Self‐Care for the Prevention and Management of Cardiovascular Disease and Stroke: A Scientific Statement for Healthcare Professionals From the American Heart Association

Self‐care is most commonly understood as a naturalistic decision‐making process in which persons engage for the purpose of maintaining health and managing acute and chronic illness.^{9, 32, 33, 34, 35} Self‐care decision making is a complicated process. Better understanding of the nature of self‐care decision making will help clinicians understand how to better teach self‐care to their patients and to understand how self‐care fails and how to improve it. Naturalistic decision making has been used to explain the process of self‐care in individuals with CVD, most commonly HF, as well as other chronic illnesses.^{35, 36, 37, 38}

The naturalistic decision‐making framework explains that, in real‐world settings, people make decisions that are meaningful and familiar to them.³⁹ These real‐world decisions are complex; they involve uncertainty, ambiguity, dynamically evolving conditions, missing information, time stress, and high stakes. These decisions may also have ill‐defined, shifting, or competing goals and involve multiple individuals. The naturalistic decision‐making process explains how individuals make decisions given this complexity and how they develop the skills necessary to succeed when faced with similar situations.⁴⁰ Naturalistic decision making emphasizes how individuals use their experience and personal values in decision making.³⁹ Experience emerges from situational awareness,⁴¹ or perception of the situation, as well as comprehension of the significance of a specific situation. One's experience with the situation, each option, and past response create a set of patterns that include relevant actions and expected outcomes associated with each possible response.³⁶ In this way, past experience and personal values lead to the actions taken in specific situations.⁴²

Teaching and supporting self‐care should be a major activity in our healthcare system. Yet, complexities in its conceptualization and practice result in underappreciation of self‐care by clinicians and healthcare systems. As a consequence, clinicians have not emphasized self‐care, and the vast majority of people do not perform self‐care behaviors well.^{16, 70, 71}

Self‐care research and clinical efforts have been hindered by the perceptions of both patients and providers that pharmacological interventions are more effective than lifestyle change.^{72, 73} Widespread failure of clinicians to follow, or give more than token attention to, CVD prevention guidelines has resulted in little change or worsening in CVD risk factors over time in many countries and points to a compelling need for a greater emphasis on self‐care.⁷⁴ Further evidence of this failure is evident in practice‐based approaches to CVD risk reduction.^{75, 76} Some reasons for the lack of effectiveness of clinical efforts to influence CVD risk include the limited training of providers about patient education and use of effective behavior change strategies, lack of time for patient encounters, lack of support in the clinic environment for a self‐care–based approach, growth of healthcare systems focusing on care of acute events with little appreciation of the chronicity of most conditions, and better reimbursement for treatment than for prevention.⁷⁷

Our clinical and societal focus on an illness‐ or disease‐driven model of episodic care has resulted in what Goldman et al call “investing in sickness rather than health.”⁴⁹ This investment and our changing demography have resulted in people living longer with multiple chronic conditions that are not well controlled because self‐care is at the heart of control of chronic illnesses.

Self‐measurement and monitoring of BP is a major emphasis in the self‐care of hypertension. In a recent meta‐analysis, there was evidence that self‐measurement of BP was superior to usual care (24 trials) in reducing systolic (−3.1 mm Hg) and diastolic (−2.0 mm Hg) BP at 6 months.¹⁵⁷ In the meta‐analysis, additional support beyond self‐measurement (24 trials) resulted in even greater reductions in systolic (−3.4 to −8.9 mm Hg) and diastolic (−1.9 to −4.4 mm Hg) BP compared with usual care up to 12 months.¹⁵⁷ Another meta‐analysis of 11 trials of self‐measurement of BP interventions versus usual care demonstrated a significant reduction in diastolic (weighted mean difference, −2.02; 95% confidence interval [CI], −2.93 to −1.11), but not systolic, BP.¹⁵⁸ The researchers also provided evidence that self‐measurement of BP interventions compared with usual care (13 trials) improved antihypertensive medication adherence (standardized mean difference, 0.21; 95% CI, 0.08–0.34). A third meta‐analysis¹⁵⁹ provided evidence that digital interventions designed to improve hypertension self‐care were superior to usual care in reducing systolic (weighted mean difference being −3.74  mm Hg [95% CI, −2.19 to −2.58]) and diastolic BP (−2.37  mm Hg [95% CI, 0.40 to −4.35]). Hence, there is evidence across multiple trials favoring self‐care in general, and self‐measurement of BP in particular, as efficacious means of reducing BP and improving medication adherence in hypertension. Another Cochrane review on self‐monitoring for improving control of BP in patients with hypertension is currently underway.¹⁶⁰

