology, 82, 1329-1332. Methot, J., Hamelin, B. A., Bogaty, P., Arsenault, M., Plante, S., & Poirier, P. (2004). Does hormonal status influence the clinical presentation of acute coronary syndromes in women? Journal of Women's Health, 13, 695-702. Milavetz, J. J., Giebel, D. W., Christian, T. F., Schwartz, R. S., Holmes, D. R., & Gibbons, R. J. (1998). Time to therapy and salvage in myocardial infarction. Journal of the American College of Cardiology, 31, 1246-1251. Milner, K. A., Funk, M., Arnold, A., & Vaccarino, V. (2002). Typical symptoms are predictive of acute coronary syndromes in women. American Heart Journal, 143, 283-288.   156 Milner, K. A., Funk, M., Richards, S., Wilmes, R. M., Vaccarino, V., & Krumholz, H. M. (1999). Gender differences in symptom presentation associated with coronary heart disease. The American Journal of Cardiology, 84, 396-399. Milner, K. A., Vaccarino, V., Arnold, A. L., Funk, M., & Goldberg, R. J. (2004). Gender and age differences in chief complaints of acute myocardial infarction (Worcester Heart Attack Study). The American Journal of Cardiology, 93, 606-608. Morgan, D. M. (2005). Effect of incongruence of acute myocardial infarction symptoms on the decision to seek treatment in a rural population. Journal of Cardiovascular Nursing, 20, 365-371. Mork, A. A., Haufe, S. M. W., & Yancey, W. B. (2004). Sometimes (what seems to be) a heart attack is (really) a pain in the neck. Journal of the American Board of Family Practice, 17, 74-77. Morris, D. B. (1991). The culture of pain. Berkeley, CA: University of California Press. Mosca, L., Ferris, A., Fabunmi, R., & Robertson, R. M., American Heart Association. (2004). Tracking women's awareness of heart disease: An American Heart Association national study. Circulation, 5, 573-579. Moser, D. K., McKinley, S., Dracup, K., & Chung, M. L. (2005). Gender differences in reasons patients delay in seeking treatment for acute myocardial infarction symptoms. Patient Education and Counseling, 56(1), 45-54. Murphy, S. A., Chen, C., Cannon, C. P., Antman, E. M., & Gibson, C. M. (2002). Impact of gender on angiographic and clinical outcomes after fibrinolytic therapy in acute myocardial infarction. American Journal of Cardiology, 90, 766-770. Muthén & Muthén (2007). MPlus® [Computer software, version 5.21]. Los Angeles, CA: Muthén & Muthén. Newby, L. K., Rutsch, W. R., Califf, R. M., & et al. for the GUSTO-1 Investigators. (1996). Time from symptom onset to treatment and outcomes after thrombolytic therapy. Journal of the American College of Cardiology, 27, 1646-1655.   157 O'Donnell, S., Condell, S., Begley, C., & Fitzgerald, T. (2006). Prehospital care pathway delays: Gender and myocardial infarction. Journal of Advanced Nursing, 53, 268-276. Omran, S., & Al-Hassan, M. (2006). Gender differences in signs and symptoms presentation and treatment of Jordanian myocardial infarction patients. International Journal of Nursing Practice, 12, 198-204. Oprah.com Plus. (2009). The facts about heart disease for women. Retrieved May 2, 2009, from http://www.oprah.com/article/oprahshow/health_yourbody_heartfactold O'Rourke, R. A. (1994). Chest pain. In A. R. W. Schlant R.C. (Ed.), The heart: Arteries and veins. (pp. 459-467). New York: McGraw-Hill. Ottesen, M. M., Kober, L., Jorgensen, S., & Torp-Pedersen, C. (1996). Determinants of delay between symptoms and hospital admission in 5978 patients with acute myocardial infarction. European Heart Journal, 17, 429-437. Panting, J. R., Gatehouse, P. D., Yang, G. Z., Grothules, F., Firmin, D. N., Collins, P., et al. (2002). Abnormal subendocardial perfusion in cardiac syndrome X detected by cardiovascular magnetic resonance imaging. New England Journal of Medicine, 346, 1948-1953. Patel, H., Rosengren, A., & Ekman, I. (2004). Symptoms in acute coronary syndromes: Does sex make a difference? American Heart Journal, 148(1), 27-33. Peterson, J. C., Adler, S., Burkart, J. M., Greene, T., Hebert, L. A., Hunsicker, L. G. et al. (1995). Blood pressure control, proteinuria, and the progression of renal disease: The Modification of Diet in Renal Disease study. Annals of Internal Medicine, 123, 754-762. Pilote, L., Dasgupta, K., Guru, V., Humphries, K. H., McGrath, J., & Norris, C., et al. (2007). A comprehensive view of sex-specific issues related to cardiovascular disease. Canadian Medical Association Journal, 176(6 (Suppl)), S1-S144.   158 Pope, J. H., Aufderheide, T. P., Ruthazer, R., Woodlard, R. H., Feldman, J. A., Beshansky, J. R., et al. (2000). Missed diagnosis of acute cardiac ischemia in the ED. New England Journal of Medicine, 342, 1163-1170. Public Health Agency of Canada. (2008). Diabetes in Canada: Highlights from the national diabetes surveillance system, 2004-2005. Retrieved June 23, 2009, from http://www.phac- aspc.gc.ca/publicat/2008/dicndss-dacsnsd-04-05/pwdd-iadd-eng.php Quyyumi, A. A. (2006). Women and ischemic heart disease: Pathophysiologic implications from the Women’s Ischemia Syndrome Evaluation (WISE) study and future research steps. Journal of the American College of Cardiology, 47, 66S-71S. Ratner, P. A., Johnson, J. L., Mackay, M., Tu, A. W., & Hossain, S. (2008). Knowledge of “heart attack” symptoms in a Canadian urban community. Clinical Medicine: Cardiology, 2, 201-213. Ratner, P. A., Tzianetas, R., Tu, A. W., Johnson, J. L., Mackay, M., Buller, C. E., et al. (2006). Acute myocardial infarction symptom recognition by the lay public: The role of gender and ethnicity. Journal of Epidemiology and Community Health, 60, 606-615. Reimann, R., & Jahrmarker, H. (1969). [On the clinical findings in myocardial infarct: Symptomatology and complications.]