Self-management of cardiac pain in women: an evidence map.

OBJECTIVE: To describe the current evidence related to the self-management of cardiac pain in women using the process and methodology of evidence mapping. DESIGN AND
SETTING: Literature search for studies that describe the self-management of cardiac pain in women greater than 18 years of age, managed in community, primary care or outpatient settings, published in English or a Scandinavian language between 1 January 1990 and 24 June 2016 using AMED, CINAHL, ERIC, EMBASE, MEDLINE, Proquest, PsychInfo, the Cochrane Library, Scopus, Swemed+, Web of Science, the Clinical Trials Registry, International Register of Controlled Trials, MetaRegister of Controlled Trials, theses and dissertations, published conference abstracts and relevant websites using GreyNet International, ISI proceedings, BIOSIS and Conference papers index. Two independent reviewers screened using predefined eligibility criteria. Included articles were classified according to study design, pain category, publication year, sample size, per cent women and mean age.
INTERVENTIONS: Self-management interventions for cardiac pain or non-intervention studies that described views and perspectives of women who self-managed cardiac pain. PRIMARY AND SECONDARY OUTCOMES MEASURES: Outcomes included those related to knowledge, self-efficacy, function and health-related quality of life.
RESULTS: The literature search identified 5940 unique articles, of which 220 were included in the evidence map. Only 22% (n=49) were intervention studies. Sixty-nine per cent (n=151) of the studies described cardiac pain related to obstructive coronary artery disease (CAD), 2% (n=5) non-obstructive CAD and 15% (n=34) postpercutaneous coronary intervention/cardiac surgery. Most were published after 2000, the median sample size was 90 with 25%-100% women and the mean age was 63 years.
CONCLUSIONS: Our evidence map suggests that while much is known about the differing presentations of obstructive cardiac pain in middle-aged women, little research focused on young and old women, non-obstructive cardiac pain or self-management interventions to assist women to manage cardiac pain. PROSPERO REGISTRATION NUMBER: CRD42016042806. © Article author(s) (or their employer(s) unless otherwise stated in the text of the article) 2017. All rights reserved. No commercial use is permitted unless otherwise expressly granted.

This evidence map provides an overview of a broad range of research based on a systematic search of 20 databases including grey literature sources published in English or a Scandinavian language between 1 January 1990 and 24 June 2016.

Including only studies published in English or Scandinavian language may be seen as a limitation; however, we chose to include a robust search of the grey literature in an attempt to reduce a potential publication bias.

Despite having healthcare providers and researchers involved in establishing the scope of the evidence map, our search strategy may not have targeted what women themselves see as the most important aspects of the self-management of cardiac pain.

Additional search terms and keywords (eg, breathlessness, fatigue, sleep problems) and emotional sequelae (eg, anxiety, depression) may have identified additional self-management interventions targeted to cardiac pain and symptoms in women.

This comprehensive evidence map identified evidence gaps and future research needs specifically focused on the self-management of cardiac pain in women.

Background

In 2015, more than 110 million people worldwide were affected by coronary artery disease (CAD),1 and the number is expected to increase over the next decades.2 CAD is increasing in women less than 55 years of age due to rising obesity and diabetes rates.3 In 2013, 30% more women died of CAD than cancer (including breast cancer) in the USA.4 According to most recent mortality statistics in Europe5 (2015), CAD is the most common cause of death and accounts for 20% of all deaths among women.

Cardiac pain has been considered the primary indicator of CAD,6 and each year 4.5 million individuals are evaluated for cardiac pain in emergency departments in the USA.7 Cardiac pain disproportionately burdens more women than men, and women have a varied pattern and distribution of pain symptoms associated with both obstructive and non-obstructive CAD.8 9 Women also have a higher prevalence of clinically relevant cardiac pain after percutaneous coronary interventions (PCIs) and cardiac surgery and report more persistent pain of moderate to severe intensity up to 2 years after cardiac surgery.10–12 Poorly controlled acute pain (ischaemic and procedural) is a risk factor for persistent pain, a debilitating complication for women following cardiac surgery.13 14

The identification and management of cardiac pain associated with CAD are vital to minimise risk of a major adverse cardiac event (MACE). Women with obstructive and non-obstructive CAD have cardiac pain that differs from that of men. Women describe their pain as sharp and burning, with additional symptoms of discomfort in the jaw, neck, shoulders, arms, back and epigastric area.6 8 Women’s pain and symptoms vary in frequency and distribution, making it difficult for them to interpret as cardiac specific.8 15 Non-obstructive CAD (cardiac syndrome X) is angina-like chest discomfort without evidence of coronary artery obstruction.16 Non-obstructive CAD is more prevalent in younger, middle-aged women and evidence suggests that more extensive, non-obstructive CAD, hypertension and diabetes are associated with MACE similar to those with obstructive CAD.9 17 These women suffer from persistent and incapacitating cardiac pain, are at risk for impaired function, depression, poor health-related quality of life (HRQL) and death.18 They are also frequent users of healthcare services (emergency room, hospitalisation and repeat diagnostic evaluation).19

