Symptom Distress and Quality of Life in Patients with Advanced Congestive Heart Failure

Abstract

Little is known about the burden of illness associated with advanced congestive heart failure (CHF). Understanding the needs of this population requires further information about symptoms and other factors related to quality of life. We studied a convenience sample of 103 community-dwelling patients with New York Heart Association Class III/IV CHF. The primary outcome, quality of life (QOL), was measured with the Multidimensional Index of Life Quality (MILQ). Potential correlates of QOL included overall symptom burden (Memorial Symptom Assessment Scale, MSAS), including global symptom distress (MSAS Global Distress Index, GDI); psychological state (Mental Health Inventory-5, MHI-5); functional status (Sickness Impact Profile, SIP); spirituality (Functional Assessment of Chronic Illness Therapy-Spirituality Scale, FACIT-Spirituality); and co-morbid conditions (Charlson Comorbidity Index). Patients had a mean age of 67.1 years (SD = 12.1); were mostly white (72.8%), male (71.8%), and married (51.5%); and had a mean ejection fraction of 22.3% (SD = 6.8). The most prevalent symptoms were lack of energy (66%), dry mouth (62%), shortness of breath (56%), and drowsiness (52%). Pain was reported by about one-third of patients. For each of these symptoms, high symptom-related distress was reported by 14.1−54.1%. Quality of life was moderately compromised (MILQ composite, median = 56, possible range 12−84). Impairment in quality of life was strongly associated with global symptom distress (MSAS GDI; r = 0.74, P < 0.001); burden of comorbid conditions (r = −0.32, P = 0.002), female sex (r = −0.22, P = 0.03), functional impairment, particularly psychological impairment (r = −0.55, P < 0.001), and poorer psychological well-being (r = 0.68, P < 0.001). In multivariate analyses, impairment in quality of life was significantly related to high symptom distress, poorer psychological well-being, and poor functional mobility (R² = 0.67; P=0.002 for all). Distressful symptoms related to impaired quality of life included lack of energy (P=0.04), irritability (P=0.03) and drowsiness (P=0.02). Community-dwelling patients with advanced CHF experience numerous symptoms, significant symptom distress, and a compromised quality of life. Overall quality of life was strongly associated with symptom distress, psychological well-being and functional status. A focus on ameliorating prevalent physical symptoms and psychological distress, along with supportive measures that promote functional mobility, may lead to an improvement in the overall quality of life in this patient population.

Introduction

Congestive heart failure (CHF) is one of the leading causes of death in the Western world (1,2). The prevalence of CHF in developed countries is 1−2% of the adult population overall, and approximately 6−10% in the elderly (3-12). In the United States, CHF is responsible for more than 30,000 deaths and 700,000 hospitalizations annually (13), and has an estimated annual cost exceeding $20 billion (14). It is now the leading cause of hospitalization in the population older than 65 years. With an aging population, the prevalence of CHF is likely to rise in coming decades, and increasing costs will be attributable to both the expanding number of patients and the greater sophistication of therapy.

Despite advances in the treatment of CHF, the prognosis from the time of diagnosis is still poor. Community-based surveys have recorded mortality rates of 30−40% within one year of diagnosis and 60−70% within 5 years (12,15-17). Among those who are hospitalized, mortality rates are even higher, exceeding that of most cancers. Although some recent studies suggest that prognosis is improving (18), a very large proportion of patients with CHF require end-of-life care within a relatively short time of diagnosis.

Despite the prevalence and public health effects of CHF, few studies have evaluated the burden of illness for the individual with advanced disease (19-21). The extant literature suggests that diverse symptoms are common and distressing during the last six months of life, and that suffering from both physical and psychological symptoms increases as death approaches. Such studies suggest a need for greater attention on palliative care in advanced CHF. The impact of symptoms on quality of life is not straightforward, however, and additional data are needed to plan appropriately for the health care of this population and the need for specialist-level palliative care, including hospice (22). Studies that illuminate the complex relationships among disease status and symptoms, psychological and functional condition, spiritual concerns and global quality of life can inform a broad range of issues, including those involving the scope and focus of clinical care, access to care, and health care policy. These relationships were explored in a prospective observational study of patients with advanced CHF.

Methods

Baseline data for CHF patients enrolled in a longitudinal observational study of patients with end-stage CHF or chronic pulmonary disease known as the Heart and Lung Disease (HALD) study were used for this analysis. The study protocol was approved by the Institutional Review Boards at Beth Israel Medical Center in New York (BIMC-NY), the Massachusetts General Hospital (MGH), Boston, and the New England Research Institutes. All patients provided written informed consent before participation.