. Münchener Medizinische Wochenschrift, 111, 923-930. Richards, S. B., Funk, M., & Milner, K. A. (2000). Differences between blacks and whites with coronary heart disease in initial symptoms and in delay in seeking care. American Journal of Critical Care, 9, 237-244. Riegel, B., Moser, D. K., Carlson, B., Deaton, C., Armola, R., Sethares, K., et al. (2003). Gender differences in quality of life are minimal in patients with heart failure. Journal of Cardiac Failure, 9(1), 42-48. Rose, G. A. (1962). The diagnosis of ischaemic heart pain and intermittent claudication in field surveys. Bulletin of the World Health Organization, 27, 645-658.   159 Rosenfeld, A. G. (2006). State of the heart: Building science to improve women's cardiovascular health. American Journal of Critical Care, 15, 556-567. Rosenfeld, A. G., Perrin, N. A., & Darney, B. G. (2005). Symptom clusters in women with acute myocardial infarction. Circulation, 111(4), e56. Ryan, C. J., DeVon, H. A., Horne, R., King, K. B., Milner, K., Moser, D. K., et al. (2007). Symptom clusters in acute myocardial infarction: A secondary data analysis. Nursing Research, 56, 72-81. Sedlak, T. L., Pu, A., Aymong, E., Gao, M., Khan, N., Quan, H., et al. (in press). Sex differences in coronary catheterisation and revascularization following acute myocardial infarction: Time trends from 1994 to 2003 in British Columbia, Canada. Canadian Journal of Cardiology. Shaw, L. J., Merz, C. N., Pepine, C. J., Reis, S. E., Bittner, V., & Kelsey, S. F. et al., for the WISE Investigators. (2006). Insights from the NHLBI-sponsored Women’s Ischemia Syndrome Evaluation (WISE) study. Part I: Gender differences in traditional and novel risk factors, symptom evaluation, and gender-optimized diagnostic strategies. Journal of the American College of Cardiology, 47, 4S-20S. Sheifer, S. E., Rathore, S. S., Gersh, B. J., Weinfurt, K. P., Oetgen, W. J., Breall, J. A., et al. (2000). Time to presentation with acute myocardial infarction in the elderly: Associations with race, sex, and socioeconomic characteristics. Circulation, 102, 1651-1656. Sheskin, D. J. (2007). Handbook of parametric and non-parametric statistical procedures (4th ed.). Boca Raton, FL: Chapman & Hall/CRC. Shin, J. Y., Martin, R., & Howren, B. (2009). Influence of assessment methods on reports of gender differences in AMI symptoms. Western Journal of Nursing Research, 31, 553-568. Skrondal, A., & Rabe-Hesketh, S. (2004). Generalized latent variable modeling. Multilevel, longitudinal, and structural equation models. New York: Chapman & Hall/CRC.   160 Spitzer, D. L. (n.d.). In Gender and Sex-Based Analysis Advisory Committee, Institute of Gender and Health (Ed.), Gender and sex-based analysis in health research: A guide for CIHR researchers and reviewers. Ottawa, ON: Canadian Institutes of Health Research.  Retrieved from http://www.cihr-irsc.gc.ca/e/32019.html#6 Statistics Canada. (2006). Canadian community health survey, cycle 3.1. Retrieved November 7, 2008, from http://data.library.ubc.ca/java/jsp/database/production/detail.jspid=1010/usr/datalib/dat afiles/survey/statscan./cchs/cycle3.1 Statistics Canada. (2007a). Age-standardized mortality rates by selected causes, both sexes. (Table No. Table 102-0552, Catalogue no. 84F0209X). Ottawa, ON: Statistics Canada. Retrieved from http://www40.statcan.ca/l01/cst01/health30a.htm Statistics Canada. (2007b). Mortality: Summary list of causes. (The Daily new release No. April 27, 2007). Ottawa, ON: Statistics Canada. Retrieved from http://www.statcan.ca/Daily/English/070427/d070427b.htm Statistics Canada. (2008a). 2006 census of population: Ethnic origin, single and multiple ethnic origin responses and sex for the population of Canada, provinces, territories, census metropolitan areas and census agglomerations. Retrieved August 10, 2008, from http://www12.statcan.ca/english/census06/data/topics/RetrieveProductTable.cfm?TPL= RETR&ALEVEL=3&APATH=3&CATNO=&DETAIL=0&DIM=&DS=99&FL=0&FR EE=0&GAL=0&GC=99&GK=NA&GRP=1&IPS=&METH=0&ORDER=1&PID=923 33&PTYPE=88971&RL=0&S=1&ShowAll=No&StartRow=1&SUB=801&Temporal=20 06&Theme=80&VID=0&VNAMEE=&VNAMEF=#FN4 Statistics Canada. (2008b). Appendix G. mother tongue, home language and language of work: Classifications from 2006, 2001 and 1996. Retrieved August 11, 2008, from http://www12.statcan.ca/english/census06/reference/dictionary/app007.cfm   161 Steg, P. G., Bonnefoy, E., Chabaud, S., Lapostolle, F., Dubien, P. Y., Cristofini, P., Leizorovicz, A., & Touboul, P. (Comparison of Angioplasty and Prehospital Thrombolysis In acute Myocardial infarction [CAPTIM] Investigators). (2003). Impact of time to treatment on mortality after prehospital fibrinolysis or primary angioplasty: Data from the CAPTIM randomized clinical trial. Circulation, 108, 2851-2856. Stephen, S. A., Darney, B. G., & Rosenfeld, A. G. (2008). Symptoms of ACS in women with diabetes: An integrative review of the literature. Heart & Lung, 37, 179-189. Stern, S., Behar, S., Leor, J., Harpaz, D., Boyko, V., Gottlieb, S., et al. (2004). Presenting symptoms, admission electrocardiogram, management, and prognosis in acute coronary syndromes: Differences by age. American Journal of Geriatric Cardiology, 13, 188-196. Stern, U. M., Salzer, B., Schuch, S., & Hornstein, O. P. (1998). Sex-dependent differences in sweating of normal probands and atopic patients in cardiovascular stress. Hautarzt, 49, 209- 215. Stocks, J. T. (2001). Statistics for social workers. In B. A. Thyer (Ed.), The handbook of social work research methods (pp. 81-131). Thousand Oaks, CA: Sage. Stone, M. A., Khunti, K., Squire, S., & Paul, S. (2008). Impact of co-morbid diabetes on quality of life and perception of angina pain in people with angina registered with general practitioners in the UK. Quality of Life Research, 17, 887-894. Teoh, M., Lalondrelle, S., Roughton, M., Grocott-Mason, R., & Dubrey, S. W. (2007). Acute coronary syndromes and their presentation in Asian and Caucasian patients in Britain. Heart, 93, 183-188. The Criteria Committee of the New York Heart Association, Inc. (Kossman, C. E., chairman). (1964). Nomenclature and criteria for diagnosis. Diseases of the heart and blood vessels. (6th ed., pp. 110-114). Boston, MA: Littel, Brown.   