The risk of future coronary events for women who present with cardiac pain, additional symptoms and/or cardiac pain equivalents can be classified into risk categories that can be used to guide further assessment and evaluation.20 21 In addition to having difficulty interpreting cardiac pain, women minimise their symptoms, prefer to consult with family and friends and have caring responsibilities and concerns for their family.22 As a result, women delay assessment and diagnostic evaluation23; the time from symptom onset to emergency department arrival for women is 85–320 min, and this has not changed in the last decade.24 Over 25% of women will die within a year of their first myocardial infarction (MI) compared with 19% of men, and 47% of women will die within 5 years of their first MI compared with 36% of men.25

The under-recognition of women’s symptoms as well as the difficulty in diagnosing cardiac pain in women contribute to poorer outcomes and greater mortality rates in women as compared with men.6 Outcomes are also associated with a person’s ability to self-manage their condition in everyday life. A variety of new skills have to be learnt: to comply with medication regimes, to establish and sustain new and healthier lifestyle routines, monitor and manage symptoms and recognise when to seek help when symptoms occur.26 Self-management programmes are designed to allow people to take an active part in the management of their condition through problem solving, decision making, action planning, self-tailoring and the formation of patient–provider partnerships.27 28 Due to the complexity of cardiac pain in women, there is an increased need to develop mechanisms to assist women to recognise and manage their pain. We conducted an comprehensive review of the literature to understand the current body of knowledge on self-management programmes for women with cardiac pain using methods described by the Evidence for Policy and Practice Information (EPPI) and the Coordinating Centre at the Institute of Education.29–33 In this paper, we will describe the first step in the EPPI review process, to establish the current evidence related to the self-management of cardiac pain in women using the process and methodology of evidence mapping.

Methods

The main purpose of evidence mapping is to provide an overview of a broad range of research and identify evidence gaps and future research needs.34 The evidence map is broad in scope and is primarily focused on identifying and describing the characteristics of the evidence base.35 It is the first step in conducting an integrated mixed-methods systematic review,29–33 and it does not necessarily include quality appraisal of the included studies.36 Six steps were used to construct an evidence map of cardiac pain in women37: (1) identify the scope of the evidence map; (2) define the key variables; (3) establish a comprehensive search strategy; (4) identify study inclusion and exclusion criteria; (5) systematically retrieve, screen and classify the evidence; and (6) report the findings in an evidence map.

Identify the scope of the evidence map

The initial scope of the work was established by the research team to focus on three types of cardiac pain in women: (1) obstructive CAD, (2) non-obstructive CAD and (3) postprocedure (PCI and cardiac surgery). The research question and the study eligibility criteria were discussed in a consultation session with healthcare providers (physicians and nurses) and researchers working in cardiology, cardiac surgery and adult multidisciplinary chronic pain clinics. An overarching review question was established and purposefully kept broad to ensure a comprehensive review of the evidence: What is known about the self-management of cardiac pain in women? This question could be answered by a broad range of quantitative and qualitative evidence, including systematic reviews, randomised controlled trials (RCTs), cohort studies, cross-sectional, case control studies and case series/reports.

Define the key variables

We used the PICO framework38 to focus our research question and to facilitate the literature search. The PICO question elements included population, intervention, comparison and outcomes. Keywords and the National Library of Medicine’s Medical Subject Headings (MeSH) were combined under the three PICO categories: (P) women, (I) self-management and (O) cardiac pain (table 1). In this instance, we did not search using a comparator to maintain breadth of our evidence map.

Table 1

PICO search strategy

	Population	Intervention	Outcome
	‘Woman’	‘Self-management’	‘Cardiac pain’
	MeSH terms	MeSH terms	MeSH terms
	1. MH ‘Women+’ 2. MH ‘Female’ 3. MH ‘Identity+’ 4. MH ‘Population Characteristics+’ 5. MH ‘Sex Characteristics’ 6. 1 OR 2 OR 3 OR 4 OR 5	1. MH ‘Self Care+’ 2. MH ‘Self Concept+’ 3. MH ‘Self-Assessment’ 4. MH ‘Pain Management’ 5. MH ‘Disease Management+’ 6. MH ‘Health Behaviour+’ 7. MH ‘Activities of Daily Living+’ 8. MH ‘Life Style+’ 9. MH ‘Adaption, Psychological+’ 10. 1 OR 2 OR 3 OR 4 OR 5 OR 5 OR 7 OR 8 OR 9	1. MH ‘Angina Pectoris+’ 2. MH ‘Myocardial Ischemia+’ 3. MH ‘Sternotomy’ 4. MH ‘Cardiovascular Surgical Procedures+’ 5. MH ‘Chest pain+’ 6. MH ‘Acute, Pain’ 7. MH ‘Pain, Postoperative+’ 8. MH ‘Chronic Pain’ 9. (2 OR 3 OR 4) AND (5 OR 6 OR 7 OR 8) 10. 1 OR 5 OR 9
	Keywords	Keywords	Keywords
Combined with OR	1. ‘women’ 2. ‘woman’ 3. ‘female*’ 4. ‘sex diff*’ 5. ‘gender diff*’ 6. 1 OR 2 OR 3 OR 4 OR	1. ‘self-manage*’ 2. ‘self-assess*’ 3. ‘disease manage*’ 4. ‘lifestyle*’ 5. ‘coping’ 6. ‘self-concept*’ 7. ‘(pain* manage*)’ 8. ‘self-care’ 9. ‘health behavior*’ 10. 1 OR 2 OR 3 OR 4 OR 5 OR 6 OR 7 OR 8 OR 9	1. ‘angina*’ 2. ‘sternotomy’ 3. ‘myocardial ischemia*’ 4. ‘heart surgery’ 5. ‘chest pain*’ 6. ‘heart attack’ 7. ‘acute pain*’ 8. ‘postoperative pain*’ 9. ‘(persistent pain* OR chronic pain*)’ 10. (2 OR 3 OR 4 OR 6) AND (5 OR 7 OR 8 OR 9) 11. 1 OR 5 OR 10
	Combined with AND

