Heart Failure Knowledge, Symptom Perception, and Symptom Management in Patients with Heart Failure

Abstract

BACKGROUND:

Prevention of heart failure (HF) exacerbations requires that patients carefully self-manage their condition. Symptom perception is a key component in self-care for patients with HF that involves monitoring for HF symptoms and recognizing symptom changes. Heart failure knowledge is a prerequisite for better symptom perception and symptom management. However, the relationships among HF knowledge, symptom perception, and symptom management remain unclear.

OBJECTIVE:

To explore the inter-relationships among HF knowledge, symptom monitoring, symptom recognition, and symptom response in patients with HF.

METHOD:

We included 185 patients with HF in this study. Heart failure knowledge was measured using the Dutch HF Knowledge Scale. Symptom monitoring, symptom recognition, and symptom response were measured using the Self-Care of HF Index. Structural Equation Modeling was used for data analyses.

RESULTS:

Heart failure knowledge was associated with symptom monitoring (β= .357, p <.001). Symptom monitoring was directly associated with both symptom recognition (β= .371, p <.001) and symptom response (β= .499, p <.001). Symptom recognition was directly associated with symptom response (β= .274, p < .001). Heart failure knowledge was not directly associated with symptom recognition, nor with symptom response. Heart failure knowledge was indirectly associated with symptom recognition and symptom response through symptom monitoring.

CONCLUSION:

Symptom monitoring is associated with both symptom recognition and symptom response and is a mediator between HF knowledge and symptom recognition and between HF knowledge and symptom response. This finding suggests that it is important for clinicians not just to increase patients’ HF knowledge but also to enhance their skills of symptom monitoring and symptom recognition and promote symptom monitoring among patients in order to improve symptom response in self-care.

INTRODUCTION

Heart failure (HF) is a chronic, progressive clinical syndrome characterized by high prevalence of multiple HF symptoms that lead to poor quality of life, high rates of hospitalization, increased medical cost, and high mortality.^{1, 2} Patients experience worsening HF symptoms with an acute HF exacerbation, which be caused by multiple factors, including excessive sodium intake, noncompliance with prescribed medication, uncontrolled hypertension, infections, or cardiac arrhythmia.³ Prevention of HF exacerbations requires that patients engage in self-management of their symptoms.⁴ Self-management of symptoms requires that patients with HF monitor and track whether they have any signs and symptoms of worsening HF on a regular basis. When HF symptoms occur, it is important that patients with HF accurately interpret symptoms, recognize these symptoms are related to HF, and take prompt and appropriate action(s).^5-7

Symptom perception is a key component of self-care for patients with HF.^{4, 8} In the original Theory of HF Self-Care,⁶ self-care is a naturalistic decision-making process (i.e., how people make decisions in real-world settings), which includes self-care maintenance and self-care management. Self-care maintenance includes behaviors to maintain physiologic stability to prevent an acute exacerbation of HF, such as symptom monitoring and treatment adherence (e.g., adherence to prescribed medications, following a low salt diet). Self-care management includes behaviors and actions to respond to HF symptoms when they occur, such as recognizing symptoms, evaluating these symptoms, deciding to take action, implementing a treatment plan, and evaluating the effectiveness of the treatment implemented.⁶ In the Revised Theory of HF Self-Care,⁷ symptom perception has been added and is comprised of symptom monitoring (which was included in self-care maintenance in the original theory) and symptom recognition (which was included in self-care management in the original theory). Symptom perception is important across the symptom maintenance and management continuum. Patients need to continuously monitor whether they have any signs and symptoms of HF exacerbation. When patients have signs/symptoms, they must recognize whether their symptoms are HF-related and if so, what action(s) they need to take to respond to these symptoms appropriately.

Heart failure knowledge is a prerequisite for better self-care.^{9, 10} Therefore, there have been a number of intervention studies focusing on enhancing patients’ HF knowledge and/or self-management/self-care skills to improve outcomes (e.g., symptoms, quality of life, and rehospitalization/death).^11-13 For example, the Family Partnership and Education trial offered educational interventions to HF patients (and their family partners) in the intervention groups to increase their HF knowledge and to improve self-care behaviors (e.g., medication adherence).¹³ Patients in the intervention group had better HF knowledge immediately after intervention than patients in the usual care control group. However, no significant differences were found between groups in medication adherence.¹³ The findings indicate HF knowledge is a prerequisite, but not sufficient for self-care behaviors.

