Abstract
Poor self-care is prevalent in patients with heart failure and adversely affects heart failure symptoms, hospitalization, and mortality. Adherence rates to different types of self-care vary in patients with heart failure. The purpose of this study was to examine factors associated with medication adherence, dietary sodium adherence, and symptom management in patients with heart failure. Data were collected using questionnaires, Medication Electronic Monitoring System, and 24-hour urine sodium output (N = 94). In logistic regression analysis, social support, education level, and race were associated with medication adherence (p = .002). Gender, race, perceived control, and depressive symptoms were associated with dietary sodium adherence (p = .030). Gender, education level, and depressive symptoms were associated with symptom management (p = .006). Different factors were associated with each type of self-care. Thus, clinicians need to manage each type of self-care by considering factors associated with it.
The most common types of self-care in patients with heart failure are taking medication as prescribed (medication adherence), following a low sodium diet (dietary sodium adherence), and monitoring and managing signs and symptoms (symptom management; Lennie et al., 2008; Riegel et al., 2004; Wu et al., 2013). Typically, adherence to the three types of self-care in patients with heart failure is poor, although medication adherence and dietary adherence rates differ depending on the subjective and objective measures used. Medication adherence assessed by an objective measure, such as a Medication Event Monitoring System, was poor, ranging from 28% to 33%, while the adherence rates based on self-report ranged from 57% to 75% (Tang et al., 2014). Dietary sodium adherence assessed by an objective measure, such as 24-hour urine sodium intake, was also poor (25%), while self-reported adherence was 75% (Lennie et al., 2008). Both studies imply that self-reported medication adherence and dietary sodium adherence may not be reliable. Symptom management is also inadequate, ranging from 58 out of 100 points (scores for adequate symptom management > 70; Tawalbeh et al., 2017).
Poor self-care has been known to adversely affect heart failure symptoms and/or hospitalization and mortality rates. Poor medication adherence is associated with higher hospitalization and mortality rates (Wu et al., 2013). Too low or too high sodium intake is associated with more severe heart failure symptoms or higher rates of hospitalization and mortality (Doukky et al., 2016; Heo et al., 2014). It has been shown that adequate symptom management using a self-care log reduced length of stay by 58% (Eastwood et al., 2007). Thus, self-care needs to be improved to reduce heart failure symptoms, and hospitalization and mortality rates.
To improve self-care, it is critical to determine modifiable factors affecting it. Psychological, social, and physical factors, such as depressive symptoms, perceived control, social support, and functional status, may affect self-care, including medication adherence, dietary sodium adherence, and symptom management. More severe depressive symptoms are associated with poorer medication adherence assessed by a questionnaire (Ragbaoui et al., 2017). The relationship between perceived control and self-care has not been frequently examined in patients with heart failure or other populations. In patients with diabetes, higher level of perceived control was significantly associated with higher medication adherence assessed by a questionnaire in a bivariate analysis (Gonzalez et al., 2015). Social factors, such as social support, may affect medication adherence. In patients with heart failure, social support is significantly associated with medication adherence assessed by a questionnaire (Ragbaoui et al., 2017). Moreover, one physical factor, New York Heart Association functional classification, is significantly associated with self-care maintenance (Yang & Kang, 2018), including management of medication.
Depressive symptoms, perceived control, and social support may affect dietary sodium adherence. More severe depressive symptoms are associated with poorer dietary adherence, including sodium intake assessed by a questionnaire in patients with heart failure (Luyster et al., 2009). The relationships between perceived control or social support and dietary sodium adherence have not been frequently examined in patients with heart failure. Perceived control was significantly associated with overall self-care, including dietary adherence, assessed by a questionnaire in patients with heart failure (Hwang et al., 2014). Patients with heart failure reported that social support, such as social pressure and support from others and social situations, affected their dietary adherence (Heo et al., 2009). In addition, one physical factor, New York Heart Association functional classification, is significantly associated with self-care maintenance, including management of dietary sodium (Yang & Kang, 2018).