Patient Selection and Procedures

A convenience sample of patients was recruited from an eligible pool identified through the heart failure programs at both BIMC-NY and the MGH. All eligible patients between 2000 and 2002 were approached for recruitment at these sites. To be eligible, CHF patients had to be classified as New York Heart Association (NYHA) class III or IV, and have an ejection fraction <35%. Patients with co-morbidities known to be associated with short life expectancy (e.g., advanced renal, hepatic or cerebrovascular disease) were automatically ineligible. Other eligibility criteria included: life expectancy estimated to be more than three months; community (i.e., non-institutional) residence; English speaking; and no evidence of a cognitive or psychiatric disorder sufficient to preclude accurate data collection.

The research staff identified potentially eligible patients through routine medical record reviews. If a patient met screening eligibility criteria, the treating physician was contacted to verify preliminary findings and obtain consent to contact the patient. The patient was then contacted and asked to participate in the study. All medical information (e.g., NYHA class, ejection fraction) was confirmed through medical records. Trained interviewers collected data at each clinic visit.

Measures

Demographic, disease-related and treatment-related information was obtained from the medical record and the patient. Patients were also asked to complete several standardized instruments.

Charlson Comorbidity Index (CCI)

The CCI was used to characterize comorbid medical conditions obtained from medical records, thereby allowing inferences about the extent to which symptoms and other factors related to quality of life may be attributable to conditions other than CHF. The CCI designates comorbid conditions in 19 categories, each of which is primarily defined using ICD-9-CM diagnoses codes (23). Each category has an associated weight based on the adjusted risk of one-year mortality, and the overall comorbidity score reflects the cumulative increased likelihood of one-year mortality. Charlson scores can range between 0 and 35, with a higher score indicating increase in burden of comorbid disease.

Short Portable Mental Status Questionnaire (SPMSQ)

The SPMSQ is a 10-item questionnaire administered to patients that assesses cognitive status (24) and was used to screen for the impact of subtle impairment in the ability to address questions related to quality of life. The SPMSQ assesses several areas of cognitive functioning, including both short-term and long- term memory, orientation to surroundings and current events and the capacity to perform serial mathematics. An SPMSQ score of 0−2 indicates normal mental functioning; 3−4, mild cognitive impairment; 5−7, moderate cognitive impairment; and 8 or more, severe cognitive impairment.

Memorial Symptom Assessment Scale (MSAS)

The MSAS is a reliable and valid patient-rated instrument for the assessment of symptom prevalence, characteristics and degree of distress (25). The MSAS uses two or three Likert scales per symptom to measure the frequency, intensity and/or distress associated with 35 common symptoms. For the purpose of this study, an additional symptom dealing with pain localized in the chest area was added to supplement the usual pain item. Several indices can be calculated based on responses to the MSAS, including measures of overall physical symptom distress (MSAS PHYS), overall psychological symptom distress (MSAS PSYCH), and global symptom distress (MSAS Total and MSAS Global Symptom Distress or GDI). The MSAS GDI focuses on prevalent physical and psychological symptoms and is considered to be a reliable and valid indicator of global symptom distress in a variety of patient populations with chronic disease (25, 26). The MSAS GDI includes the average of the frequency scores for four psychological symptoms (e.g., feeling sad, worrying, feeling irritable and feeling nervous) and the distress scores for six physical symptoms (e.g., lack of appetite, lack of energy, pain, feeling drowsy, constipation, and dry mouth) (26). All distress scores for individual symptoms, and the symptom distress indices such as MSAS GDI, range from 0 (not at all) to 4 (very much).

Mental Health Inventory-5 (MHI-5)

The five-item, patient-rated MHI-5 measures psychological state and psychological well-being. The range of possible scores is 0 to 30, with a higher score indicating greater psychological well-being. A score of ≤18 is considered an indicator of severe depressive symptoms (27).

Sickness Impact Profile (SIP)

The 68-item SIP is a generic health measure administered to patients that assesses sickness-related impairment in somatic autonomy, mobility control, psychological autonomy and communication, social behavior, feelings, and mobility (28). The SIP also provides overall measures of physical and psychological dysfunction along with a total dysfunction score. All subscores and total score are scaled from 0 to 100, with a higher score indicating greater impairment.