In the Theory of HF Self-Care, HF knowledge is one of the most influential factors in developing patients’ expertise in decision-making process for performing their self-care behaviors (i.e., self-care maintenance, symptom perception, self-care management).^{6, 7} However, the inter-relationships among HF knowledge, symptom monitoring, symptom recognition, and symptom response remain unclear. Accordingly, the purpose of this study was to explore the inter-relationships among HF knowledge, symptom monitoring, symptom recognition, and symptom response in patients with HF.

METHODS

Design

This was a secondary data analysis of an observational prospective study.¹⁴ The parent study was an observational study with the aim of describing HF patients’ subjective experience in dealing with symptom scenarios. Data were collected on patients’ perceptions of symptoms, evaluations of symptoms, and responses to symptoms.¹⁴ In the current study, we used structural equation modeling to explore the relationships among HF knowledge, symptom monitoring, symptom recognition, and symptom response.

Sample and Setting

One hundred eighty-five patients hospitalized with an exacerbation of HF were recruited from one large academic medical center (945 beds) and one large community hospital (605 beds) hospital located in two urban cities in a southeastern state between 2014 and 2017. We enrolled patients during their hospitalization with a confirmed diagnosis of HF in the medical record from a heart failure cardiologist who were admitted with a HF exacerbation. Patients who had 1) a diagnosis of dementia, 2) a history of left ventricular assist device placement or heart transplant, 3) a co-existing terminal illness or hospice care, or 4) severe psychiatric disorders other than depression were excluded.

Procedures

Institutional review board approval was obtained for conducting research and recruitment in both hospitals. Patient eligibility was confirmed by research assistants. Research assistants approached eligible patients during the first 3 days of hospital admission and explained research procedures to them. Patients who agreed to participate in this study signed a written informed consent form and completed the questionnaires. After completing the study, each participant was compensated for his/her time.

Measures

Demographic and clinical variables.

Data on demographic and clinical variables were collected using a standardized demographic and clinical questionnaire by patient interview and medical record review.

Heart Failure Knowledge was measured using the Dutch Heart Failure Knowledge Scale.¹⁵ The Dutch HF Knowledge Scale contains 15 multiple-choice items regarding HF knowledge in general, knowledge on HF treatment, and HF signs/symptoms. Each item has three response options, one of which is correct. Scores are summed and transformed linearly to a 0 to 100 range, higher scores indicate more HF knowledge. This instrument is reliable and valid and has been used in studies of patients with HF.¹⁵ The Cronbach’s alpha of the scale in this study was .72.

Symptom Monitoring was measured using two items from the maintenance scale of the Self-Care of Heart Failure Index (version 6.2).¹⁶ We asked patients to rate "How often do you 'weigh yourself daily' and 'check your ankles for swelling'?” using a 4-point Likert scale ranging from 1 (never or rarely) to 4 (always). Scores are summed with higher scores indicate more frequent symptom monitoring.

Symptom Recognition was measured using one item from the management scale of the Self-Care of Heart Failure Index (version 6.2).¹⁶ When patients had symptom of trouble breathing or ankle swelling, we asked them to recall the last time they had trouble breathing or ankle swelling and to rate "How quickly did you recognize it as a symptom of HF?" using a 5-point Likert scale ranging from 0 (did not recognize it) to 4 (very quickly) with higher scores indicate more quickly recognize HF symptom.

Symptom Response was measured using four items from the management scale of the Self-Care of Heart Failure Index (version 6.2).¹⁶ If patients had trouble breathing or ankle swelling, we asked them to rate "How likely are you to try one of the four remedies?" using a 4-point Likert scale ranging from 1 (not likely) to 4 (very likely). Examples of remedies include: ‘reduce the salt in your diet’, and ‘take an extra water pill’. Higher scores indicate better self-care responses to the HF symptom(s).

Data analysis

Data analysis began with a descriptive examination of all variables using SPSS (Armonk, NY), version 28.0, including frequency distributions, means, standard deviations, medians, and interquartile ranges, as appropriate to the level of measurement of the variables. An alpha of < .05 was denoted a priori.