Psychological and physical factors can be also associated with symptom management. In patients with heart failure (Hwang et al., 2014), depressive symptoms and perceived control were significantly associated with self-care, including management. One physical factor, New York Heart Association functional classification, is also significantly associated with self-care, including symptom management (Yang & Kang, 2018).
However, the relationships of all these modifiable factors to self-care have not been examined in patients with heart failure, especially using objective measures. In a study (Albert et al., 2014), medication adherence rate (the highest) differed from adherence to daily weight monitoring (the lowest). In another study (Dunbar et al., 2013), family partnership and education improved dietary sodium adherence, but not medication adherence. The findings of both studies imply that factors affecting adherence to each type of self-care may differ. However, factors affecting important types of self-care in patients with heart failure, including taking medication as prescribed, following a low sodium diet, and symptom management, have been rarely examined in a single study. Thus, we do not know whether different factors affect each of the three types of self-care or not, and, in turn, we also do not know whether we have to use different strategies or the same strategies to improve each type of self-care in this population.
The purpose of this study was to determine modifiable factors (i.e., New York Heart Association functional classification, perceived control, social support, and depressive symptoms) associated with medication adherence and dietary sodium adherence assessed by objective measures, and symptom management, controlling for non-modifiable covariates. Covariates were selected based on literature. Age (Juarez et al., 2015), gender (Chung et al., 2006), marital status (Ni et al., 1999), educational level (Juarez et al., 2015), race (Juarez et al., 2015), and comorbidity (Masterson Creber et al., 2014) were associated with adherence to self-care.
Methods
Sample
The research team recruited research participants in the longitudinal study from heart failure clinics or units of four hospitals in three cities in the United States (2008–2012). The inclusion criteria were a diagnosis of heart failure based on medical records, New York Hear Association functional classifications II, III, or IV based on an in-depth, face-to-face interview, the ability to read and speak English based on medical record or a phone call or a face-to-face interview, no cognitive problems (Mini-Mental State Examination test score < 24; Riegel et al., 2002), no history of severe psychiatric diseases (e.g., schizophrenia) with the exception of depression, no terminal illness (e.g., cancer), and no referrals for heart transplantation or interventions.
The research team and the health care providers at the hospitals reviewed medical records to screen eligible participants. Then, the health care providers recruited the eligible patients. One research team member contacted the patients who were referred by the health care providers or who were self-referred by phone or at the clinics after checking their eligibility through medical record reviews. Considering an effect size of 0.28 (Luyster et al., 2009), a 5% significance level, 90% power, and 10 independent variables, the sample size of the current study was 83 (G * Power 3.1.5; Faul et al., 2007). Among 123 cases, 29 were excluded because of withdraw, ineligibility, or missing data.
Measurements
Medication adherence was assessed using one objective measure (Medication Event Monitoring System). In this study, medication adherence was defined as the percentage of days that each participant took the correct number of tablets for three consecutive months (adherence: ≥ 80%) (Heo et al., 2014). The medication system has two parts: A vial for medication and a cap, which is a microelectronic monitoring device. The device records the time when the cap is opened and removed from the vial (Heo et al., 2014). A research team member discussed with the patient to select one cardiac medication (e.g., beta blocker or angiotensin-converting enzyme inhibitor/angiotensin II receptor blocker). The research team member asked the participant to put the medication into the vial and use the system for three consecutive months. The patient refilled the medication if the medication in the vial was all gone or almost gone.
Sodium intake was assessed using 24-hour urine and defined as the amount of sodium excretion in 24-hour urine per day (≤ 2,500 mg). Urine sodium is an objective and reliable measure of sodium intake in comparison to self-report (Lennie et al., 2008). Participants received written and verbal instructions regarding the purpose and methods of 24-hour urine collection from the research team member. Participants collected urine using a container or containers for consecutive 24 hours and recorded the time and volume of each urination on a urine-collection sheet. The research team picked up the urine container(s) from the participant’s home to deliver it or them to a certified clinical lab to assess the amount of sodium in urine.