162 The Task Force on the Management of Stable Angina Pectoris of the European Society of Cardiology. (2006). Guidelines on the management of stable angina pectoris: Full text [Electronic verstion]. European Heart Journal, 27, 1341-1381. The TIMI Study Group. (1984). The Thrombolysis In Myocardial Infarction (TIMI) trial: Phase I findings. New England Journal of Medicine, 33, 523-530. Then, K. L., Rankin, J. A., & Fofonoff, D. A. (2001). Atypical presentation of acute myocardial infarction in 3 age groups. Heart & Lung: The Journal of Acute and Critical Care, 30, 285-293. Thuresson, M., Jarlov, M. B., Lindahl, B., Svensson, L., Zedigh, C., & Herlitz, J. (2005). Symptoms and type of symptom onset in acute coronary syndrome in relation to ST elevation, sex, age, and a history of diabetes. American Heart Journal, 150, 234-242. Tomai, F., Crea, F., Gaspardone, A., Versaci, F., De Penta, D., & et al. (1994). Ischemic preconditioning during coronary angioplasty is prevented by glibenclamide, a selective ATP- sensitive K+ channel blocker. Circulation, 909, 700-705. UCLA Academic Technology Services. Statistical computing seminars: Latent class analysis in MPlus. Retrieved January 29, 2008, from http://www.ats.ucla.edu/stat/mplus/seminars/lca/default.htm Uebersax, J. (2006). LCA: Frequently asked questions. Retrieved January 29, 2008, from http://ourworld.compuserve.com/homepages/jsuebersax/faq.htm Ungvari, G. S., Goggins, W., Leung, S., Lee, E., & Gerevich, J. (2009). Schizophrenia with prominent catatonic features ('catatonic schizophrenia') III. Latent class analysis of the catatonic syndrome. Progress in Neuro-Psychopharmacology & Biological Psychiatry, 33(1), 81-85. Unruh, A. M. (1996). Gender variations in clinical pain experience. Pain, 65, 123-167. Vaccarino, V., Parsons, L., Every, N. R., Barron, H. V., & Krumholz, H. M. (1999). Sex-based differences in early mortality after myocardial infarction: National Registry of Myocardial Infarction-2 participants. New England Journal of Medicine, 341, 217-225.   163 Van de Werf, F., Bax, J., Betriu, A., Blomstrom-Lundqvist, C., Crea, F., Falk, V., et al. (2008). Management of acute myocardial infarction in patients presenting with persistent ST- segment elevation. European Heart Journal, 29, 2909-2945. Wagner, G. S., & Marriot, H. J. L. (2008). Practical electrocardiography. Philadelphia: Wolters Kluwer Health/Lippincott Williams & Wilkins. Wizemann, T. M., & Pardue, M. L. (Committee on Understanding the Biology of Sex and Gender Differences, Board on Health Sciences Policy, Institute of Medicine) (Eds.). (2001). Exploring the biological contributions to human health: Does sex matter? Washington, DC: National Academies Press. doi:http://www.nap.edu/catalog/10028.html World Health Organization. (1998). Gender and health: Technical paper. Geneva, Switzerland: World Health Organization. World Medical Association. (2008). Declaration of Helsinki. Retrieved November 4, 2008, from http://www.wma.net/e/policy/b3.htm Yong, H. H. (2006). Can attitudes of stoicism and cautiousness explain observed age-related variation in levels of self-rated pain, mood disturbance and functional interference in chronic pain patients? European Journal of Pain, 10, 399-407. Zerwic, J. J., Ryan, C. J., DeVon, H. A., & Drell, J. (2003). Treatment-seeking for acute myocardial infarction: Differences in delay across sex and race. Nursing Research, 52, 159-167.   164            APPENDIX A:  Research Ethics Boards Approvals   165   166   167   168            APPENDIX B:  Informed Consent Forms   169      October 2004 CONSENT FORM   Title of Study: Patients’ Sensations during Angioplasty  Principal Investigators: Pamela A. Ratner, PhD, RN Associate Professor, School of Nursing, UBC 604-822-7427  Martha Mackay, MSN, RN Adjunct Professor, School of Nursing, UBC Clinical Nurse Specialist, St. Paul’s Hospital 604-682-2344 (63127)  Co-Investigators: Christopher Buller, MD, Clinical Professor, Cardiology, Department of Medicine, UBC Rebecca Fox, PA, MSc, Research Coordinator, Vancouver Hospital Joy Johnson, PhD, RN, Professor, School of Nursing, UBC Bruno Zumbo, PhD, Professor, Educational & Counseling Psychology, UBC  Researchers at the School of Nursing at UBC, St. Paul’s Hospital and Vancouver Hospital are currently conducting research to evaluate patients’ experiences during angioplasty. You are being invited to take part in this research study because you are scheduled for this procedure. Your participation is entirely voluntary so it is up to you to decide whether to take part in this study. Before you decide, it is important for you to understand what the research involves. This consent form tells you about the study and what will happen to you during the study, if you decide to participate.  If you wish to participate, you will be asked to sign this form. If you do decide to take part in this study, you are still free to withdraw at any time and without giving any reasons for your decision.  If you do not wish to participate, you do not have to provide any reason for your decision nor will you lose the benefit of any medical care to which you are entitled or are presently receiving.  