In this study, structural equation modeling was constructed using Amos (Analysis of Moment Structures) version 28.0 (Armonk, NY) to explore the inter-relationships among HF knowledge, symptom monitoring, symptom recognition, and symptom response in patients with HF (Figure 1). The model parameters were estimated using maximum likelihood estimation. The measurement model was a confirmatory factor analysis of indicators used to represent the single latent construct, the measure of symptom response, with multiple indicators (≥ 3) to account for measurement error. To identify latent constructs, the scale of each observed variable was set to the metric of the first indicator variable. Latent variance in the measurement model was fixed at 1.0.

Figure 1: Conceptual framework tested.

Variables in oval shape represent latent variables; Variables in rectangle shape represent observed variables

Model fit was assessed using the following indices: the Normed Chi-square (NC) (χ² /df), the Confirmatory Fit Index (CFI), Normed fit index (NFI), and the Root Mean Square Error of Approximation (RMSEA). A model was considered to have a reasonable error if approximation of the CFI was ≥ .90 and the NFI ≥ .90. According to Kline, a model is considered to have a reasonable error of approximation if the NC is < 5. Values of RMSEA in the range of .05 to .08 indicated an adequate fit, .08 to .10 indicated a fair fit. In the measurement model, factor loading ≥ .40 indicated good fit.^{17, 18}

Structural equation modeling is inherently a large-sample technique. We had a sample of 185 with 7 measured variables (three observed variables and 4 variables for a latent variable). The ratio of participants to free parameters was 26:1, which was much greater than a preferable ratio of 20:1.¹⁷ Thus, the power was more than sufficient for this analysis.

RESULTS

Sample Characteristics

A total of 185 patients with HF were enrolled in this study. The mean age of patients in the sample was 62 years (SD = 13). About half of the sample was female (49%). The majority of patients were Caucasian (76%) and unmarried (56%). The average left ventricular ejection fraction was 34%. More than half (57%) had reduced ejection fraction and three quarters of patients were classified in NYHA functional class III/IV (72%). Full characteristics of the sample are described in Table 1.

Table 1.

Sample characteristics, N = 185

Characteristics	Mean ± SD (range) or n (%)
Age, years	62 ± 13 (31-93)
Gender, Female	91 (49)
Race, Caucasian	141 (76)
Marital status, Married	82 (44)
Education, years	12.6 ± 2.7 (4-21)
New York Heart Association class III/IV	122 (72)
Charlson Comorbidity Index	4.1 ± 2.2 (1-11)
Body mass index, kg/m2	33.3 ± 9.3 (19-60)
Left ventricular ejection fractions, %	33.6 ± 14.6 (8-60)
Left ventricular ejection fractions < 40%	91 (57)

Measurement model

The latent variable of symptom response had four indicators, so this variable was tested for goodness of fit with confirmatory factor analysis. The factor loadings for symptom response were .77, .89, .56, and .37. The factor loadings can be interpreted as standardized regression coefficients, regressing the factors on the measure (symptom response here). Therefore, we expect factor loadings close to 1.0.¹⁷ The model produced a non-significant Chi-square of 5.845 (p = .054). Other indices indicated a good fit of the model to the data (NC = 2.92, NFI = .97, CFI = .98, RMSEA = .10). In order to determine if the measurement model could be improved, the item with the smallest factor loading was dropped (item 4 [Call your doctor or nurse for guidance] factor loading = .37). After dropping that one item, the fit indices did not improve (NFI = 1.0, CFI = 1.0, RMSEA = .369), and the Cronbach’s alpha only increased slightly (.74 to .76). Thus, the 4-item scale was retained for the final structural regression modeling analysis.

Structural equation modeling

The conceptual framework tested (Figure 1) included one exogenous variable (HF knowledge) and three endogenous variables (symptom monitoring, symptom recognition, and symptom response). The model produced a significant Chi-square of 36.495 (df = 11, p < .001). Other indices indicated a fair fit of the model to the data (NC = 3.32, NFI = .89 CFI = .92, RMSEA = .112).

Relationships among HF knowledge, symptom monitoring, symptom recognition, and symptom response

Heart failure knowledge was directly associated with symptom monitoring (β= .357, p <.001; Figure 2). When patients with HF had more HF knowledge, they monitored their symptoms more frequently. Symptom monitoring was directly associated with both symptom recognition (β= .371, p <.001) and symptom response (β= .499, p <.001). When HF patients monitored their symptoms more frequently, they 1) could recognize their HF symptoms more quickly; and 2) were more likely to take action to respond to their HF symptoms.