Symptom management was assessed by the self-care management subscale (six items) of the Self-Care of Heart Failure Index, which assesses individuals’ involvement in monitoring and managing heart failure symptoms (Riegel et al., 2004). The total possible standardized scores range from 16.7 to 100, and higher scores indicate better symptom management (adherence ≥ 70). Reliability (Cronback’s alpha = .70) and validity (a significant difference in the scores between new patients with heart failure and experienced patients with heart failure, p < .05) in this population were acceptable (Riegel et al., 2004), and Cronbach’s alpha in the current study was .62.
Psychological variables were assessed by questionnaires. Depressive symptoms were assessed by the Patient Health Questionnaire–9 (Hammash et al., 2013). Total possible scores range from 0 to 27, and higher scores indicate more severe depressive symptoms. Reliability (Cronback’s alpha = .83) and validity (significant relationship between depresisve symptoms assessed by this instrument and the Beck Depression Inventory II, Kappa = .64, p < .001) in this population were acceptable (Hammash et al., 2013), and reliability was also acceptable in the current study (Cronbach’s alpha: .89). Perceived control was assessed by the Control Attitudes Scale-Revised (eight items; Moser et al., 2009), which assesses the levels of control of heart failure and life. Total possible scores range from 8 to 40, and higher scores indicate higher levels of perceived control. Reliability (Cronback’s alpha = .76) and validity (significant relationships of perceivedd control to anxiety and depressive symptoms, p < .05) in this population were acceptable (Moser et al., 2009), and reliability was also acceptable in the current study (Cronbach’s alpha: .79). Social support was assessed by the Multidimensional Scale of Perceived Social Support (12 items), which assesses people’s perceptions of the degree of support received from others (Zimet et al., 1990). Total possible scores range from 12 to 84, and higher scores indicate higher level of support. Reliability (Cronbach’s alpha = .93) and validity (signficant relationship between social support and symptom management, p < .01) in this population were acceptable (Shumaker et al., 2017). Cronback’a alpha in the current study was .94.
Sociodemographic and clinical characteristics were assessed using standardized sociodemographic and clinical questionnaires. Comorbidity was assessed using the Carlson Comorbidity Index (Charlson et al., 2014).
Procedures
This was a cross-sectional, correlational study that examined factors associated with three types of self-care, including medication adherence, dietary sodium adherence, and symptom management, in patients with heart failure using baseline data of a longitudinal, correlational study (Heo et al., 2014). The purpose of the longitudinal study was to examine modifiable factors, including physical factors, psychosocial factors, and behavioral factors, associated with health-related quality of life in patients with heart failure (Heo et al., 2014). The longitudinal study was approved by relevant institutional review boards. The research team obtained written informed consent from all the research subjects. The investigation was conducted according to the principles outlined in the Declaration of Helsinki. In the longitudinal study, data were collected twice at baseline and at 12 months.
Analysis
Descriptive statistics, t tests, and chi-square tests were used to compare and describe the characteristics of participants based on self-care adherence depending on the levels of measures. Logistic regression analyses were used to examine modifiable factors associated with each of medication adherence, dietary sodium adherence, and symptom management adherence, controlling for age, gender, marital status, education level, race, and comorbidity. All data analyses were done using IBM SPSS Statistics for Windows (Version 24.0 Armonk, NY). Two-tailed tests with a significance level p < .05 were used for all the tests.
Results
The sample characteristics based on self-care adherence are presented in Table 1 and Table 2. When the sample (N = 94) was divided into two groups of adherents and non-adherents for each type of self-care, the mean age ranged from 53.6 to 60.3 years, and the mean years of education ranged from 13.7 to 15.4 years (Table 1). The comorbidity scores ranged from 3.0 to 3.6. The mean scores of perceived control ranged from 29.2 to 30.4. The mean scores of social support ranged from 62.3 to 72.0. The mean scores of depressive symptoms ranged from 6.4 to 10.2. Approximately 25% to 52% of each group were males (Table 2). Approximately 30% to 56% were married. Approximately 39% to 71% were Caucasian. Approximately 52% to 70% were at New York Heart Association classification III or IV.
Sample Characteristics Based on Self-Care Adherence (N = 94): Continuous Variables.
Sample Characteristics Based on Self-Care Adherence (N = 94): Categorical Variables.
