SEARCH

SEARCH BY CITATION

Abstract

  1. Top of page
  2. Abstract
  3. Methods
  4. Results
  5. Discussion
  6. Conclusion
  7. References

Primary caregivers of patients with congestive heart failure withstand enormous burden, often sacrificing their own quality of life. The relationship between caregiver burden and depression and patient quality of life and depression in this setting is unknown. Fifty outpatients were prospectively administered the Minnesota Living with Heart Failure Questionnaire and Beck Depression Inventory II (BDI-II). Caregivers were administered the Zarit Caregiver Burden Interview and BDI-II. The mean quality of life score was 35, and 26% had a BDI-II score >10. The mean Zarit Caregiver Burden Interview score was 16. Minnesota Living with Heart Failure Questionnaire, BDI-II, and Zarit Caregiver Burden Interview scores were all associated with lower ejection fraction, need for hospitalization, increased number of medications, and comorbidities. Patient Minnesota Living with Heart Failure Questionnaire score correlated with patient BDI-II, caregiver BDI-II, and Zarit Caregiver Burden Interview scores. Caregiver burden score correlated with both caregiver BDI-II and patient BDI-II. Death or hospitalization at 6 months was associated with caregiver burden and depressive symptoms and with patient quality of life and depressive symptoms. Caregivers of patients with congestive heart failure experience high caregiver burden and prevalence of depressive symptoms, which are related to the patient disease burden.

Patients who have congestive heart failure (CHF) suffer from high morbidity and mortality1–4 and poor quality of life (QOL)5–7 and a 20%–40% prevalence of self-reported depressive symptoms.1,4,8–11 Frequently overlooked in this setting is the primary caregiver. Caregiver burden is a broad term used to describe the emotional, medical, and financial challenges of providing care. Caregivers withstand enormous burden in caring for CHF patients, often putting their own well-being and QOL at risk.12 With projections for higher prevalence of CHF13,14 this will become a major cost issue for families, hospitals, and governments. Caregivers absorb a great deal of these costs15 by enabling patients to remain in the community, outside of hospitals and long-term care facilities. In our current era of increasing medical care, technology-based therapies, shorter hospital stay, home-based care, and increasing reliance on caregivers, consideration of caregiver burden and QOL become increasingly important.

There is a paucity of literature regarding the caregiver—patient relationship for patients with CHF, especially with respect to caregiver burden, depressive symptoms, and QOL. Caregiver burden may be especially heavy in the setting of chronic heart failure, as patients with this illness suffer from a high comorbid illness burden, experience polypharmacy, and have poor functional capacity and high mortality.16 This study is designed to describe levels of caregiver burden and depressive symptoms, and to correlate this with patient QOL and depression in an outpatient cohort with CHF. We hypothesized that greater caregiver burden would be associated with increasing disease burden and depressive symptoms in both the patients and caregivers.

Methods

  1. Top of page
  2. Abstract
  3. Methods
  4. Results
  5. Discussion
  6. Conclusion
  7. References

This study protocol was approved by the Queen Elizabeth II Health Sciences Centre Research Ethics Board. Study participants were outpatients attending the Heart Function Clinic at a large tertiary care hospital, the Queen Elizabeth II Health Sciences Centre located in Halifax, Nova Scotia, between July 15 and August 20, 2002. Exclusion criteria included: 1) patients living under professional care; 2) patients who identified no primary caregiver; 3) patients or caregivers unable to understand English; or 4) if either the patient or primary caregiver was not present, did not provide informed consent, or did not complete the questionnaire. After the routine clinic visit and following informed consent, patients and their primary caregivers were taken to separate rooms and given the self-administrated questionnaires to complete. Efforts were made to allow both participants to complete their self-administered questionnaires in a quiet environment, separately. Patients and caregivers were monitored during the time they completed the study to identify any potential problems. In two cases, the questionnaires were read to participants and their answers were recorded, although no assistance was provided in deciding which response was given.