Cardiac rehabilitation is a commonly prescribed element of care in coronary heart disease. Karmali et al¹⁶¹ conducted a systematic review of interventions to improve uptake of cardiac rehabilitation in coronary heart disease. Eight of 10 studies included in this review increased the uptake of cardiac rehabilitation, 3 of 8 studies included improved adherence to cardiac rehabilitation, and no studies found a difference in quality of life.¹⁶¹ More recently, Janssen et al¹⁶² completed a meta‐analysis of lifestyle modification programs for patients with coronary heart disease versus usual care (23 trials). In this analysis, usual care was associated with worse outcomes compared to lifestyle modification programs, including all‐cause mortality (odds ratio, 1.34 [95% CI, 1.10–1.64]), cardiac mortality (odds ratio, 1.48 [95% CI, 1.17–1.88]), and the composite of cardiac readmissions and subsequent nonfatal infarctions (odds ratio, 1.35 [95% CI, 1.17–1.55]); improvements in dietary and exercise behavior were greater for programs that incorporated goal setting, self‐monitoring, planning, and feedback techniques.¹⁶² In summary, there is evidence across multiple trials to support self‐care interventions that are focused on lifestyle modifications as a means to improve clinical outcomes in coronary heart disease. In contrast, interventions focused solely on improving adherence to cardiac rehabilitation only seem to influence uptake, but not adherence behavior or quality of life.

Li et al¹⁴¹ recently reviewed the efficacy of structured home‐based exercise programs in patients with peripheral artery disease. Across 5 trials, structured home‐based exercise programs improved walking time (668 seconds [95% CI, 5.2–128.4]) and pain‐free walking time (57.8 seconds [95% CI, 20.4–95.1]), as well as mean difference in Walking Impairment Questionnaire distance (8.7; 95% CI, 3.9–13.5) and speed scores (8.1; 95% CI, 4.5–11.6). Hence, there is evidence across multiple trials that home‐based exercise programs improve walking ability in patients with peripheral artery disease. Moreover, there is an additional Cochrane review on disease management interventions for improving self‐care in lower‐limb peripheral artery disease currently underway.¹⁶³

A recent systematic meta‐review of 13 systematic reviews representing 101 individual trials of self‐care support interventions with stroke survivors provide high‐quality evidence to support self‐care.¹⁶⁴ The researchers concluded that support for self‐care in the context of therapy rehabilitation delivered soon after a stroke resulted in short‐term (<1 year) improvement in activities of daily living and reductions in the risk of dependence and death poststroke.¹⁶⁴ More recently, Fryer et al¹⁶⁵ completed a meta‐analysis of 6 trials of stroke self‐care programs and provided evidence that such interventions improve quality of life (standardized mean difference, 0.34 [95% CI, 0.05–0.62]) and improve self‐efficacy (0.33 [95% CI, 0.04–0.61]) compared with usual care. Hence, there is evidence across multiple systematic reviews and multiple trials to support self‐care as a means of improving activities of daily living, quality of life, and self‐efficacy, as well as reducing dependency and premature death in stroke.

Heneghan et al¹⁶⁶ completed a meta‐analysis of self‐testing and self‐care of oral anticoagulation in AF. Across 11 trials, there was a significant reduction in thromboembolic events associated with self‐monitoring versus usual care (hazard ratio [HR], 0.51 [95% CI, 0.31–0.85]), but no difference for major hemorrhagic events (HR, 0.88 [95% CI, 0.74–1.06]) or death (HR, 0.82 [95% CI, 0.62–1.09]); younger patients (aged <55 years) and those with mechanical heart valves had the greatest reductions in thromboembolic events in response to self‐testing and self‐care of oral anticoagulation.¹⁶⁶ ClarkeSmith et al¹⁶⁷ subsequently completed a meta‐analysis of 8 trials focused on the efficacy of educational or behavioral intervention versus usual care in AF on time in therapeutic range of oral anticoagulation therapy. In this meta‐analysis, self‐monitoring did not improve time in therapeutic range compared with usual care (mean difference of 6.31 [95% CI, −5.63 to 18.25]).¹⁶⁷ The researchers suggested that they may not have found benefit because 4 of the trials included mixed indication cohorts and 10 further trials were excluded because they did not provide AF‐specific data. They also suggest some patient‐specific factors that could have impacted intervention effectiveness (eg, older age and inaccurate beliefs). Although there is evidence across multiple trials that self‐testing and self‐care of oral anticoagulation reduces the risk of thromboembolic events at large, more research is needed into mechanisms other than time in therapeutic range by which self‐care interventions are successful in AF.