Medication adherents (n = 48) compared with non-adherents (n = 46) perceived a higher level of social support (p = .004) and a lower level of depressive symptoms (p = .006). In addition, more medication adherents kept marital status (p = .012) and were Caucasian than other races (p = .002). More dietary sodium adherents (n = 30) compared with non-adherents (n = 64) were females (p = .023). There were no differences in the sample characteristics between symptom management adherents (n = 20) and non-adherents (n = 74). In this sample, medication adherence (51%), dietary sodium adherence (32%), and symptom management adherence (21%) were poor. The mean standardized score of symptom management was 51.2 out of 100 (± 20.8).
In logistic regression analysis, education level (p = .030), race (p = .011), and social support (p = .011) were significantly associated with medication adherence (χ2 = 27.277, p = .002, Nagelkerke R2 = .336) (Table 3). Greater social support was associated with a greater likelihood of medication adherence (odds ratio = 1.050), while higher education level (odds ratio = .817) and races other than Caucasian (odds ratio = .248) were associated with less likelihood of medication adherence. Gender (p = .013), race (p = .033), perceived control (p = .041), and depressive symptoms (p = .016) were significantly associated with dietary sodium adherence (χ2 = 19.891, p = .030, Nagelkerke R2 = .267) (Table 4). Female gender (odds ratio = 4.060), greater perceived control (odds ratio = 1.152), and depressive symptoms (odds ratio = 5.420) were associated with a greater likelihood of dietary sodium adherence, while races other than Caucasian (odds ratio = .262) were associated with a less likelihood of dietary sodium adherence. Gender (p = .031), education level (p = .039), and depressive symptoms (p = .041) were significantly associated with symptom management adherence (χ2 = 24.625, p = .006, Nagelkerke R2 = .357) (Table 5). Female gender (odds ratio = 4.413), higher education level (odds ratio = 1.284), and depressive symptoms (odds ratio = 1.125) were associated with a greater likelihood of performing adequate symptom management.
Factors Associated with Medication Adherence.
Note: χ2 = 27.277, p = .002, Nagelkerke R2 = .336.
Depressive symptoms (reference group): no depressive symptoms. Gender (reference group): male. Marital status (reference group): married status. New York Heart Association functional classification (reference group): II. Race (reference group): Caucasian.
Factors Associated with Dietary Sodium Adherence.
Notes: χ2 = 19.891, p = .030, Nagelkerke R2 = .267.
Depressive symptoms (reference group): no depressive symptoms. Gender (reference group): male. Marital status (reference group): married status. New York Heart Association functional classification (reference group): II. Race (reference group): Caucasian.
Factors Associated with Symptom management Adherence.
Notes: χ2 = 24.625, p = .006, Nagelkerke R2 = .357.
Depressive symptoms (reference group): no depressive symptoms. Gender (reference group): male. Marital status (reference group): married status. New York Heart Association functional classification (reference group): II. Race (reference group): Caucasian.
Discussion
The findings of the current study demonstrate that different modifiable and non-modifiable factors are significantly associated with each type of self-care in patients with heart failure. Modifiable factors of social support, perceived control, and depressive symptoms were significantly associated with at least one type of self-care. Higher level of social support was associated with medication adherence, while higher level of perceived control was associated with dietary sodium adherence. Unexpectedly, more severe depressive symptoms were associated with dietary sodium adherence and symptom management adherence. Among non-modifiable factors, education level, race, and gender were associated with self-care. Education level was associated with both medication adherence and symptom management, but the directions were opposite. Races other than Caucasian were associated with both poor medication and dietary sodium adherence. Female gender was associated with better adherence to dietary sodium intake and symptom management. All of these findings imply that self-care can be enhanced by improving those modifiable factors and that the relationships of those modifiable and non-modifiable factors to each type of self-care are complicated. Thus, clinicians and researchers need to assess each type of self-care separately and provide different strategies to improve each type of self-care in patients with heart failure.