Please take time to read the following information carefully and the research coordinator will be happy to answer any of your questions.   170 Purpose This study is being conducted to understand what patients’ sensations and feelings are when they are undergoing angioplasty.  Study Procedures You will be asked to sign a separate consent for the angioplasty procedure (if you have not already done so), once it has been explained to you by the doctor. Then your angioplasty procedure will be carried out in the usual way. We are most interested in what patients feel as the balloon is inflated (blown up) during the angioplasty. For this reason, it is possible that the balloon might be inflated for slightly longer than usual (no more than two (2) minutes), so that we can ask you about what you are feeling. You will be asked questions that will allow you to describe the sensations you are having both during today’s procedure and also when you have had heart symptoms in the past. Participation in this study involves responding to these questions about what you are feeling. You will not be required to spend any additional time in the Catheterisation Lab or with the Research Coordinator beyond that required for your usual medical care. Other information about your heart condition, the results of your angioplasty, and some general information about your background (age, education, language spoken, etc.) will be recorded onto forms.  Risks and Benefits Inflating the balloon in the heart artery causes a temporary block in blood supply to the area of the heart beyond the balloon. This happens with all angioplasty procedures. Most often the balloon is left inflated for 15 to 45 seconds. When the balloon is blown up, some patients feel discomfort or tightness in their chest or other areas, or may feel short of breath or sweaty. This is expected, and passes quickly (within seconds) after the balloon is deflated. In this study, the balloon inflation may last slightly longer than required for a usual procedure, but NO MORE THAN 2 minutes. Until recently it was very common to use balloon inflations of 2 minutes or more in all patients, and even up to 15 minutes in some patients. These longer inflations are still used in some patients today, as part of the standard angioplasty procedure, with no ill-effect, provided they are properly observed. Although patients may experience some discomfort from this, years of experience has shown that these symptoms pass within seconds after the balloon is deflated, even with the very long inflations. It has also been shown that there are no lasting side effects from having the balloon inflated for this amount of time, other than the possible side effects from a standard angioplasty procedure that have been explained to you by the cardiologist. It is unlikely that you will experience more severe discomfort or show any signs of not tolerating the balloon inflation, but if you do, the cardiologist will immediately deflate the balloon.  Your participation in this study, however, may help increase our understanding of patients’ experiences during angioplasty. We hope that the information learned from this study can be used in the future to benefit other people in similar situations.   Withdrawal from the Study Your participation in this research is entirely voluntary and you may withdraw from this study at any time. If you decide to enter the study and to withdraw at any time in the future, there will be no penalty and your future medical care will not be affected.    171 The study investigators may decide to discontinue the study at any time, or withdraw you from the study at any time, if they feel that it is in your best interests.  Confidentiality Your confidentiality will be respected.  No information that discloses your identity will be released or published without your specific consent to the disclosure.  However, research records and medical records identifying you may be inspected in the presence of the Investigators and the Providence Health Care/UBC Research Ethics Board for the purpose of monitoring the research.  However, no records which identify you by name or initials will be allowed to leave the Investigators’ offices.  For More Information If you have any questions or desire further information, you can contact Dr. Pamela Ratner at 604-822-7427 or Ms. Martha Mackay at 604-682-2344 ext. 63127, the Co-Principal Investigators. If you have any concerns about your treatment or your rights as a research subject, you may contact Dr. Steve Shalansky, Chair of the Research Ethics Board at Providence Health Care at 604-682-2344 ext. 62325.   172 Consent • I have read and understood the consent form. • I have been given a copy of this consent form. • I have had sufficient time to consider the information provided and to ask questions and have had satisfactory responses to my questions. • I understand that all of the information will be kept confidential and that the results will only be used for scientific purposes. • I understand that my participation is voluntary and that I am completely free to refuse to participate or to withdraw from this study at any time without changing in any way the quality of care that I receive. • I understand that there is no guarantee that this study will provide any benefits to me. • I understand that I am not waiving any of my legal rights as a result of signing this consent form. • I have read this form and I freely consent to participate in this study.    _________________________ ______________________  ____________ Printed Name of Participant  Signature    Date    _________________________ ______________________  ____________ Printed Name of Witness  Signature    Date    _________________________ ______________________  ____________ Printed Name of    Signature    Date Principal Investigator or Designate Obtaining Consent     If you would like to receive a summary of the research findings please tick the box and complete the information where you would like the summary mailed.  Yes, I would like to receive a summary of the research findings: o  Where to mail the summary:________________________________________________  ______________________________________________________________________    173    November 2003  CONSENT FORM   Title of Study: Patients’ Sensations during Angioplasty  Principal Investigators: Pamela A. Ratner, PhD, RN Associate Professor, School of Nursing, UBC 604-822-7427  Martha Mackay, MSN, RN Adjunct Professor, School of Nursing, UBC Clinical Nurse Specialist, St. Paul’s Hospital 604-682-2344 (63127)  Co-Investigators: Christopher Buller, MD, Clinical Professor, Cardiology, Department of Medicine, UBC Rebecca Fox, PA, MSc, Research Coordinator, Vancouver Hospital Joy Johnson, PhD, RN, Professor, School of Nursing, UBC Bruno Zumbo, PhD, Professor, Educational & Counselling Psychology, UBC  Researchers at the School of Nursing at UBC, St. Paul’s Hospital and Vancouver Hospital are currently conducting research to evaluate patients’ experiences during angioplasty. You are being invited to take part in this research study because you are scheduled for this procedure. Your participation is entirely voluntary so it is up to you to decide whether to take part in this study. Before you decide, it is important for you to understand what the research involves. This consent form tells you about the study and what will happen to you during the study, if you decide to participate.  If you wish to participate, you will be asked to sign this form. If you do decide to take part in this study, you are still free to withdraw at any time and without giving any reasons for your decision.  If you do not wish to participate, you do not have to provide any reason for your decision nor will you lose the benefit of any medical care to which you are entitled or are presently receiving.  Please take time to read the following information carefully and the research coordinator will be happy to answer any of your questions.     174 Purpose This study is being conducted to understand what patients’ sensations and feelings are when they are undergoing angioplasty.   Study Procedures You will be asked to sign a separate consent for the angioplasty procedure (if you have not already done so), once it has been explained to you by the doctor. Then your angioplasty procedure will be carried out in the usual way. We are most interested in what patients feel as the balloon is inflated (blown up) during the angioplasty. For this reason, it is possible that the balloon might be inflated for slightly longer than usual (no more than two (2) minutes), so that we can ask you about what you are feeling. You will be asked questions that will allow you to describe the sensations you are having both during today’s procedure and also when you have had heart symptoms in the past. Participation in this study involves responding to these questions about what you are feeling. You will not be required to spend any additional time in the Catheterisation Lab or with the Research Coordinator beyond that required for your usual medical care. Other information about your heart condition, the results of your angioplasty, and some general information about your background (age, education, language spoken, etc.) will be recorded onto forms.  Risks and Benefits Inflating the balloon in the heart artery causes a temporary block in blood supply to the area of the heart beyond the balloon. This happens with all angioplasty procedures. Most often the balloon is left inflated for 15 to 45 seconds. When the balloon is blown up, some patients feel discomfort or tightness in their chest or other areas, or may feel short of breath or sweaty. This is expected, and passes quickly (within seconds) after the balloon is deflated. In this study, the balloon inflation may last slightly longer than required for a usual procedure, but NO MORE THAN 2 minutes. Until recently it was very common to use balloon inflations of 2 minutes or more in all patients, and even up to 15 minutes in some patients. These longer inflations are still used in some patients today, as part of the standard angioplasty procedure, with no ill-effect, provided they are properly observed. Although patients may experience some discomfort from this, years of experience has shown that these symptoms pass within seconds after the balloon is deflated, even with the very long inflations. It has also been shown that there are no lasting side effects from having the balloon inflated for this amount of time, other than the possible side effects from a standard angioplasty procedure that have been explained to you by the cardiologist. It is unlikely that you will experience more severe discomfort or show any signs of not tolerating the balloon inflation, but if you do, the cardiologist will immediately deflate the balloon.  Your participation in this study, however, may help increase our understanding of patients’ experiences during angioplasty. We hope that the information learned from this study can be used in the future to benefit other people in similar situations.   175 Withdrawal from the Study Your participation in this research is entirely voluntary and you may withdraw from this study at any time. If you decide to enter the study and to withdraw at any time in the future, there will be no penalty and your future medical care will not be affected.  