For more than a decade, there has been evidence from meta‐analyses of randomized, controlled trials that interventions focused on enhancing self‐care in HF are efficacious versus usual care in improving outcomes. McAlister et al¹⁶⁸ first provided evidence across 4 trials that interventions focused on enhancing HF self‐care reduced all‐cause hospitalizations and HF hospitalizations. A subsequent meta‐analysis of trials focused on HF self‐care provided evidence that interventions focused on improving self‐care (5 trials) reduced all‐cause readmission and HF readmissions (3 trials).¹⁶⁹ Grady¹⁷⁰ reviewed trials focused on self‐care interventions and quality of life in HF. Nine of the 17 trials included in her review showed an improvement in quality of life compared to usual care; but, methodological heterogeneity across the trials reviewed interfered considerably with the ability to draw any strong conclusions.¹⁷⁰ Ditewig et al¹⁷¹ completed a systematic review without a meta‐analysis on effectiveness of self‐management programs in HF and concluded that self‐care interventions generally had a positive effect on all‐cause readmission, mortality, and quality of life. These researchers also highlighted significant methodological heterogeneity across studies that interfered with any strong summative conclusions. Another meta‐analysis focused on structured telephone support interventions to improve HF self‐care.¹⁷² In this meta‐analysis, self‐care focused structured telephone support interventions reduced all‐cause mortality (22 trials; HR, 0.87 [95% CI, 0.77–0.98]) and HF hospitalizations (16 trials; HR, 0.85 [95% CI, 0.77–0.93]). Recently, Jonkman et al¹⁷³ completed an individual patient data meta‐analysis of HF self‐management interventions. Self‐management interventions (20 trials) reduced the composite risk of HF hospitalization or all‐cause death (HR, 0.80 [95% CI, 0.71–0.89]), the risk of HF hospitalization (HR, 0.80 [95% CI, 0.69–0.92]), and improved 12‐month HF‐related quality of life (standardized mean difference, 0.15 [95% CI, 0.00–0.30]).¹⁷³ In subsequent analyses, it was concluded that no specific intervention characteristics other than longer intervention duration were associated with better self‐management intervention efficacy.¹⁷⁴ In summary, there is sufficient information across many trials that interventions targeting HF self‐care are associated with better clinical outcomes and improvements in quality of life, suggesting readiness for implementation studies.

Healthy self‐care behaviors lower the risk of incident disease, and the mechanisms underlying these effects can be broadly summarized as cardioprotection.¹⁷⁵ This is particularly true of the self‐care maintenance behaviors of smoking cessation, maintaining normal BMI, routine physical activity, reducing dietary sodium intake, decreasing alcohol use, and maintaining a healthy diet, a low cholesterol, a normal BP, and a normal fasting plasma glucose. Other self‐care maintenance behaviors help reduce inflammation attributed to infection in CVD such as routine preventive dental care and annual influenza vaccination.^{176, 177} Medication adherence is perhaps the most obvious cardioprotective self‐care behavior. For example, it has been proposed that HF patients who are adherent to prescribed therapies may avoid the escalation of loop diuretics,^{178, 179, 180} avoid the need for inotropic therapy,^{181, 182} and potentially avoid having efficacious therapies discontinued during hospitalization,¹⁸³ all of which are associated with worse outcomes. When the emphasis of self‐care changes from preventing to managing CVD and secondary prevention, these health behaviors become no less important because they address common cardiovascular pathophysiological mechanisms such as inflammation.¹⁸⁴

A growing body of research points to a significant impact of social relationships on all‐cause and disease‐specific morbidity and mortality.^{250, 251, 252} Both the size and diversity of one's social network or capital (informal connections available for support, help, and information) are strongly and prospectively linked with CVD morbidity and mortality. In particular, epidemiological data suggest that people who have larger, more integrated social networks are at reduced risk for mortality and ischemic heart disease²⁵³ and stroke²⁵⁴ and have a better prognosis after myocardial infarction,²⁵⁵ compared with more socially isolated individuals.