The adherence rates of the three types of self-care differed, but all were poor. The medication adherence rate (51%) assessed by an objective measure in the current study was higher than to that (28% to 33%) in a prior heart failure study (Tang et al., 2014), but it was similar to that in another heart failure study (59%) that used the same measure (Medication Event Monitoring System; Wu et al., 2013). All of these findings demonstrate poor medication adherence, even though the medication adherence rate is relatively high in comparison to dietary sodium adherence and symptom management. In the current study, adherence to dietary sodium intake assessed by an objective measure (24-hour urine) was very poor (32%) and similar to that in another heart failure study (25%; Lennie et al., 2008). Adherence to symptom management (21%) was the poorest among the three types of self-care, and the standardized score (51.2) was similar to that in other studies (58; adequate symptom management ≥ 70; Tawalbeh et al., 2017). All of these findings demonstrate poor adherence in all three types of self-care and the need for improvements.
The findings of the current study demonstrate different modifiable and non-modifiable targets of interventions to improve each type of self-care. The findings imply that improvement in social support and consideration of education level and race can be important targets of interventions to improve medication adherence assessed by an objective measure. In a prior heart failure study (Ragbaoui et al., 2017), social support was also significantly associated with medication adherence assessed by a questionnaire. The findings of both studies demonstrate the important role of social support in medication adherence regardless of the measures. Social support may help patients with heart failure remember to take their medication as prescribed because forgetfulness is one of the most common factors affecting non-adherence to taking medication as prescribed (Aggarwal et al., 2015). Thus, social support needs to be enhanced to improve medication adherence. Among the non-modifiable factors, higher level of education and races other than Caucasian were associated with less likelihood of medication adherence. In a heart failure study (Juarez et al., 2015), bivariate analysis showed that higher level of education (categorical variable) was associated with medication adherence assessed by average medication possession ratios. In the current study and also in a prior study (Juarez et al., 2015), races other than Caucasian were associated with less likelihood of medication adherence. Thus, interventions need to be provided to patients with heart failure regardless levels of education and who are races other than Caucasian, to improve medication adherence.
Improvement in perceived control and consideration of depressive symptoms, gender, and race can be important targets of interventions to improve dietary sodium adherence. Among the modifiable factors, higher levels of perceived control and depressive symptoms were associated with a greater likelihood of dietary sodium adherence in the current study. The relationship between perceived control and dietary sodium adherence assessed by an objective measure has not been frequently examined in heart failure. In a heart failure study (Hwang et al., 2014), higher level of perceived control was associated with overall good self-care. These findings demonstrate that patients’ perceptions of how much control they have over their symptoms and life can have an important role in dietary adherence and also overall self-care. Unexpectedly, a higher level of depressive symptoms was significantly associated with a greater likelihood of dietary sodium adherence. In prior studies (Luyster et al., 2009; Ragbaoui et al., 2017), more severe depressive symptoms were associated with poorer medication adherence or dietary sodium adherence. These inconsistent findings may be due to differences in the measures. In the current study, both medication adherence and dietary sodium adherence were assessed by objective measures, such as Medication Event Monitoring System and 24-hour urine sodium intake, while in the prior studies, they were assessed by subjective measures. The medication and dietary sodium adherence rates assessed by objective and subjective measures considerably differed (Lennie et al., 2008; Tang et al., 2014). The effects of depressive symptoms on appetite may, in part, explain the association between more depressive symptoms and dietary sodium adherence in the current study. In a heart failure study (Andreae et al., 2018), depressive symptoms interacted with appetite. In another heart failure study (Song et al., 2015), more patients with depressive symptoms compared with non-depressed patients had micronutrient deficiency. These findings may imply that depressive symptoms may decrease appetite, which can lead to less dietary sodium intake. Thus, both perceived control and depressive symptoms need to be considered to improve dietary sodium adherence. Among the non-modifiable factors, female gender was associated with dietary sodium adherence, which is consistent with the finding in the prior study (Chung et al., 2006). This may be because females, typically, prepare foods for themselves so that it is easier for them to follow a low sodium diet. Or females may take less foods than males, which can lead to less intake of dietary sodium. Races other than Caucasian were associated with less likelihood of dietary sodium adherence. Thus, more attention needs to be paid to male patients and those from races other than Caucasian to improve dietary sodium adherence.