*Comparison: none.

MeSH, Medical Subject Headings.

Establish a comprehensive search strategy

The literature on the self-management of cardiac pain in women was systematically searched using keywords and MeSH headings in accordance with the search criteria in the bibliographic databases. Publications needed to be available in English or a Scandinavian language and published between 1 January 1990 and 24 June 2016 (inclusive). Searches were conducted in July 2016 using selected databases: AMED (Allied and Complementary Medicine), CINAHL, ERIC, EMBASE, MEDLINE, Proquest, PsychInfo, the Cochrane Library, Scopus, Swemed+ and Web of Science. For ongoing and recently completed clinical trials, we searched the Clinical Trials Registry, International Register of Controlled Trials and the MetaRegister of Controlled Trials. Grey literature sources included theses and dissertation, published conference abstracts and relevant websites using GreyNet International, ISI proceedings, BIOSIS and Conference papers index. Publication citations were exported from electronic search interfaces to Endnote.

Identify study inclusion and exclusion criteria

The inclusion criteria were kept broad, and studies were included if they focused on the self-management of cardiac pain in women or described the views and perspectives of women who had cardiac pain independent of the research design. Types of participants included women who were greater than 18 years of age with cardiac pain, managed in the community, primary care or outpatient settings. Types of outcomes included those related to knowledge, self-efficacy, physical and mental function, social and role function and HRQL. Lastly, the number of women included in the studies needed to be at least 25% to adequately represent the ratio of men and women who have cardiac disease.4

Systematically retrieve, screen and classify the evidence

Title and abstracts of all identified articles were screened in the first round. Studies were excluded if they were not about the self-management of cardiac pain in women or described the views and perspectives of women who had cardiac pain. A random sample of excluded studies (n=50) was discussed between the two reviewers to establish screening accuracy and to confirm understanding of the study eligibility criteria. The second round was based on the full-text screening and followed all the predefined study eligibility criteria. The articles were single-screened due to the large number and time constraints. If the reviewer was unsure about inclusion/exclusion of an article, the second reviewer was consulted to confirm inclusion/exclusion. The included articles were classified and described according to study design as per the hierarchy of evidence (ie, systematic reviews/meta-analyses, intervention studies (RCTs) and non-intervention studies (prospective, retrospective/case control and cross-sectional)) and according to pain category (ie, obstructive CAD, non-obstructive CAD, postprocedure (PCI/cardiac surgery) and mixed). To ensure that characteristics of primary research studies were not over-represented, descriptions of the systematic reviews and systematic reviews with meta-analyses were not included in the final evidence map.39

Report the findings in an evidence map

Only characteristics of the primary research studies (ie, country, study design, pain category, publication year, sample size, per cent women and mean age) were included in the final evidence map. Primary research study bubble plots (ie, weighted scatter plots) were used to graphically illustrate the relationships between: (1) per cent women, study design and type of cardiac pain across year of publication, weighted by sample size and (2) age and type of cardiac pain across year of publication, weighted by sample size. All analyses were performed using Stata statistical software V.13.40

Results

Identification of studies

In total, 6582 eligible citations were identified from searching commercially available bibliographic databases and grey literature sources (Figure 1). After the first screening, a total of 1368 citations were deemed to meet the eligibility criteria. Full-text reports were obtained and processed for 1125 (82%) of the citations. Seventy-four per cent (n=837) of these did not meet the eligibility criteria mainly because the outcomes were not related to cardiac pain or the number of women included was less than 25%. In addition, 6% (n=68) were excluded because they were systematic reviews or systematic reviews with meta-analyses. Twenty per cent (n=220) were primary research studies and were included in the final evidence map: 22% (n=49) were intervention studies (RCTs) and 78% (n=171) were non-intervention studies (prospective, retrospective/case control and cross-sectional).

Figure 1

Flow chart.

Thirty-three per cent (n=73) of the studies were identified through grey literature sources. The majority of these (n=60, 82%) were dissertations related to obstructive CAD.