Patients completed the 21-item Beck Depression Inventory II (BDIII) (Harcourt Assessment, Inc., San Antonio, TX), a well-established questionnaire for measuring depressive symptoms,17,18 and the Minnesota Living With Heart Failure Questionnaire (MLHFQ)19 (University of Minnesota; Minneapolis, MN), a disease-specific QOL tool. Caregivers completed the BDI-II and the modified Zarit Caregiver Burden (ZCB)20 interview.

The BDI-II is a 20-element questionnaire that may be self-completed. Respondents are asked to answer yes or no to each element as it relates to the past 2 weeks of their lives. Each affirmative answer incurs one point (except with two elements where the reverse is true) and the total score is additive. Scores of 10 or higher on the BDI-II are associated with increased likelihood of clinical depression. The reliability and clinical applicability are well documented and this tool has been successfully evaluated in several disease states.21,22 The MLHFQ is the most extensively used disease-specific QOL instrument for heart failure and has been extensively used. The reliability and validity of MLHFQ have been established, especially in a relatively stable, outpatient population.23–25 This questionnaire can also be self-completed and consists of 21 questions to which the respondent may provide an ordinal response to each element on a Likert scale, from zero to five. To assess caregiver burden, the ZCB interview was used.20,26 The interview consists of 22 questions for the caregiver, who scores them on a five-point scale and is the most widely referenced scale in studies of caregiver burden.12 Each questionnaire typically requires 10 minutes to complete.

Patient data obtained from the chart included: age; gender; number of persons living with patient; marital status; previous medically treated depression; number of years with CHF; whether hospitalized in the past 6 months; number of visits to the Heart Function Clinic in the past 6 months; ejection fraction; serum creatinine; New York Heart Association (NYHA) functional class; and presence of diabetes mellitus, hypertension, cancer, or dialysis. Data obtained from the caregiver included age, gender, whether employed, number living in the home, and relationship to the patient. Six-month composite death and hospitalization for the patients and their caregivers were also recorded.

Comparisons of continuous variables were made using Student t testing, and the chi-square statistic was used for categoric comparisons. We used simple linear correlation to describe the relationships between caregiver burden score, each of the other three questionnaire scores, and the various clinical characteristics. While the questionnaire scores were primarily treated as continuous variables, we also analyzed each score in categoric fashion. The BDI-II was assessed using a cut-point of 10, as this value is associated with high likelihood of clinical depression and adverse outcomes in other studies. The MLHFQ scores were divided into those above and below the median of 30, with higher scores indicating a poorer QOL. The ZCB scores were divided into those above and below 20, with higher scores stated to be associated with high caregiver burden, such as with advanced Alzheimer's disease. Statistical program SPSS 9.0 (SPSS Inc., Chicago, IL) was used for analysis.

Results

  1. Top of page
  2. Abstract
  3. Methods
  4. Results
  5. Discussion
  6. Conclusion
  7. References

Patient Demographics and Clinical Characteristics. A total of 50 patients and 50 primary caregivers participated in the survey. During the study period, nine patients were excluded due to lack of caregiver present, three for lack of informed consent, and one due to inability to fully complete the questionnaires. Participants noted no untoward or adverse events during the study. The average age of CHF patients was 72±11 years (median, 73 years); 72% were male. Mean caregiver age was 61±14 years (median, 62 years); 80% were female. Seventytwo percent of patients were in NYHA class III or IV at the time of interview. They had a longer duration of disease than those in NYHA class I–II but were not older. Other clinical characteristics were similar. Our patient cohort experienced seven hospitalizations and five deaths during 6-month follow-up, with none occurring in the NYHA class I–II group. There were four hospitalizations (one for acute coronary syndrome) and no deaths in the caregiver group. Tables I and II set forth the baseline characteristics of patients and caregivers, respectively.