Multidimensional Index of Life Quality (MILQ)

The MILQ is a validated, patient-rated instrument that contains 35 items that cover nine domains: physical health, physical health, physical functioning, mental health, cognitive functioning, social functioning, intimacy, productivity, financial status, and relationship with health professionals (29). Each item is scored on a 7-point scale, reflecting the respondent's degree of satisfaction from 1 (very dissatisfied) to 7 (very satisfied). All subscores of the MILQ have a range of 4 to 28. The MILQ composite score ranges from 12 to 84 and is a weighted sum that measures global quality of life.

Functional Assessment of Chronic Illness Therapy-Spirituality Scale (FACIT-Spirituality)

The FACIT-Spirituality is a four-item scale administered to patients that measures comfort and strength derived from one's faith (30). The score is an average of the items and ranges from 0 to 4 with higher score indicating greater spirituality.

Statistical Methods

Data are described as mean ± standard deviation (SD) for normally distributed variables, as median (minimum, maximum) for skewed variables, and as frequency (percent) for categorical variables. Although skewed for the most part, MSAS scores are presented as mean ± SD for the sake of historical comparisons with other patient populations.

To explore factors associated with overall quality of life, the MILQ composite score was considered the primary outcome for bivariate and multivariate statistical analyses. In univariate analyses, Pearson correlations were calculated between the MILQ and other variables. Categorical variables were dummy coded in order to calculate correlations. These univariate correlations yielded a potential pool of predictors for a series of multivariate analyses in which the MILQ composite score was the dependent variable. Several multiple regression models with stepwise entry were then analyzed to select the best set of non-redundant variables for predicting the MILQ composite score. All standardized measures or subscales of standardized measures were included in the univariate analyses along with the patient characteristics listed in Table 1. All univariate predictors with a P-value < 0.05 were then included for multivariate modeling. Interim stepwise analyses of subscales from the same measure were done to pre-select those subscales and minimize the number of predictors entered into the final multivariate model. Standardized regression coefficients are presented along with non-standardized coefficients in order to show the magnitude of importance of each predictor in the regression model. All tests used a significance level of P=0.05. Analyses were done using SPSS 13.0 (SPSS, Inc, Chicago).

Table 1.

Demographics and Clinical Characteristics of CHF Patients

Female		29/103 (28 %)a
Age		67.1 ± 12.1b
Marital Status	Married	53/100 (53 %)
	Divorced, Separated, Widowed	31/100 (31 %)
	Single Never Married	16/100 (16 %)
Lives Alone		27/103 (26 %)
Race	White	75/103 (73 %)
	Black	13/103 (13 %)
	Hispanic	10/103 (10 %)
	Other	5/103 (5 %)
Ejection Fraction Percent		22.3 ± 6.8b
NYHA Class	III	46/103 (44 %)
	III or IV	51/103 (50 %)
	IV	6/103 (6 %)
Charlson Comorbidity Index		3 (1, 10)c
SPMSQ		0 (0, 5)c
History of myocardial infarction		52/98 (53 %)
Peripheral vascular disease		11/99 (11 %)
History of stroke		16/99 (16 %)
Asthma		11/99 (11 %)
Emphysema or other lung disease		11/99 (11 %)
Ulcer disease		12/99 (12 %)
Diabetes		33/99 (33 %)
Kidney disease		14/98 (14 %)
Connective tissue disease		13/99 (13 %)
Cancer		12/99 (12 %)
	Leukemia or	3/99 (3 %)
	Lymphoma
	Solid Tumors	9/99 (9 %)

^aDescriptives presented as frequency/sample size (%) unless otherwise noted.

^bMean ± SD

^cMean (Minimum, Maximum)

Results

One hundred three patients with CHF were enrolled in the study (Table 1). The mean age was 67.1 years (SD = 12.1). The majority of the patients were white (72.8%) and male (71.8%), and approximately half were married (51.5%). Approximately half of the patients had a history of myocardial infarction and 16% had a history of stroke. The mean ejection fraction was 22.3% (SD = 6.8), and by definition, all patients had significant disability related to the disease (NYHA class III or IV). Overall comorbidity burden was relatively modest, as evidenced by a Charlson Comorbidity Index score of three or less in 65% of patients. A large majority of the patients had no cognitive impairment (SPMSQ median = 0, with 96 % scoring ≤ 2), and more than 75% of the patients were alive two years after data were collected.