Characteristics of the primary research studies

Of the 220 primary research studies, 98 (45%) were conducted in North America and 44 (20%) in Western Europe. No relevant studies from the African continent and only one study from South America (ie, Brazil) were identified. Thirteen per cent (n=29) of the studies were published before year 2000, 52% (n=114) between 2000 and 2010 and 35% (n=77) were published after 2010. Sixty-nine per cent (n=151) of the included studies on cardiac pain in the evidence map represented obstructive CAD, 2% (n=5) non-obstructive CAD, 15% (n=34) post-PCI or cardiac surgery pain and 14% (n=30) of studies were a mixed sample of obstructive CAD and post-PCI/cardiac surgery pain. Characteristics of the primary research studies according to pain category are outlined in table 2.

Table 2

Summary of characteristics of primary research studies (n=220)

Pain category	Design (n)	Age Mean Median Range	Sample size Mean Median Range	% Women Mean Median Range	Focus of report (n)	Country (n)
Obstructive CAD	RCT (19)	63 64 48–76	318 165 53–1343	48 40 26–100	Alternative and complementary medicine (2) eHealth or other educational resources (6) Self-management education programmes (11)	Brazil (1), Canada (2), China (1), Denmark (1), Iran (1), Sweden (1), UK (6), USA (6)
	RCT pilot (4)	61 60 55–68	35 30 10–70	63 63 25–100	eHealth or other educational resources (3) Self-management education programmes (1)	Canada (1), USA (3)
	RCT quasi-experimental (3)	63 57 55–76	57 67 7–98	88 100 64–100	eHealth or other educational resources (2) Self-management education programmes (1)	Australia (1), Korea (1), USA (3)
	Prospective cohort (26)	64 62 47–77	176 77 5–1097	64 54 25–100	Clinical presentation (1) eHealth or other educational resources (1) Ethnicity and symptom experience (3) Risk factors (2) Strategies/responses after cardiac events (10) Symptom experience/description (9)	Canada (6), Sweden (4), UK (3), USA (13)
	Retrospective cohort (33)	65 64 54–76	249 62 5–1579	51 43 25–100	Delay in seeking treatment (7) eHealth or other educational resources (1) Ethnicity and symptom experience (5) Self-management education programmes (2) Strategies/responses after cardiac events (8) Symptom experience/description (8)	Canada (2), Denmark (2), Germany (2), Hong Kong (1), Norway (1), Korea (1), Oman (2), Poland (1), Sweden (8), UK (3), USA (10)
	Case control (3)	60 60 55–66	353 30 70–767	31 30 64–100	Alternative and complementary medicine (1) Risk factors (1) Strategies/responses after cardiac events (1)	Australia (1), Iran (1), Sweden (1)
	Cross-sectional (63)	62 62 47–76	330 83 7–7093	59 50 25–100	Alternative and complementary medicine (3) Delay in seeking treatment (2) Ethnicity and symptom experience (8) Risk factors (4) Self-management education programmes (2) Strategies/responses after cardiac events (18) Symptom experience/description (28)	Australia (1), Canada (7), Czech Republic (1), Finland (1), Germany (1), Hong Kong (2), Iran (1), Jordan (1), Norway (2), UK (14), USA (31), Across (1)
Non-obstructive CAD	RCT (2)	56–57	24–53	100	Alternative and complementary medicine (1) Self-management education programmes (1)	UK (1), Sweden (1)
	RCT quasi-experimental (1)	56	9	100	Alternative and complementary medicine (1)	UK (1)
	Cross-sectional (2)	53–62	159–963	100	Clinical presentation (1) Risk factors (1)	Denmark (1), USA (1)
Post-PCI/CABG	RCT (5)	65 66 57–72	117 100 54–250	50 50 28–100	Alternative and complementary medicine (2) eHealth or other educational resources (3)	Germany (1), Greece (1), Iran (1), Thailand (1), USA (1)
	RCT pilot (1)	56	34	38	Clinical presentation (1)	Sweden (1)
	RCT quasi-experimental (1)		56	25	Alternative and complementary medicine (1)	Korea (1)
	Prospective cohort (15)	64 64 56–72	109 52 6–326	60 37 25–100	Alternative and complementary medicine (1) eHealth or other educational resources (1) Risk factors (2) Strategies/responses after cardiac events (6) Symptom experience/description (5)	Australia (3), Canada (3), Germany (2), Norway (1) UK (2), USA (4)
	Retrospective cohort (5)	69 67 63–80	481 57 8–2269	55 44 31–100	Symptom experience/description (2) Strategies/responses after cardiac event (3)	Australia (1), Canada (1), Switzerland (1), USA (2)
	Cross-sectional (7)	62 64 56–65	237 100 8–753	48 37 26–100	Strategies/responses after cardiac events (6) Symptom experience/description (1)	Australia (1), Korea (1), Norway (2), Taiwan (1), Turkey (1), USA (1)
Mixed sample	RCT (7)	64 62 61–67	177 196 70–250	63 40 26–100	eHealth or other educational resources (3) Self-management education programmes (4)	Australia (1), Iran (1), Sweden (2), UK (2), USA (1)
	RCT pilot (2)	60	47–72	28–40	Alternative and complementary medicine (1) Self-management education programmes (1)	USA (2)
	RCT quasi-experimental (4)	64 64 60–68	126 107 50–240	52 28 28–100	Self-management education programmes (4)	Sweden (1), USA (3)
	Prospective Cohort (8)	62 62 49–74	1267 826 31–4204	36 33 28–55	Alternative and complementary medicine (1) Clinical presentation (1) Risk factors (2) Self-management education programmes (2) Strategies/responses after cardiac events (1) Symptom experience/description (1)	Australia (1), Canada (2), USA (5)
	Retrospective cohort (3)	69 69 65–72	378 72 39–1024	60 51 30–100	Self-management education programmes (1) Strategies/responses after cardiac event (1) Risk factors (1)	Australia (1), Canada (1), Sweden (1)
	Cross-sectional (6)	60 59 58–63	604 96 10–2798	45 35 26–100	Ethnicity and symptom experience (1) Self-management education programmes (1) Strategies/responses after cardiac event (4)	Lithuania (1), UK (1), USA (3), across (1)