Table I.  Baseline Patient Characteristics (N=50)
Characteristic 
Age (M ± SD) (yr)72±11
Male gender (%)72
Number of clinic visits in the past 6 mo (M ± SD)5.7±7
Duration of CHF diagnosis (M ± SD) (yr)3.3±3.3
NYHA functional class III or IV (%)72
History of hypertension (%)68
Current smoker (%)6
Diabetes (%)34
Prior myocardial infarction (%)60
Ischemic etiology of CHF (%)76
Ejection fraction (M ± SD)0.33±0.14
Serum creatinine (M ± SD) (μmol/L)168±100
Number of prescribed medications (M ± SD)8.7±3.4
Hospitalized within the previous 6 months (%)45
Income <Can $20,000 (?US $16,300] (%)54
Education (%) 
Not high school graduate70
High school graduate30
Marital status 
Currently married (%)76
M=mean; CHF=congestive heart failure; NYHA=New York Heart Association
Table II.  Baseline Caregiver Characteristics (N=50)
Characteristic 
Age (M ± SD) (yr)61±14
Female gender (%)80
Relative of patient (%)96
Spouse of patient (%)66
Marital status (%) 
Currently married90
Widowed/divorced/separated10
Employment (%) 
Full-time20
Retired66
Part-time/disability/unemployed14
ZCB score (M ± SD)16.0±14.4
BDI-II score (M ± SD)5.6±6.4
History of treated episode of depression (%)14
M=mean; ZCB=Zarit Caregiver Burden interview; BDI-II=Beck Depression Inventory II

BDI-II: Depressive Symptoms Scores. The mean caregiver BDI-II index was 5.6, and the mean patient BDI-II index was 7.1. Of the patient cohort, 13 of 50 (26%) scored >10 (high likelihood of clinical depression) vs. nine of 50 (18%) among caregivers (p=0.18). Patients with BDI-II scores of >10 used a significantly greater number of prescribed medications and had higher MLHFQ scores; trends were seen with the presence of diabetes and increasing age (Table III). Associations with BDI-II scores >10 in caregivers were lower age, greater number of patient medications, and higher caregiver burden score. Gender was not by itself predictive, but male caregivers of female patients had much lower BDI-II scores than female caregivers of male patients (6.8 vs. 23.5; p<0.05).

Table III.  Patient and Caregiver Characteristics by Beck Depression Inventory II (BDI-II) Score
 Patient BDI-II ScoreCaregiver BDI-II Score*
Patient Characteristic<10 (n=37)≥10 (n=13)p Value<10(n=41)>10 (n=9)p Value
Age (M) (yr)71760.0774640.006
Male gender (%)70770.1469890.10
Time since CHF diagnosis (M) (yr)3.23.70.353.33.70.37
NYHA functional class III–IV (%)68850.1376560.18
Hypertension (%)68850.0776780.24
History of diabetes mellitus (%)41150.0537220.21
Income <Can $20,000 (?US $16,300) (%)54460.4351670.2
Education
Not high school graduate (%)73620.0968780.19
Marital status
Married (%)73850.3376780.45
Prior myocardial infarction (%)62540.4224220.45
Ejection fraction (M)0.330.290.310.320.310.44
Serum creatinine (M) (μmol/L)1531940.141621980.16
Number of medications (M)8.110.40.028.310.30.05
MLHFQ score (M)30.848.60.0001   
Caregiver BDI-II score (M)5.26.70.246.78.90.17
Caregiver ZCB score (M)14.919.20.183.216.4<0.001
6-Mo death/hospitalization (n)210<0.001570.002
M=mean; CHF=congestive heart failure; NYHA=New York Heart Association; MLHFQ=Minnesota Living With Heart Failure Questionnaire; ZCB=Zarit Caregiver Burden interview; *characteristics listed are for the patients the caregivers cared for

MLHFQ: QOL Scores. The mean patient MLHFQ score was 35 (median, 30). This is generally higher (indicating a poorer QOL) than many reported scores for outpatient cohorts, which tend to score <30. Poorer QOL score was associated with greater BDI-II score, number of medications, comorbid conditions, serum creatinine, and NYHA class, and with recent hospitalization, male gender, and lower ejection fraction (see Table IV).