Symptom Prevalence and Characteristics

On the MSAS, patients reported a high prevalence of many symptoms, as well as relatively high levels of symptom intensity, frequency and distress (Table 2). The median number of symptoms per patient was nine (range 0−26). Symptoms with the highest prevalence were lack of energy (66%), dry mouth (62%), shortness of breath (56%), and feeling drowsy (52%). Other prevalent symptoms included numbness or tingling in hands and feet (48.5%), difficulty sleeping (44.1%), cough (40.8%), anorexia (31.1%) and a cluster of psychological symptoms: worrying (43.7%), feeling sad (42.7%), feeling nervous (35.9%), difficulty concentrating (33%), and feeling irritable (33%). Pain was also prevalent in this population, with “chest pain or pressure” reported in 29% of patients, and “other pain” reported in 37% of patients.

Table 2.

Prevalence and Characteristics of Symptoms (n = 103)

Symptom	Overall Prevalence (%)	Frequencya	Severityb	Distressc	MSAS Scored
Lack of energy	66	75	38.2	44.8	1.76 ± 1.40
Dry mouth	62.1	37.5	17.2	14.1	1.22 ± 1.10
Shortness of breath	56.3	53.4	25.9	43.1	1.37 ± 1.35
Feeling drowsy	52.4	44.4	17	24.1	1.12 ± 1.18
Numbness or tingling in hands and feet	48.5	44	16	22	1.01 ± 1.20
Difficulty sleeping	44.1	77.8	42.2	44.4	1.22 ± 1.48
Worrying	43.7	52.3	22.2	33.3	1.02 ± 1.29
Feeling sad	42.7	43.2	27.3	34.1	0.99 ± 1.25
Cough	40.8	35.7	14.3	16.7	0.80 ± 1.06
Other pain	37.9	64.9	32.4	54.1	0.96 ± 1.36
Feeling nervous	35.9	43.2	20	38.9	0.84 ± 1.24
Itching	34.3	42.9	20	22.9	0.72 ± 1.09
Difficulty concentrating	33	29.4	11.8	29.4	0.68 ± 1.03
Feeling irritable	33	23.5	8.8	26.5	0.67 ± 1.02
Swelling of arms or legs	32	N/A	18.2	33.3	0.68 ± 1.10
Lack of appetite	31.1	62.5	26.7	25.8	0.71 ± 1.16
Chest pain or pressure	29.1	33.3	6.7	26.7	0.58 ± 0.99
Feeling bloated	28.2	34.5	17.2	17.2	0.57 ± 1.00
Dizziness	27.2	17.9	17.9	32.1	0.57 ± 0.95
Problems with urination	26.2	59.3	33.3	29.6	0.60 ± 1.07
Problems with sexual interest or activity	26	76.9	57.7	46.2	0.73 ± 1.31
Constipation	25.2	N/A	38.5	38.5	0.56 ± 1.11
“I don't look like myself”	25.2	N/A	15.4	23.1	0.49 ± 0.94
Headaches	24.5	36	12	24	0.51 ± 0.97
Sweats	21.4	31.8	13.6	18.2	0.40 ± 0.84
Weight loss	19.4	N/A	10	25	0.31 ± 0.77
Change in skin	19.4	N/A	25	30	0.42 ± 0.92
Wheezing	18.6	36.8	22.2	22.2	0.38 ± 0.88
Feeling fearful	17.5	27.8	33.3	38.9	0.40 ± 0.93
Change in the way food tastes	15.5	N/A	25	18.8	0.35 ± 0.89
Nausea	14.6	20	26.7	40	0.30 ± 0.80
Difficulty swallowing	11.7	25	16.7	16.7	0.24 ± 0.73
Mouth sores	10.7	N/A	0	9.1	0.17 ± 0.53
Hair loss	10.7	N/A	18.2	27.3	0.20 ± 0.65
Diarrhea	9.8	10	30	40	0.18 ± 0.58
Vomiting	2.9	0	33.3	33.3	0.08 ± 0.46

^aPercentage of patients with symptom describing the frequency of the symptom as “frequently” or “almost constantly.”

^bPercentage of patients with symptom describing the severity of the symptoms as “severe” or “very severe.”

^cPercentage of patients with symptom describing the distress associated with the symptom as “quite a bit” or “very much.”

^dMean ± standard deviation provided for historical comparison. MSAS Score ranges from 0−4.

The MSAS symptom descriptors for distress, and for intensity and/or frequency, provide a more detailed understanding of symptom burden (Table 2). Symptom distress, defined as “quite a bit” and “very much,” was reported by a substantial proportion of those patients with the most prevalent symptoms: lack of energy 44.8%, dry mouth 14.1%, shortness of breath 43.1%, and feeling drowsy 24.1%. High symptom-associated distress was similarly reported by 26.7% of patients with chest pain, 54.1% of those with other types of pain, and 33.3% and 34.1% of those who endorsed the symptoms of worrying and feeling sad, respectively.