CABG, coronary artery bypass graft; CAD, coronary artery disease; PCI, percutaneous coronary intervention; RCT, randomised controlled trial.

There were a total of 61 891 participants (29 552 (48%) women) included across studies, and the mean age was 63 years (range 47–80 years). The median sample size was 90 (range 7–7093), the mean proportion of women was 56% and 58 (26%) of studies included only women (n=15 071). The relationship between the per cent women, study design, type of cardiac pain and year of publication are depicted in figure 2. Figure 3 illustrates the relation between mean age and sample size across years of publication and pain categories.

Figure 2

Percent women, study design and type of cardiac pain across year of publication, weighted by sample size. RCT, randomised controlled trial.

Figure 3

Age and type of cardiac pain across year of publication, weighted by sample size. CABG, coronary artery bypass graft; CAD, coronary artery disease; PCI, percutaneous coronary intervention.

Characteristics of the intervention studies

Forty-nine (22%) of the studies on cardiac pain focused on interventions, including 33 (67%) RCTs with parallel group design, 7 (14%) RCT pilots and 9 (18%) quasi-experimental RCTs. Twenty-six (53%) of the RCTs assessed the effectiveness of interventions targeting men and women with obstructive CAD. Three RCTs (6%) included women with non-obstructive CAD. Postprocedural pain was the main or secondary outcome of 7 (14% (cardiac surgery: n=4, PCI: n=2, mixed: n=1)) RCTs, and 13 (27%) RCTs had a mixed sample of both obstructive and post-PCI/cardiac surgery pain. Interventions investigating cardiac pain related to obstructive CAD and postprocedure (PCI/CABG) included 25%–100% of their sample as women. In comparison, interventions targeting cardiac pain due to non-obstructive CAD only included women (ie, no men). The RCTs evaluated outcomes of different self-management interventions that were broad and targeted cardiac pain and symptom management through complementary and alternative medicine (CAM) interventions or self-management support delivered in groups, over the Internet, face to face or with help of other educational resources (eg, information sheets, videos). The median duration of the interventions was 2 months (range <1 week to 24 months), and the median follow-up time was 4 months (range <1 week to 9 years). The interventions and outcomes according to pain category and year of publication are outlined in table 3.

Table 3

Interventions and outcomes in RCTs across years of publication (n=49)