The study investigators may decide to discontinue the study at any time, or withdraw you from the study at any time, if they feel that it is in your best interests.  Confidentiality Your confidentiality will be respected.  No information that discloses your identity will be released or published without your specific consent to the disclosure.  However, research records and medical records identifying you may be inspected in the presence of the Investigators and the UBC Research Ethics Board for the purpose of monitoring the research. However, no records which identify you by name or initials will be allowed to leave the Investigators’ offices.  For More Information If you have any questions or desire further information, you can contact Dr. Pamela Ratner at 604-822-7427 or Ms. Martha Mackay at 604-682-2344 ext. 63127, the Co-Principal Investigators. If you have any concerns about your treatment or your rights as a research subject, you may contact the Research Subject Information Line at the University of British Columbia at 604-822-8598.    176 Pamela A. Ratner Martha Mackay Consent • I have read and understood the consent form. • I have been given a copy of this consent form. • I have had sufficient time to consider the information provided and to ask questions and have had satisfactory responses to my questions. • I understand that all of the information will be kept confidential and that the results will only be used for scientific purposes. • I understand that my participation is voluntary and that I am completely free to refuse to participate or to withdraw from this study at any time without changing in any way the quality of care that I receive. • I understand that there is no guarantee that this study will provide any benefits to me. • I understand that I am not waiving any of my legal rights as a result of signing this consent form. • I have read this form and I freely consent to participate in this study.    _________________________ ______________________  ____________ Printed Name of Participant  Signature    Date   _________________________ ______________________  ____________ Printed Name of Witness  Signature    Date   _________________________ ______________________  ____________ Printed Name of    Signature    Date Principal Investigator   _________________________ ______________________  ____________ Printed Name of    Signature    Date Principal Investigator   If you would like to receive a summary of the research findings please tick the box and complete the information where you would like the summary mailed.  Yes, I would like to receive a summary of the research findings: o  Where to mail the summary:________________________________________________  _______________________________________________________________________   177            APPENDIX C:  Data Collection Tools   178                Patients’ Sensations During Angioplasty DEMOGRAPHIC AND SYMPTOM FORM                  Please return to NAHBR   Attention: Pam Ratner, PhD, RN Email: pam.ratner@ubc.ca   School of Nursing Fax: (604) 822-7869   University of British Columbia Tel: (604) 822-7427   T201-2211 Wesbrook Mall, Vancouver, BC V6T2B5   179 Patients’ Sensations During Angioplasty  Centre:   St. Paul’s OR    Vancouver Hospital      R.A. Initials __________  PATIENT INFORMATION  DEMOGRAPHIC INFORMATION 1-What was the first language that you spoke?      2- What is your highest level of education? Language   (Check one) Education (Check one1)  Chinese1                    Croatian2            Hindi3  English4                     Farsi5                            Punjabi6  French7                       German8                    Italian9  Other10,specify: __________________________ 3- Most people in Canada describe themselves as Canadian first but also identify themselves based on their background or the nationality of their ancestors. What would you say is your main ethnic background? (Check all that apply)  Aboriginal/1st Nations1                           African2            Australian3        British4                Canadian5                            Caribbean6               Chinese7            European8                    Fijian9                          French10             Hindi11  Hispanic12                Japanese13        Korean14  Middle Eastern15                                          Punjabi16            Russian17          Scottish18          Sri Lankan19  Taiwanese20            Filipino21          Welsh22  Other23,specify: __________________________  No Schooling1  Elementary incomplete2  Elementary complete (to Grade 7)3  Junior High incomplete4  Junior High complete (Grades 8-9)5  High School incomplete6  High School complete (Grades 10-13)7  Non-university incomplete8  Non-university complete9  University incomplete10  University diploma/certificate11  University bachelor’s degree12  University professional degree (law or medicine)13  University master’s degree14  University doctorate15  Other16,specify: __________________________ 4-What is your occupation, that is the kind of work you do?____________________________________ ___________________________________________ 5- How many years have you lived in Canada? ______________________________________________ ______________________________________________  Place Patient Sticker Here   Date of  PCI:     |__|__| |__|__|__| |__|__|                  dd               mmm              yr   180 Patients’ Sensations During Angioplasty__________________          (Cont’d) INDEX QUESTION Question 1: Thinking back over the last several months, we would like to know about any discomfort you have been experiencing that has led to your being referred for this procedure.  