Figure 2: Conceptual framework tested with significant relationships.

Solid line=significant relationship; dashed line=non-significant relationship

Symptom recognition was directly associated with symptom response (β= .274, p < .001). Those who recognized their HF symptoms more quickly were more likely to take actions to respond to their HF symptoms.

Symptom monitoring was also indirectly associated with symptom response through symptom recognition. This means that HF patients who monitored their symptoms more frequently were more likely to take actions to respond to their HF symptoms because they recognized their HF symptoms more quickly.

HF knowledge was not directly associated with symptom recognition (β= .122, p = .102), nor with symptom response (β= −.001, p =.985). HF knowledge was associated with symptom recognition and symptom response indirectly through symptom monitoring. This means that HF patients who had more HF knowledge would recognize their symptoms more quickly and would be more likely to take actions to respond to their HF symptoms when they monitored their symptoms more frequently.

DISCUSSION

In this study, we explored the inter-relationships among HF knowledge, symptom monitoring, symptom recognition, and symptom response. We identified that symptom monitoring was the foremost component in symptom perception, as symptom monitoring was directly associated with both symptom recognition and symptom response and was a mediator between HF knowledge and symptom recognition and between HF knowledge and symptom response. We also identified that symptom recognition is a mediator between symptom monitoring and symptom response. That is, patients who had more HF knowledge monitored their HF symptoms more frequently, recognized their HF symptoms more quickly, and then were more likely to respond to their HF symptoms appropriately, which is consistent with the Situation-Specific Theory of Heart Failure Self-Care.^{7, 19} Our finding is valuable as we demonstrated symptom monitoring and symptom recognition are the keys to symptom response, which provides insights for clinicians and researchers to design HF self-care intervention.

To our knowledge, there is no published study examining the relationships among HF knowledge, symptom monitoring, symptom recognition, and symptoms response; therefore, we are not able to compare our findings with prior studies. Researchers have found that patients with HF had difficulty with sensing, labeling, interpreting and recognizing HF symptoms; and therefore, commonly fail to take prompt and appropriate actions to respond to their HF symptoms.^{7, 14, 19} Symptom monitoring is a foundational step in symptom perception and symptom management and plays an important role in HF outcomes. In the HF literature, adherence to weight monitoring (i.e., weight monitoring ≥ 6 out of 7 days preceding a HF-related emergency department [ED] visit or hospitalization) was associated with lower odds of HF-related ED visits or hospitalizations (odds ratio = .049).²⁰ Similarly, adherence to weight recording in a diary ≥ 80% of days during the 12 month study period was associated with fewer HF-related hospitalizations compared to patients who completed < 80% of weight diaries.²¹ Therefore, when HF patients monitor their HF symptoms more frequently (i.e., once per day), they can recognize their HF symptoms more quickly, are more likely to respond to their symptoms, and have lower HF-related ED visits or hospitalizations.

In the Theory of HF Self-Care, symptom recognition is assumed to be the key to successful self-care management (i.e., response to HF symptoms).^{6, 7} In our study, we found symptom recognition was indeed associated with symptom response and was a mediator between HF knowledge and symptom response. Patients need to regularly and frequently monitor their symptoms in order to recognize HF symptoms. When patients recognize HF symptoms quickly, they respond promptly to their symptoms. Thus, it is critical for clinicians to train their HF patients on how to monitor for HF symptoms (including daily weight monitoring) and instruct patients to monitor their HF symptoms on a daily basis diligently in order to recognize changes of their HF symptoms and respond promptly and accurately.