Pain category design	Alternative and Complementary medicine Intervention → outcome (year)	EHealth or other educational resources Intervention → outcome (year)	Self-management education programmes Intervention → outcome (year)
Obstructive CAD
RCT parallel group	Chiropractic treatment → muscle/skeletal pain, HRQL (2012) Topical heat therapy → acute chest pain (2014)	Cognitive interactive nursing intervention → risk factors, health behaviours, self-efficacy (1999) Discharge intervention, a self-directed slide presentation → knowledge about disease management (2009) Educational video and information sheets → prehospital delays Information sheets for patients with chest pain → anxiety, depression, HRQL, patients’ satisfaction (2002) Self-management intervention: sessions, home visit and telephone follow-up → readmission, HRQL, risk factors, health behaviours, depression, illness perception, self-efficacy, symptoms and well-being (2011) Supportive education programme → HRQL, symptom self-management (2002)	A cognitive–behavioural nurse-facilitated self-help intervention → angina, mood, knowledge, risk factors, health behaviours, HRQL and health service utilisation (2011) Disease management programme → physical functioning, symptoms and psychosocial status (2000) Lay-facilitated angina management programme → chest pain, depression, risk factors and health behaviours (2012) Health education in general practice (face to face) → risk factors, health behaviours, chest pain and impact on everyday life Nurse-led secondary prevention clinics → chest pain, depression and HRQL (1994) Follow-up care intervention in general practice → risk factors, health behaviours, chest pain and HRQL (1990) Self-care intervention→ HRQL, social support, risk factors and health behaviours (2010) Stress management training or exercise training → psychosocial functioning, depression, anxiety, HRQL, risk factors and health behaviours (2005) Home-based psychosocial intervention programme → psychosocial distress (2002) Chinese cardiac rehabilitation programme → risk factors, health behaviours and chest pain (2004) Stress management programme postdischarge→ fatigue, depression, risk factors, health behaviours and HRQL (2002)
RCT-pilot		Motivational interviewing → self-care adherence, self-efficacy and symptom burden (2015) Smartphone iOS application → discharge instruction compliance, symptoms changes (2014) Symptom diary intervention → cardiac events, HRQL and self-care maintenance (2012)	Coaching intervention in a cardiac rehabilitation programme → attendance and rehabilitation intake, self-efficacy (2012)
RCT quasi-experimental		Individualised nurse coaching intervention → health beliefs, self-care behaviours and self-efficacy (2003) Mobile phone and Internet intervention → weight loss, risk factors (2012)	Aboriginal cardiac rehabilitation programme → risk factors, health behaviours, disease and symptom management (2013)
Non-obstructive CAD
RCT parallel group	Autogenic training → anxiety, depression and HRQL (2009)		Physical training and relaxation therapy → exercise capacity, sense of coherence, stress and HRQL (2002)
RCT quasi-experimental	Transcendental meditation → chest symptoms, HRQL and electrocardiographic changes (2012)
Post-PCI/CABG
RCT parallel group	Acupuncture → lung function and postoperative pain (2013) Guided imagery → pain and anxiety (2015)	Educational audiotape: symptom management postdischarge → symptoms, depression and HRQL (2010) Symptom management telehealth intervention → cardiac symptoms, HRQL, physical activity (2007)
RCT pilot		Lifestyle intervention, work-book and sessions (diet, smoking cessation and exercise) → cardiac events, depression, risk factors and health behaviours (2016)	Behavioural health educational programme → stress, HRQL, sense of coherence, anxiety and depression (2012)
RCT quasi-experimental	Aromatherapy → vital signs, anxiety and sleep quality (2013)
Mixed sample
RCT parallel group		Telephone counselling intervention postdischarge → psychosocial adjustment, depression, anxiety, perceived control, cardiac events (2003) Web-based cardiac rehabilitation programme → angina symptoms, risk factors, depression, and self-efficacy (2014) Work-book and audio tape with relaxation programme → angina symptoms, depression (2002)	Cardiac rehabilitation family-centred empowerment model → perceived threat, self-efficacy, self-esteem, HRQL, stress, anxiety, risk factors and health behaviours (2016) Lifestyle modification programme → exercise capacity, chest pain, depression, risk factors and health behaviours (2006) Stress management programme postdischarge → stress behaviour, social support and depression (2007)
RCT pilot	Spiritual retreat → depression, psychological symptoms, stress, hope, gratitude, HRQL, risk factors and health behaviours (2011)	Illness representation intervention → illness perception, depression, social support and HRQL (2007)
RCT quasi-experimental			Cardiac rehabilitation and education programme → self-efficacy and activity (1997) Cardiac rehabilitation and education programme → HRQL, risk factors, health behaviours and readmission (2005) Cardiac rehabilitation and education programme → HRQL, risk factors and health behaviours (1997) Cardiac rehabilitation and education programme → risk factors, health behaviours, recurrent cardiac events (2000)

CABG, coronary artery bypass graft; CAD, coronary artery disease; HRQL, health-related quality of life; PCI, percutaneous coronary intervention; RCT, randomised controlled trial.

Patient-reported outcomes across the RCTs included cardiac pain frequency and/or intensity, bodily pain, HRQL or other psychosocial factors associated with cardiac pain including fatigue, stress, depressive symptoms, anxiety, catastrophising, coping, self-efficacy, sense of coherence and personality traits. A majority of the studies also included objective outcomes related to cardiac risk factors (eg, blood pressure, weight, cholesterol levels, blood glucose and ischaemic stress tests) and patient self-reports about health behaviours (eg, smoking, activity and dietary habits). The most common data collection methods used in approximately 90% of the intervention studies were validated self-reported questionnaires collected on site by the research team or sent out by mail. Outcomes were also extracted from medial charts or registries. Ten per cent of the studies collected data through telephone surveys or face-to-face interviews using standardised.

Characteristics of non-intervention studies

In total, we identified 171 non-intervention studies: 49 (29%) prospective, 41 (24%) retrospective and 78 (46%) cross-sectional. In addition, three (2%) studies had a case–control design (table 2). The majority (n=125, 73%) of the non-intervention studies focused on women with obstructive CAD, 63 (50%) of these were cross-sectional studies. Two (1%) cross-sectional studies investigated cardiac pain and symptoms in women with non-obstructive CAD. Similar to the intervention studies, non-interventions studies investigating cardiac pain related to obstructive CAD and postprocedure (PCI/CABG) included 25%–100% of their sample as women, and all non-interventions studies in non-obstructive CAD included women only. Outcomes assessed were similar to those outcomes described for the intervention studies. Validated self-reported questionnaires were used in approximately 75% of the quantitative non-intervention studies, and 25% used structural interviews delivered face to face or by telephone. In addition, 52 (30%) of the non-intervention studies used qualitative interviews (ie, semistructured or in-depth interview), and 11 (7%) used focus group interviews to collect data. A combination of quantitative and qualitative methods was found in 11 (7%) studies. The majority of the qualitative and mixed method studies were related to obstructive CAD (n=54, 73%) that explored women’s cardiac pain and symptom experiences. Cardiac pain appraisal and pain self-management strategies were also investigated, primarily with a focus on prehospital delays, barriers/challenges (eg, knowledge deficits) or sequelae after the cardiac event (eg, anxiety, depression, fatigue) and impact on everyday life. Self-management strategies primarily included education and eHealth interventions and strategies targeted to CAM. No qualitative study was found to focus on women with non-obstructive CAD. Five studies explored women’s experiences after cardiac surgery, and two studies explored women’s experiences with cardiac rehabilitation in mixed samples of women after cardiac events (eg, MI and cardiac surgery).