The MSAS indices revealed mild to moderate distress in the overall sample, with substantial variation in each index. The median (range) for the MSAS GDI was 0.90 (0−2.80), and the medians (ranges) for the MSAS-PHYS, MSAS-PSYCH and MSAS-Total were 0.62 (0−3.22), 0.63 (0−2.77) and 0.52 (0−2.13), respectively.

Although most patients reported relatively high levels of psychological well-being (MHI-5 median = 23.5, range 5−30), 24% were at or below the cut-off for depression (Table 3). Overall, patients noted a relatively mild degree of functional impairment (SIP median = 25.7, range 0−100), and a moderate compromise in global quality of life (MILQ composite median = 56, range 12−84). The FACIT-Spirituality scale showed a high level of spirituality in this group (median = 3, range 0−4), suggesting that they were able to derive comfort and strength from their faith.

Table 3.

Descriptive Statistics for Potential Correlates of Quality of Life

Measure		Sample Size	Median (Min, Max)	Possible Range
Number of MSAS symptoms per patient		103	9 (0, 26)	0 − 36
MSAS PHYS		103	0.62 (0, 3.22)	0 − 4
MSAS PSY		103	0.63 (0, 2.77)	0 − 4
MSAS Total		103	0.52 (0, 2.13)	0 − 4
MSAS GDI		103	0.90 (0, 2.80)	0 − 4
MHI-5		100	23.5 (6, 30)	5 − 30
SIP	Somatic Autonomy	100	0 (0, 60)	0 − 100
	Mobility Control	100	44 (0, 100)	0 − 100
	Psychological Autonomy	100	18 (0, 100)	0 − 100
	Social Behavior	100	64 (0, 100)	0 − 100
	Feelings	100	17 (0, 100)	0 − 100
	Mobility	100	40 (0, 100)	0 − 100
	Overall Physical	100	21 (0, 74)	0 − 100
	Overall Psychological	100	31 (2, 91)	0 − 100
	Total Score	100	25.7 (1, 79)	0 − 100
MILQ	Mental Health	99	20 (6, 28)	4 − 28
	Physical Health	99	16 (4, 28)	4 − 28
	Physical Function	97	15 (4, 28)	4 − 28
	Social Function	92	20 (7, 28)	4 − 28
	Partner Intimacy	59	22 (9, 28)	4 − 28
	Cognitive Function	98	22 (4, 28)	4 − 28
	Financial Status	99	19 (4, 28)	4 − 28
	Health Professionals	98	25 (9, 28)	4 − 28
	Work/Productivity	96	13 (4, 28)	4 − 28
	Composite Score	97	54.4 (18, 84)	12 − 84
FACIT Spirituality		98	3 (0, 4)	0 − 4

Quality of Life: Univariate Relationships

Quality of life, as measured by the MILQ composite score, demonstrated a strong negative association with all indices of symptom distress (Table 4); the score on the MSAS GDI had the greatest absolute value (r = −0.74, P < 0.001), suggesting the strongest association between global symptom distress and impaired quality of life. Impairment in quality of life also was associated with the number of comorbidities (CCI, r = −0.32, P = 0.002), female sex (r= −0.22, P=0.03), and functional impairment, as depicted on the SIP. Almost all scales of the SIP, except somatic autonomy, showed a significant negative correlation with the MILQ; overall psychological dysfunction appeared to show a greater association than physical dysfunction (r = −0.55 versus −0.41, both P< 0.001). Psychological well being, as measured on the MHI-5, was positively related to the MILQ (r = 0.68, P < 0.001).

Table 4.

Univariate Correlations with MILQ Composite Score

Measure	MILQ Composite Score
	Correlation r	P-value
Age	0.08	0.41
Female	−0.22	0.03
Separated, Divorced, Widowed	−0.14	0.16
Lives alone	−0.13	0.20
White	0.09	0.36
Ejection Fraction	−0.12	0.23
NYHA Class	−0.14	0.17
Charlson Comorbidity Index	−0.32	0.002
Ulcer	−0.11	0.29
Connective tissue disease	−0.12	0.24
SPMSQ	−0.09	0.29
MHI-5	0.68	<0.001
SIP Somatic Autonomy	−.18	0.08
SIP Mobility Control	−.52	<0.001
SIP Psychological Autonomy	−.41	<0.001
SIP Social Behavior	−.40	<0.001
SIP Feelings	−.51	<0.001
SIP Mobility	−.40	<0.001
SIP Overall Physical	−.41	<0.001
SIP Overall Psychological	−.55	<0.001
SIP Total Score	−0.51	<0.001
FACIT Spirituality	0.16	0.11
Number of MSAS symptoms	−0.62	<0.001
MSAS Phys score	−0.63	<0.001
MSAS Psych Score	−0.68	<0.001
MSAS Total Score	−0.68	<0.001
MSAS GDI	−0.74	<0.001