a- Tell me where the main location has been for any discomfort you have had? ___________________________________________________  b- How severe was this discomfort on a scale from 1 to 10, 1 being very minimal to 10 being most intense?  ___________________________________________________  c- In 1 or 2 words, how would you describe your discomfort?  ___________________________________________________  Examples:  squeezing, pressing, burning, constricting, a weight, knife, sharp, tearing, aching, exploding, etc. Question 2:  a- Did you feel discomfort elsewhere? (Check all “discomfort locations” that apply or specify other.)  b- How severe was this discomfort on a scale from 1 to 10, 1 being very minimal to 10 being most intense?  c- In 1 or 2 words, how would you describe your discomfort?  Did you feel anything else?  Repeat question 2 until the patient no longer expresses discomfort.  Check all that apply Intensity (1-10) Descriptor DISCOMFORT LOCATION Central chest Other Chest Back Right Arm Left Arm Throat Jaw Other 1 (specify) Other 2 (specify) ASSOCIATED SYMPTOM Diaphoresis Dizziness, lightheadedness Dyspnea Epigastric, indigestion Fatigue, weakness Nausea, vomiting Other 1 specify Other 2 specify  FEMORAL PUNCTURE Note the patient’s reaction to femoral puncture (describe response):   Then ask patient:  On a scale of 1 to 10, 1 being hardly noticeable, and 10 being extreme, how uncomfortable did that make you?   181 Patients’ Sensations During Angioplasty__________________         (Cont’d)  1st BALLOON INFLATION Coronary Artery:   RCA   Circ   LAD      Inflation duration (seconds)  _______________  Question 1: At this time, we would like to know about any discomfort you are experiencing.  a. Tell me where the main location is for any discomfort you are having? ___________________________________________________  b. How severe is this discomfort on a scale from 1 to 10, 1 being very minimal to 10 being most intense? ___________________________________________________  c. In 1 or 2 words, how would you describe your discomfort? ________________________________  Examples:  squeezing, pressing, burning, constricting, a weight, knife, sharp, tearing, aching, exploding, etc.   Question 2:  d- Do you feel discomfort elsewhere? (Check all “discomfort locations” that apply or specify other.)  e- How severe is this discomfort on a scale from 1 to 10, 1 being very minimal to 10 being most intense?  f- In 1 or 2 words, how would you describe your discomfort?  Do you feel anything else?  Repeat question 2 until the patient no longer expresses discomfort.  Check all that apply Intensity (1-10) Descriptor DISCOMFORT LOCATION Central chest Other Chest Back Right Arm Left Arm Throat Jaw Other 1 (specify) Other 2 (specify) ASSOCIATED SYMPTOM Diaphoresis Dizziness, lightheadedness Dyspnea Epigastric, indigestion Fatigue, weakness Nausea, vomiting Other 1 specify Other 2 specify Please make note of medications administered during the procedure.     NOTE:  APPEND: 1) A PHOTOCOPY OF THE BC CARDIAC REGISTRY FORM   2) Copy of one ST window with numeric values AND one concurrent real-time 7 of 12 leads (I, II, III, aVR, aVL, aVF, V3) at 3 different times:          I.  Before Procedure;    II.  Before Balloon Inflation;    III.  After Balloon Inflation. Medications & Dosage: Comments:   182   BC Cardiac Registry Data Collection DIAGNOSTIC CATHETERIZATION & INTERVENTIONAL PROCEDURES BCCR number: Referring physician: Date:  Diagnostic only   PCI or other intervention with/without diagnostic Urgency at time of procedure:  Emergent – Must be done without delay (e.g. STEMI; Call back; Hold lab for procedure)  Urgent – Inpatient (including referring hospital); done before discharge  Elective – Outpatient; next available time PART 1:  Physician to complete sections 1 - 7 1. NYHA Functional Class of Heart Failure:  I  II  III  IV 2. CCS Angina:  0  I  II  III  IVa   IVb   IVc   Atypical 3. INDICATIONS FOR PROCEDURE: NON ACS:  Stable angina  Heart failure  Valvular heart disease   Mitral Aortic  Congenital heart disease  Non-ischemic cardiomyopathy  Serious arrhythmia  Other:  ACS:  Acute coronary syndrome ECG changes Ischemia Complicating factors:  STEMI   Ongoing  Hemodynamic instability  ↓ ST   Recurring  Cardiogenic shock  No ST changes  Provokable  CHF  Ambiguous ST  None  None 4. RESULTS OF PROCEDURE – EXTENT OF DISEASE:  Angiographically normal  Disease less than 50%  Single vessel disease  more than 75% Prox LAD  Double vessel disease  more than 75% Prox LAD  Triple vessel disease  more than 75% Prox LAD  Left main disease  more than 50%        more than 70% 5. ANGIOGRAPHER’S INITIAL RECOMMENDATION  Medical management  PCI  Surgery  URGENCY (for PCI and Surgery):  Emergent   Urgent   Semi-Urgent   Elective 6. RESULTS OF PROCEDURE – PCI REVASCULARIZATION  Complete revascularization achieved  Complete revascularization NOT achieved   By intent  Failed PCI  Planned staging/evaluation 7. EJECTION FRACTION:     % Source:    LV angiogram (cath lab)    MIBI    Echocardiogram report   MUGA    Estimated  Physician signature:     Printed name:     183   BC Cardiac Registry Data Collection DIAGNOSTIC CATHETERIZATION & INTERVENTIONAL PROCEDURES  Date: PART 2: RN to complete: Height:    inches    cm  Weight:    lbs   kg   Hemoglobin: OUTCOME DETERMINANTS:  Unknown YES NO Renal function:  On dialysis  Creatinine: ________________  GFR: ________________ Congestive heart failure (current or history):  Prior myocardial infarction:  Hypertension: Hyperlipidemia:  Total cholesterol:    mmol/L  Triglycerides:    mmol/L  HDL:     mmol/L  LDL:     mmol/L Diabetes:  Unknown YES NO  Type I  Type II Insulin  Type II – No insulin    Unknown YES NO Peripheral vascular disease: Cerebrovascular disease: Smoking:  Never    Current    Former – Quit date:  ____________________  Prior PCI: Prior CABG:  Co-morbidities Pulmonary disease   (e.g. COPD, asthma,                                chronic infection) Liver or GI disease   (e.g. hepatitis, peptic                                ulcer, history of GI bleed) Malignancy:  MEDICATIONS IN THE LAST 24 HOURS:  Unknown YES NO Beta blockers IV nitroglycerin Long acting nitrates Calcium antagonists Lipid lowering agents ACE inhibitors/ATII blockers (ARBs)  ASA (aspirin) clopidogrel    Unknown YES NO glyburide Hormone replacement (estrogen) therapy  Anticoagulants Antidepressants Antiarrhythmics digoxin Diuretics   RN signature:     Printed name:     184           APPENDIX D: Raw Symptom Descriptors and Codes   185 Symptom codes and corresponding raw descriptors Code Descriptors Chest, not right 2-sided chest pain  Across  Across chest  Across central chest  Across upper chest  Across front chest  Across the top, across top of chest  Across diaphragm  Across top of chest  All around front of chest  Anterior chest  Both sides of chest  Breast bone  Broadly over chest  Central  Central chest  Central left chest  Central chest - back  Central chest – above esophagus  Central chest - entire chest  Central sternum  Centrally located  Centre of chest  Centre upper chest  Chest  Chest – centre  Chest/left arm  Chest – left side   186 Symptom codes and corresponding raw descriptors (cont’d) Code Descriptors Chest, not right (cont’d) Chest/back  Chest to shoulders  Chest – right in the heart  Chest pain  Chest pain, central chest  Down centre of chest  Front upper chest  High front chest  High upper chest  In the centre through to my back  Left anterior chest wall  Left breast  Left centre of chest  Left chest  Left chest – lower  Left mid chest  Left side  Left side and across chest  Left side of chest  Left side radiate to front of chest  Left upper chest  Low in chest  Lower central chest  Lower chest  Lower chest (middle)  Lower chest central – under sternum  Lower half of chest  Mid-central chest  Mid-chest   187 Symptom codes and corresponding raw descriptors (cont’d) Code Descriptors Chest, not right (cont’d) Mid-chest bilaterally  Mid-chest radiating down left arm  Mid-sternal  Mid-sternal chest pain  Middle of chest  Over centre of chest  Over heart  Over sternum  Pressure across chest  Right across chest  Sternal  Sternum  Sternum – high  Sternum, mid-chest  Top of chest  Top of stomach, central chest  Top of sternum  Under breast bones  Under breast bone on both sides of chest  Under sternum  Underneath left breast  Upper central chest  Upper chest  Upper chest, upper central chest  Upper part of chest  Upper sternum Right chest Centre to right side of chest  Right chest  Right side of chest  Right upper chest   188 Symptom codes and corresponding raw descriptors (cont’d) Code Descriptors Back Across back (upper)  Back  Back pain  Lower back  Middle of back Neck Back of neck  Both sides of neck  Left neck  Lower neck  Neck  Neck – left side  Neck – right side  Neck – back  Neck – ears  Sides of neck  Up neck Throat Back of throat  Burning in throat  Esophagus area  Throat  Throat discomfort Jaw/teeth Jaw  Jaw discomfort  Jaw and teeth  Mouth  Teeth   189 Symptom codes and corresponding raw descriptors (cont’d) Code Descriptors Left Shoulder Back of shoulder (left)  Left shoulder  Left shoulder blade  Shoulder  Scapula Right shoulder Right shoulder Both Shoulders Across shoulder blades  Shoulders  Shoulder blades Intrascapular Between shoulders  Between shoulder blades  Centre of back between shoulder blades Left arm Left arm  Left arm – elbow to shoulders  Left arm and shoulder Right arm Right arm Both arms Arms  Bilateral arm pain  Bilateral elbows  Both arms  Elbows then down to arms  Elbows Underarms Arm pit pain  Left lateral, under axilla  Left axilla  Under arms   190 Symptom codes and corresponding raw descriptors (cont’d) Code Descriptors Epigastrium, indigestion, heartburn All over indigestion  Across diaphragm  Diaphragm  Heartburn  Indigestion  Perigastric area Abdomen, liver, stomach Across stomach  Near liver  Stomach Ribs, rib cage Bottom of ribcage  Lateral sides of ribs  Left side of ribcage Headache Behind eyes and temple  Temple Vomiting Vomiting Aneurysm Area of aneurysm Right leg Down right leg  Right leg Left leg Left leg Leg Leg Right foot Right foot Left ear Left ear Right ear Right ear Both ears Ears Left hand Left hand Right hand Right hand Both hands Hands Left wrist Left wrist Right wrist Right wrist   191 Symptom codes and corresponding raw descriptors (cont’d) Code Descriptors Tongue Tongue Lungs Lungs Shortness of breath Shortness of breath  Dyspnea Anxiety Anxious  Anxiety  Feeling that nothing will help Shaking Body shaking Cold feet Cold feet Cough Cough Cramp Cramp Diarrhea Diarrhea Dry mouth Dry mouth Hot flashes Hot flashes Salivation Mouth salivation  Saliva Palpitations Palpitations Restlessness Restlessness Stomach pushing into lungs Stomach pushing into lungs Cheeks Cheeks Thirsty Thirsty No discomfort Nothing  No pain  No discomfort  No sensation  Nothing 

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}]}"
                            data-media="{[{embed.selectedMedia}]}"
                            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:
https://iiif.library.ubc.ca/presentation/dsp.24.1-0069113/manifest

Comment

Related Items