Seventeen (10%) studies explored association between ethnicity and obstructive CAD descriptions and recognition in women, and most (n=10, 58%) were cross-sectional studies. In addition, among studies conducted in the USA, the sample often included minority ethnic groups (eg, African-American, Hispanic), but outcomes were rarely reported for these groups separately.

Discussion

The studies included in our evidence map provide a very comprehensive broad overview of the evidence on the self-management of cardiac pain in women published between 1 January 1990 and 24 June 2016. Main results confirm those of the Committee on Women’s Health Research41 suggesting that there has been some progress in the increased representation of women in cardiovascular health research and increased knowledge about atherosclerotic disease in women, such as sex differences in the pathophysiology of CAD, symptom presentation and clinical outcomes. However, our results are consistent with more recent evidence42 43 that suggests a general lack of research incorporating: (1) sex and gender, (2) women younger than 50 years and older than 75 years and (3) self-management interventions.

Sex and gender terms are not interchangeable.44 Sex refers to the anatomy of an individual’s reproductive system, and gender refers to socially constructed characteristics of women and men based on behaviours, expressions and identities.45 We identified 57 (26%) studies that included only women (1990 to September 2016) and 84 (38%) studies that had fewer than 40% women in the total sample. Under-representation of women (<25%) was one of the main reasons for exclusion of studies in our second screening, and this was to ensure that we had a representative sample of studies reflecting the ratio of men and women with cardiovascular disease. Thus, despite the reported progress in the increased representation of women in cardiovascular health research, our results suggest that this gap still exists and is supported by more recent evidence.46 In addition to sex, there is a need to identify and examine gender differences in self-management interventions for cardiac pain. Pilote and Norris reported that gender was associated with higher rates of recurrent acute coronary syndrome and MACE.47 Other psychosocial factors (eg, depression) have been linked to adverse outcomes after an MI48; 50% of women less than 50 years of age and 40% of women between 50 and 60 years of age have major depression.49 In addition to depression, women with premature MI are often from minority groups and have low socioeconomic status and exposure to sexual abuse during childhood.49 Researchers must consider the complex inter-relationships between the anatomical/physiological aspects, psychological processes and the person’s interaction within the sociocultural context44 when designing self-management interventions a priori rather than adjusting for them in the analyses.50

Over 25% of women will die within a year of their first MI compared with 19% men; and 47% women will die within 5 years of their first MI compared with 36% men.25 These disparities are associated with age, ethnicity and risk factor burden (eg, depression).51 These disparities lead to women delaying to seek medical care for their cardiac pain.23 In a recent integrated review including 23 studies in 17 different countries,52 older age, female gender, ethnicity, lower socioeconomic status and symptom knowledge deficits were associated with mean delay times of 3.4 hours. Only three studies in this review52 included a subanalysis by ethnicity, and these studies reported that Blacks, Asians, Hispanics and South Asians had longer delays in seeking medical care compared with Caucasians. Ethnicity and culture affect perceptions of chest pain and aspects of a woman’s life that includes caregiving, education, employment and self-management practices.43 We found 17 studies that focused on ethnicity and symptom experience; a majority of these were cross-sectional studies conducted in North America and included women with African or Hispanic origins. Indigenous people experience greater cardiovascular disease burden,53 yet only one quasi-experimental RCT targeted indigenous women.54 This indicates that we have little research-based information on which to develop effective self-management interventions for non-Caucasians across all cardiac pain categories.

The majority of studies in the evidence map included women 50–75 years of age. In view of the increasing incidence of CAD and higher death rates in younger women with obstructive CAD compared with men,55 56 this research gap is disturbing. Women’s increased risks for macrovascular or obstructive CAD are linked to higher obesity and diabetes rates,55 and diabetes also concurs an accelerated risk for microvascular or non-obstructive CAD in younger women.17Healthcare costs attributable to non-obstructive cardiac pain are increasing,19 and the lack of recognition and assessment of early symptoms in younger women may underpin a negative trajectory.43 Younger women with CAD are more likely to have caregiving responsibilities, marital/family dysfunction and poorer perceived social support compared with men and older women.55 Our results suggest that we need to include women younger than 50 years and older than 75 years in self-management research related to cardiac pain.