Social networks simultaneously reflect both the individual's social capital and the community's social structures. People who are connected to a network or community rich in support, social trust, information, and healthy norms may have more access to resources that can help them achieve health goals.^{256, 257} Studies of urban versus rural communities suggest that rural residents have a higher burden of risk factors and encounter more challenges with self‐care than their urban counterparts because of inequalities related to socioeconomic resources and access to care.^{258, 259, 260, 261} However, other studies in lower‐ and upper‐middle‐income countries report a greater prevalence of CVD risk factors attributed to urbanization,^{262, 263, 264, 265} so this issue remains unclear. It is not yet sufficiently clear what factors at the community level are proportionately more important to CVD risk management. Several study protocols have been designed to identify these critical elements, but results are not yet available.^{187, 266, 267}

A consistent theme among successful studies addressing symptom response and a variety of other self‐care behaviors is the use of motivational approaches to promote self‐care. For example, a systematic review and meta‐analysis of interventions to improve medication adherence in adults with hypertension included 112 eligible treatment‐versus‐control group outcome comparisons of 34 272 subjects.²⁷⁵ The most promising intervention components were those linking adherence behavior with habits, giving adherence feedback to patients, self‐monitoring of BP, using pill boxes and other special packaging, and motivational interviewing. The most effective interventions used multiple components and were delivered over many days.

Recognition of the positive influence of family engagement on self‐care has supported the design and testing of family‐focused self‐care interventions for CVD patients and their family members. Although some clinical trials testing family‐ or dyad‐level interventions for CVD self‐care hold promise, they are limited in number, tend to have small sample sizes with short evaluation time frames, are based on varying theoretical frameworks, and have yielded conflicting results. The majority of studies have been with HF patients. Promising approaches focus on communication, decision making, reciprocity,³³⁰ caregiver self‐care, coping and communication, problem solving, and dealing with stress and negative emotions.³³¹ Factors that may interfere with a family‐focused approach include high levels of anxiety and lack of perceived control among the family members,^{332, 333} disease severity, and caregiver burden.³³¹

Multiple intervening and complex factors affect family‐level self‐care. The timing in the trajectory of care, pre‐existing and evolving comorbidities, psychosocial status at baseline, family context and relational variables, as well as health literacy and dyadic interdependence converge to influence self‐care. Family‐focused self‐care approaches needs to incorporate these factors when designing interventions for dyads or families. Furthermore, given that the physical and mental health of caregivers may be affected by the demands of caregiving, family‐focused interventions need to offer support without increasing burden to the family caregiver.

Community settings such as social or community centers provide an additional option to the local medical facility or the home in helping to build self‐care skills.³³⁴ Participants in community‐based self‐care training initiatives showed significant improvement in self‐care maintenance and self‐care management^{268, 335, 336} and in CVD risk reduction.³³⁷ There is, however, a paucity of studies on the direct links between community activities and CVD. One exception is a new and encouraging report from the World Health Organization, which found that 6 in 10 people globally are currently protected by at least 1 tobacco control measure–4 times more than in 2007. These community‐level approaches reflect policies, warnings, advertising bans, and taxes for example.³³⁸

Theoretically, community neighborhoods may encourage self‐care through activities and events that raise awareness and knowledge about healthy lifestyles, encourage collective activity and engagement, and alleviate social isolation. Examples include organized walks and runs, health fairs, farmers’ markets, and support groups in the local community. Some evidence from the United States suggests that adults with access to a community garden consume more fruit and vegetables,³³⁹ and community gardens have been used to improve the diet of poor communities, particularly in the developing world. Farmers’ markets and community gardens may promote active lifestyles and mental well‐being,³⁴⁰ but few well‐designed studies have been completed to confirm this assumption.

Support in the community may come from lay community health workers such as health trainers or coaches.¹⁸⁷ Dye et al³⁴¹ reported a successful initiative designed to improve hypertension self‐care among older rural residents through education and support offered by trained volunteers. Participants received baseline and follow‐up health risk appraisals with blood work, educational materials, and items such as BP monitors and pedometers. These participants demonstrated significant increases in hypertension‐related knowledge that persisted at 16 weeks, as well as significant improvements in stage of readiness to change behaviors and in actual behaviors. Furthermore, clinically significant decreases in all outcome measures were observed, with statistically significant changes in systolic BP, weight, and glucose.