In the Theory of HF Self-Care, there are links between the situational characteristics (person, problem, environment), factors influencing the decision-making process (experience, knowledge, skills, and values), and self-care actions (maintenance, symptom perception, and management) that impact self-care.⁷ Thus, HF knowledge is crucial in developing patients’ expertise in decision-making process for their self-care behaviors.^{6, 7} There were multiple intervention studies aimed on improvement of patients’ HF knowledge, self-care/self-management skills, and outcomes. For example, the Rural Education to Improve Outcomes in Heart Failure (REMOTE-HF) trial offered educational interventions to HF patients in the intervention group to increase their knowledge about HF symptoms and to improve self-care and self-management.¹¹ Patients in the intervention group had better HF knowledge than those in the usual care control group at follow-ups (3 months, 12 months, and 24 months).²² However, no significant differences were found between groups in composite events of cardiac death or HF hospitalization.¹¹

Hwang and colleague found that higher levels of HF knowledge were associated with better self-care.²³ However, one quarter of participants had high HF knowledge but poor self-care.²³ In this current study, we found HF knowledge was directly associated with symptom monitoring, but not with symptom recognition, nor with symptom response. HF knowledge was associated with symptom response indirectly through the pathways of symptom monitoring and symptom recognition. Thus, enhancement of HF knowledge is required but insufficient for appropriate symptom response. Clinicians need to increase patients’ understanding/knowledge of performing HF self-care, as well as the practical skills and scenarios on how to monitor for their HF symptoms, recognize their HF symptoms, and take prompt and accurate actions to respond to their HF symptoms.

Limitations.

This study had several limitations. First, the study sample was hospitalized patients with an acute exacerbation of HF symptoms. It is possible that our findings may not be applicable to outpatients. However, we collected patients’ data during their first 3 days of hospital admission that closely reflect their HF knowledge and how they monitored, recognized, and responded to their HF symptoms during the current event. Second, this study is a secondary data analysis. Therefore, the measurement of symptom monitoring, symptom recognition, and symptom responses are items from the SCHFI, not scales specifically designed to measure symptom monitoring, symptom recognition, or symptom response. These variables have not been used as stand-alone variables in other studies. However, the SCHFI was developed based on the Situation-Specific Theory of Heart Failure Self-Care.¹⁶ The SCHFI is a reliable, valid, and widely used scale to measure self-care behaviors in patients with HF that includes specific items related to symptom monitoring, symptom recognition, and symptom response.¹⁶ The older version of the SCHFI (v. 6.2) with limited items assessing symptom perception and symptom response may be less able to reflect HF symptom monitoring, symptom recognition, and symptom response; thus, findings of this study need to be verified in a future longitudinal study using a scale to capture multi-dimensional HF symptom perception/management.

In conclusion, symptom monitoring is associated with both symptom recognition and symptom response and is a mediator between HF knowledge and symptom recognition and between HF knowledge and symptom response. This finding suggests that it is important for clinicians not just to teach/emphasize to patients with HF the importance of symptom monitoring/recognition, but also to enhance/practice their skills of symptom monitoring and symptom recognition and promote symptom monitoring among patients in order to improve symptom response in self-care intervention/training.

Table 2:

Descriptive data of variables in the structural equation model (N=185)

Item	Mean ± SD	Range (Minimum-Maximum)
Heart failure knowledge score	67.6 ± 16.7	20-100
Symptom monitoring score	5.6 ± 2.0	2-8
Weigh yourself	2.5 ± 1.2	1-4
Check your ankles for swelling	2.0 ± 1.5	1-4
Symptom recognition score	3.1 ± 1.1	0-4
Symptom response score	10.8 ± 3.7	4-16
Reduce the salt in your diet	2.8 ± 1.2	1-4
Reduce your fluid intake	2.6 ± 1.2	1-4
Take an extra water pill	2.3 ± 1.4	1-4
Call your doctor or nurse for guidance	3.1 ± 1.1	1-4

Table 3:

Standardized direct, indirect, and total effects between heart failure knowledge and symptom response mediated by symptom monitoring and symptom recognition (N=185)

Path		Effect
HF knowledge → Symptom monitoring→ Symptom response		.178
HF knowledge → Symptom recognition→ Symptom response		.034
HF knowledge → Symptom monitoring→ Symptom recognition→ Symptom Response		.036
Total indirect effects		.248
HF knowledge→ Symptom response		−.001
(Direct effect)
Total effect		.247

What is new?

• Heart failure knowledge was directly associated with symptom monitoring.

• Heart failure knowledge was associated with symptom recognition and symptom response only indirectly through symptom monitoring.

• When heart failure patients monitored their symptoms more frequently, they recognized their heart failure symptoms more quickly, and were more likely to take appropriate actions to respond to their heart failure symptoms.