Table IV.  Patient Characteristics by Minnesota Living With Heart Failure Questionnaire (MLHFQ) Quality of Life Score*
 MLHFQ Score*
Patient Characteristic<30 (n=19)≥30 (n=31)p Value
Age (M) (yr)73.9710.19
Male gender (%)58810.04
Number of clinic visits in past 6 mo (M)3.47.10.04
Time since CHF diagnosis (M) (yr)3.23.40.39
NYHA functional class III–IV (%)58810.02
Hypertension (%)74610.25
Diabetes (%)37320.37
Prior myocardial infarction (%)58620.08
Ejection fraction (M)0.370.280.01
Serum creatinine (M) (μmol/L)1371880.04
Number of medications (M)7.19.60.004
Income <Can $20,000 (?US $16,300) (%)47580.32
Education
Not high school graduate (%)68710.25
Marital status
Married (%)68710.42
6-Mo death/hospitalization (n)390.04
M=mean; CHF=congestive heart failure; NYHA=New York Heart Association; *median MLHFQ score of 30 used for cutoff

ZCB: Caregiver Burden Scores. The mean ZCB score was 16. Table V demonstrates the associations of caregiver burden with patient characteristics. Caregivers with lower income had higher caregiver burden ZCB scores.

Table V.  Patient Characteristics by Zarit Caregiver Burden Interview (ZCB) Score
 ZCB Score
Patient Characteristic<20 (n=36)≥20 (n=14)p Value
Age (M) (yr)73710.29
Male gender (%)72790.26
Number of clinic visits in past 6 mo (M)5.95.10.37
Time since CHF diagnosis (M) (yr)3.53.00.33
NYHA functional class III–IV (%)71710.45
Hypertension (%)59860.05
Diabetes (%)38290.31
Prior myocardial infarction (%)59570.15
Ejection fraction (M)0.330.320.45
Serum creatinine (M) (μmol/L)1681950.13
Number of medications (M)9.38.40.35
Income (%)
<Can $20,000 (?US $16,300)50720.02
≥Can $20,0005028 
Education
Not high school graduate (%)65790.11
Marital status
Married (%)74790.4
6-Mo death/hospitalization (n)390.002
M=mean; CHF=congestive heart failure; NYHA=New York Heart Associatio

Correlations Among Depression, QOL, and Caregiver Burden. There was a significant correlation between patient QOL and patient depression scores (r=0.62; p<0.0001), caregiver burden and caregiver depression scores (r=0.61; p<0.0001), and caregiver burden and patient depression scores (r=0.28; p<0.05). The Figure (panels A, B, and C, respectively) show these correlations graphically. There was a trend between caregiver burden and patient MLHFQ score (r=0.25; p=0.07) and between caregiver BDI-II and patient BDI-II (r=0.24; p=0.09).

image

Figure Figure. Correlations between: A) patient quality of life (Minnesota Living With Heart Failure Questionnaire [MLHFQ]) score and patient depression (Beck Depression Inventory II [BDI-II]) score (r=0.62; p<0.0001); B) caregiver depression (BDI-II) score and caregiver burden (Zarit Caregiver Burden interview [ZCB]) score (r=0.61; p<0.0001); and C) caregiver burden (ZCB) score and patient depression (BDI-II) score (r=0.28; p<0.045)

Patient Outcomes. Interestingly, patients with worse QOL and depressive symptomatology were more likely to suffer death or hospitalization over the following 6 months. Patients of caregivers with higher caregiver burden and depressive symptomatology were similarly more likely to suffer death or hospitalization over the following 6 months (Tables III–V).

Discussion

  1. Top of page
  2. Abstract
  3. Methods
  4. Results
  5. Discussion
  6. Conclusion
  7. References

Our results show that caregivers of patients with CHF suffer a high level of caregiver burden and that this is related to their QOL. Increased caregiver burden is also associated with more advanced disease burden in their charges in terms of patient QOL, level of depressive symptoms, and disease burden (illustrated by number of medications and comorbidities). These data indicate the interrelation between patient and caregiver, their respective QOL, and disease progression, and are reported, to our knowledge, for the first time. Additionally, these factors were associated with worse patient outcomes at 6 months.