Consideration of depressive symptoms, gender, and educational level can be important targets of interventions to improve symptom management. Unexpectedly, depressive symptoms were associated with a greater likelihood of symptom management. In a heart failure study (Hwang et al., 2014), in the high knowledge and poor self-care group, depressive symptom score was high (9.5), while in high knowledge and good self-care group, depressive symptom score was low (5.7), which contradict the finding in the current study. However, in the prior study, the authors assessed overall self-care, but not symptom management alone. In addition, the groups were categorized based on the levels of both knowledge and self-care. On the other hand, depressed patients may be involved in more symptom management because they might perceive symptoms more severely than non-depressed patients because of abnormal body perception (Wiebking et al., 2010). Further studies are needed to examine the relationship between depressive symptoms and symptom management in this population.
Among the non-modifiable factors, female gender and higher level of education were associated with a greater likelihood of symptom management. Female gender was also associated with a greater likelihood of dietary sodium adherence, as addressed previously. One important factor in symptom management is to reduce sodium intake when symptoms occur. If patients can prepare own foods, it is easier for them to manage their sodium intake, which is part of symptom management. Higher level of education was associated with a greater likelihood of symptom management. In the relationship between education and medication adherence, the direction of the relationship was opposite. To take medication as prescribed, comprehensive knowledge is not required, while to perform symptom management, patients need comprehensive knowledge and skills to monitor and recognize signs and symptoms and the changes, recognize the causes, including medication, diet, and mood, and manage the causes effectively. Thus, higher level of education might be associated with a greater likelihood of symptom management. The findings of the current study suggest that more attention needs to be paid to male patients and patients with a lower level of education to improve symptom management.
The findings of the current study demonstrate the need for interventions to improve poor self-care by improving perceived control and social support by managing depressive symptoms, in addition to the need for paying attention to male patients and patients with races other than Caucasian and lower levels of education. Some studies provided interventions to improve perceived control, social support, and depressive symptoms. However, these types of interventions have not been frequently provided to patients with heart failure. One meditation pilot study that provided weekly 12 sessions (90 minutes/session) showed promising outcomes for improving all of perceived control, social support, and depressive symptoms (Heo et al., 2018). In the current study, depressive symptoms were associated with adherence to dietary sodium intake and symptom management, and poor appetite and abnormal body perception may be the reasons (Andreae et al., 2018; Song et al., 2015). Meditation may improve both appetite and abnormal body perceptions. Even though depressive symptoms were associated with better self-care in the current study, depressive symptoms still need to be managed because of the adverse effects on heart failure symptoms and health-related quality of life (Heo et al., 2014). In addition, clinicians and researchers need to assess and manage each type of self-care using reliable and valid instruments and examine intervention effects on each type of self-care separately.
There are some study limitations. The mean age of the sample was relatively young, which limits the generalizability of the findings of the current study. The current study used a cross-sectional study design, which prevents examining the causal relationships among those variables. Thus, a longitudinal study is needed to examine the causal relationships. Symptom management was assessed using a self-report questionnaire, which might not reflect the actual symptom management of the participants. However, currently there are no objective measures for symptom management.
In conclusion, the findings of this study demonstrate that different modifiable and non-modifiable factors were associated with each type of self-care. More social support, lower level of education, and Caucasian race were associated with medication adherence. Higher level of perceived control, depressive symptoms, female gender, and Caucasian race were associated with dietary sodium adherence. Depressive symptoms, female gender, and higher education level were associated with symptom management adherence. Improvement in social support may improve medication adherence. Improvement in perceived control may improve dietary adherence. More attention needs to be paid to male patients to improve dietary adherence, while more attention needs to be paid to patients with a lower education level to improve symptom management. Thus, clinicians and researchers need to assess each type of self-care separately and provide interventions that target distinct modifiable and non-modifiable factors to improve each type of self-care.
Footnotes
Acknowledgements
The study was conducted at Indiana University and University of Arkansas for Medical Sciences.
Declaration of Conflicting Interests
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: Funding for this study came from American Heart Association, Scientific Development Grant (0830104N) to Indiana University (University of Arkansas for Medical Sciences).