A recent Scientific Statement from the American Heart Association43 indicates that self-management interventions specifically tailored for women only are more effective. Self-management interventions allow people to take an active part in the management of their own conditions27 and are important predictors of successful behaviour change.57 In addition to reducing pain, self-management interventions improve HRQL.58–63 Self-management is one of the six components of the chronic care model (CCM),64–66 and personal efforts to engage in self-management cannot be made without consideration to other CCM components, such as the community and provider decision supports. We recommend using a strong theoretical model to guide the design and evaluation of self-management interventions focused on the sex and gendered aspects of cardiac pain in women. For example, the individual and family self-management theory targets context (condition-specific factors, physical and social environments), process (goal setting, self-monitoring, decision making, planning and engaging in specific behaviours, collaboration) and outcomes (health status, HRQL, cost)67 68 and could be used to design tailored interventions for women with cardiac pain.

Non-pharmacological interventions women used to self-manage obstructive and postprocedural cardiac pain included physical exercise and relaxation, cognitive–behavioural therapy, music, aromatherapy, acupuncture and peer support. Many of these interventions were often incorporated as part of rehabilitative treatment approaches. Psychological treatments were generally separated into theoretically based approaches (eg, acceptance-based and mindfulness therapy) and more specific techniques (eg, cognitive–behavioural therapy, relaxation and hypnosis).69 Non-pharmacological persistent pain interventions were in general about reducing pain intensity and pain interference and the disability associated with symptoms.69 In addition to cardiac pain frequency, outcomes emphasised HRQL and other biopsychosocial factors associated with cardiac pain. This is a promising direction for future research, particularly for women who experience cardiac pain without evidence of coronary artery obstruction or continue to have cardiac pain despite coronary interventions. Creating mechanisms to assist women to recognise and manage cardiac pain and symptoms and guide them to seek appropriate assessment and evaluation is absolutely necessary in future research studies.20 70 Although women are more likely to have cardiac pain in the absence of obstructive CAD, the evidence is inconclusive.71 For these women, elimination of physical causes of pain is not always possible. We found only five studies addressing cardiac pain in women with non-obstructive CAD; no qualitative study was found, suggesting little in-depth knowledge exists about how women describe, manage and make decisions about non-obstructive cardiac pain. Addressing this knowledge gap is essential before developing self-management interventions to meet the unique needs of these women. Cardiac pain is complex, and the need for holistic treatment approaches has been emphasised.72 In total, 19 (9%) studies across cardiac pain categories and study designs focused on complementary approaches (ie, CAM) for cardiac pain. These interventions aligned with the complementary integrative medicine interventions for persistent pain summarised in a systematic review by Delgado et al.73 Only eight of the studies in our evidence map, including 2% (n=554) of the total trial participants in the 49 intervention studies, assessed the effect of CAM interventions in RCTs. This indicates that we are lacking strong evidence to routinely include CAM interventions in cardiac pain treatment strategies for women.

Limitations

This evidence map is based on a comprehensive and systematic search of 20 databases including grey literature sources. Despite the high number of potentially eligible studies, many were either not relevant to cardiac pain or the number of women included in the studies was too low. Including only studies published in English or Scandinavian language may be seen as a limitation. However, we chose to include a robust search of the grey literature in an attempt to reduce this potential publication bias.74 The articles in the second round of full-text screening were also single-screened, which may be seen as a limitation. However, if there was uncertainty about an article, a second reviewer was consulted to confirm inclusion/exclusion.

Despite having healthcare providers and researchers involved in establishing the scope of the evidence map, our search strategy may not have targeted what women themselves see as the most important aspects of the self-management of cardiac pain. Cardiac pain is a subjective experience and is associated with physical symptoms (eg, breathlessness, fatigue, sleep problems) and emotional sequelae (eg, anxiety, depression).75 Additional search terms and keywords related to cardiac pain and symptoms may have increased our probability of identifying self-management interventions. The next step in the EPPI review process is to present results of this broad mapping and screening exercise to women with cardiac pain (obstructive CAD, non-obstructive CAD and postprocedure (PCI and cardiac surgery)), to reconfirm our search terms for a more targeted in-depth search and review including quality appraisal of intervention and non-intervention studies to enhance the self-management of cardiac pain in women.

Conclusions

We aimed to describe the current evidence related to the self-management of cardiac pain in women using the process and methodology of evidence mapping. It was not to present individual study results or synthesis of results, but to describe the process and methodology for creating an evidence map database, using the topic of self-management of cardiac pain in women. Results from this mapping process suggest that evidence about risk factors, clinical presentations and symptom experiences in women is emerging, particularly for middle-aged women with obstructive CAD. While much is known about the differing presentations of cardiac pain in women, not enough is known about young and old women, women with non-obstructive CAD, and little has been done to help women manage their pain. Self-management intervention trials are lacking across all cardiac pain categories (obstructive CAD, non-obstructive CAD and postprocedure (PCI and cardiac surgery)). Further research to develop mechanisms to assist women to recognise and manage cardiac pain, using a strong theoretical model to address contextual and process components of self-management, is needed.