Caregiver burden has been best researched with respect to caring for patients with dementia,12,26 where reported stress levels and caregiver burden are higher, and sense of well-being lower, than caregivers for patients with other chronic diseases. Our results show caregivers of CHF patients withstand a huge burden (mean ZCB score of 16), although lower than caregivers of cognitively impaired patients, where the mean scores tend to be nearly 20.15,26

Recent research in caregivers of patients with dementia suggests that patient functional impairment is correlated to the degree of behavioral problems, whereas caregiver depression relates more to caregiver burden. Caregivers of cognitively impaired individuals have been found to have significantly higher incidences of depression than the general population.27–29 Negative behavior by the patient and limited self-care are very significant factors in determining caregiver burden and depression. In 613 patients with dementia and their caregivers, Clyburn et al.30 showed that four factors primarily predicted caregiver burden and depression: functional limitations, disturbing behaviors, patient residence, and assistance given to the caregiver. The former two predicted worse depression and burden and the latter two predicted the opposite. This suggests that institutionalization of patients may mitigate the effects of illness on caregivers. This suggests further that patients with dementing illnesses may exert additional burden on their caregivers through behavioral mechanisms. This may also be present in our cohort.

Caregiver burden of spouses related to CHF increases with the number of caregiving activities and is associated with poor patient prognosis and rehospitalization rates.3,8,10,15,30 A positive correlation between marital QOL and CHF outcomes has been described,31,32 but correlation with caregiver depression or patient QOL or disease burden has not been reported. This study adds to that literature in that we have demonstrated a correlation of patient outcomes with caregiver depressive symptomatology and burden.

Our findings show that caregiver burden and caregiver depression are related. Those caregivers who had elevated depression levels (BDI-II ≥10) had much higher burden scores. It is not clear whether this indicates that being depressed makes caregiving feel more burdensome, whether the feeling of being burdened causes caregivers to become more depressed, or some combination, although there is research to indicate that increases in caregiver demands result in increases in caregiver depressive symptoms.31 Since caregiver burden and depression are related, and depressive symptomatology is known to be related to subsequent cardiovascular events, it would seem reasonable to implement strategies whereby caregiver burden is alleviated, in the same vein as other important public health activities.

The positive relationship between caregiver burden and low income may be related, at least in part, to poor social support, a factor known to be more common in those of lower socioeconomic status, and a known contributor to caregiver burden in specific patient populations.33 It is noteworthy that those in the lower income group took more medications, had lower ejection fractions, higher caregiver burden, and an increased likelihood of prior myocardial infarction when compared with those in the higher income groups. These findings are consistent with the literature34–38 and may be partially due to poorer diet, lower activity levels, and higher smoking rates. Existing resources for those who are socially disadvantaged should be a consideration in the implementation of care for CHF patients at home.

Approximately 4.3% of the general population is expected to experience depression at any given time.39 Caregivers of the CHF patients in this study had much higher levels of depressive symptoms, with 18% having BDI scores >10, a level associated with a likelihood of at least mild-to-moderate clinical depression. The scores we observed are similar to those of caregivers of other patient populations.40 Depression in caregivers was not related to the duration of the illness, contrary to findings with caregivers of patients with Parkinson's disease.35

Patient depressive symptom prevalence was similar to the past incidence of treated depression, 26% and 24%, respectively. Depression is associated with increased risk of cardiovascular events in both the general population and in those with previous cardiac events.41 Patients with CHF report higher BDI-II scores than those with other cardiovascular conditions.42 While treatment of depression has not been shown to improve cardiovascular outcomes, it still should be identified and treated, especially now that there is evidence to suggest it is likely safe to treat clinically depressed patients with recent myocardial infarction or unstable angina with selective serotonin uptake inhibitors.43

Our finding that the number of medications was associated with worse patient QOL, patient depression, and caregiver depression is interesting. The number of medications is likely a surrogate for disease burden and provides further evidence that this may be one of the better surrogates for overall disease burden. Further studies, including number of medications as a measured variable, may shed further light on this hypothesis.

Patient QOL in this study was poor and correlated with patient depression. These data are consistent with other reports in the setting of Parkinson's disease, dementia, and other illnesses. While it is intuitive that a depressed patient would have a lower QOL and vice versa, it is not known to what extent each is a primary or secondary phenomenon. Indeed, we found many interesting correlations between caregiver and patient questionnaire scores, but this study was not designed to apply causality to these findings. It is tempting, however, to speculate that as disease burden increases, so do the stresses on the family/caregiver. These stresses may lead to mood disorders and trigger adverse behaviors that may affect patient outcomes. Alternatively, depressive symptomatology in caregivers may simply reflect the reality of caring for someone with chronic severe disease. Nevertheless, it would be important to determine which caregiver factors may independently affect patient outcomes, as this would provide a target for intervention.

The consequences of high levels of caregiver burden include an increased risk of hospitalization, as well as an increase in the use of formal in-home services.44 High levels of caregiver burden correlate with increased morbidity and mortality in caregivers.26,44 We found that only four of the caregivers in our study were hospitalized (one with acute coronary syndrome) over the 6 months following our survey and none died. Several factors are known to decrease caregiver burden, such as social supports—specifically, frequent visits by other family members25—and active coping skills and management strategies. These and other strategies should be formally evaluated in CHF caregiver populations. As a result of these observations, we believe future interventions in this population that influence psychosocial factors should be designed and evaluated, incorporating outcome measures of these factors. Such interventions might include increased respite support, psychological counseling where appropriate, home care, or cardiac rehabilitation or disease management programs. Of these, increased access to heart failure clinics has been shown to improve patient outcomes but has not been evaluated for its effect on caregiver burden. Ironically, since these programs are associated with less rehospitalization, they could conceivably increase caregiver burden by increasing the time of exposure between caregiver and patient. The home hospital concept has been shown in the geriatric setting to reduce stress on the caregiver. Exercise rehabilitation for patients is a promising therapy that has shown benefit mostly in younger patients with CHF. The large, randomized study Heart Failure: A Controlled Trial Investigating Outcomes of Exercise Training (HF-ACTION) is currently underway and is designed to evaluate the effects of formal exercise rehabilitation training on total mortality in patients with CHF due to left ventricular systolic dysfunction.

This study has several limitations. We examined only stable CHF outpatients visiting a CHF clinic with their primary caregiver, and so we had a selected cohort. The relatively small sample size (n=50) of caregiver and patient pairs limited our abilities to perform multivariate analysis and further discern associations. Our determination of depressive symptoms was through a validated questionnaire, rather than the gold standard—a structured psychiatric interview. The confounding effects of medications could not be assessed. We evaluated only outpatients who were already involved in a highly structured and specialized CHF clinic. Many of these patients had experience in the discussion of mood symptoms and completing questionnaires. Consequently, they may not be a truly representative, broad-based CHF population. Our patients all had family members as primary caregivers. These results do not, therefore, apply to patients who do not require supportive care, those who have professional care, or those admitted to acute or chronic care facilities. Nevertheless, some of the same issues seen in this report could potentially be important in those populations.

Conclusion

  1. Top of page
  2. Abstract
  3. Methods
  4. Results
  5. Discussion
  6. Conclusion
  7. References

Patients with CHF and their caregivers experience poor QOL and a high incidence of depressive symptomatology. Their caregivers experience high caregiver burden and depressive symptoms, which are related to patients' QOL, disease burden, and their own caregiver burden. The number of prescribed medications and subsequent 6-month outcomes are associated with these factors. These data suggest that both patient and caregiver factors are related to patient outcomes. Future research should focus on the role of the primary caregiver in subsequent patient outcomes in the CHF population.

Acknowledgment and disclosure: The authors thank the nursing staff at the Heart Function Clinic, Anna Svendson and Marion Malloy, for their assistance and cooperation in implementing the study. This study was presented in abstract form at the 7th Annual Heart Failure Society of America Meeting in September 2003. This study was funded by the Undergraduate Internal Medicine Research Foundation (UIMRF) of Dalhousie University, Halifax, Nova Scotia.

References

  1. Top of page
  2. Abstract
  3. Methods
  4. Results
  5. Discussion
  6. Conclusion
  7. References