Quality of Life: Multivariate Relationships

Multivariate models were constructed using the global quality of life (MILQ composite score) as a dependent variable. Variables found to be significantly correlated with the MILQ composite score were used in the multivariate analyses (Table 5). An interim stepwise analysis was used to reduce the number of potential predictor variables from the SIP. In this analysis, the MILQ composite score was regressed onto the entire SIP subscores that were significantly correlated based on the univariate analyses and the total score (Table 5). This yielded a model with two SIP predictors: SIP Mobility (beta = −0.39, P < 0.001); and SIP Feelings (beta = −0.38, P < 0.001). These were then entered into a stepwise regression model (Model 1) along with being female, the CCI, the MHI-5, and the MSAS GDI (Model 1). Three predictors were selected: MSAS GDI (beta = −0.36, P = 0.002); the MHI-5 (beta = 0.33, P = 0.002); and SIP Mobility (beta = −0.24, P = 0.002). The R² for this model was 0.62, indicating that 62% of the variance of the MILQ was explained by these three predictors.

Table 5.

Factors Associated with Quality of Life in Patients with Advanced CHF: Results of Multivariate Regression Models

Predictor		Regression Coefficients		P−value
		Unstandardized (B)	Standardized (Beta)
MILQ Model 1 (R2 = 0.62)	MSAS GDI	−7.47	−0.36	0.002
	MHI-5	0.99	0.33	0.002
	SIP Mobility	−0.13	−0.24	0.002
MILQ Model 2 (R2 = 0.67)	MHI-5	1.14	0.38	<0.001
	SIP Mobility	−0.13	−0.24	0.001
	Lack of Energy	−1.94	−0.17	0.04
	Feeling Irritable	−2.60	−0.17	0.03
	Feeling Drowsy	−2.21	−0.17	0.02

To determine which symptoms of the MSAS GDI were most important in determining its association with the MILQ composite score, another interim analysis regressed the MILQ onto the ten individual symptom distress scores comprising the GDI. Scores on five symptoms showed significance: lack of energy (beta = −0.26, P = 0.006); feeling irritable (beta = −0.24, P = 0.004); feeling drowsy (beta = −0.23, P = 0.003); feeling sad (beta = −0.22, P = 0.005); and constipation (beta = −0.16, P = 0.03).

In Model 2, the MILQ composite score was regressed onto the symptoms that were significant in the model described above, along with the MSAS GDI, the MHI-5, and SIP Mobility. The MSAS symptom shortness of breath was also added to the model because of its high prevalence (56%) and potential importance in the population with CHF. The variables that remained in this model were lack of energy (beta = −0.17, P = 0.03), feeling irritable (beta = −0.17, P = 0.03), and feeling drowsy (beta = −0.17, P = 0.04); the MHI-5 (beta = 0.38, P < 0.001); and SIP Mobility (beta =−0.24, P = 0.001). The fact that MSAS GDI was not retained in the model indicated that lack of energy, feeling irritable, and feeling drowsy accounted for the variance previously explained by MSAS GDI. All variables in the model explained 67% of the variance of the MILQ composite score.

Discussion

This survey examined symptoms and other factors associated with quality of life in a population with CHF characterized by NYHA class III or IV functional impairment and a mean (SD) ejection fraction of 22.3% (±6.8%). The patients lived at home, were not cognitively impaired, and had a relatively low burden of comorbid conditions. Thus, the findings of this study should be interpreted mainly in relation to advanced CHF as a large population of community-dwelling elderly patients who are disabled by the disease but are not imminently dying experiences it.

The most prevalent symptoms were lack of energy, dry mouth, shortness of breath, drowsiness, numbness or tingling in hands and feet, insomnia, cough, anorexia and several psychological symptoms indicative of both anxiety and depressed mood. Chest pain was experienced by one-third of the patients. The median number of symptoms per patient was nine. The distress associated with each of these symptoms varied and more than one-third of those with lack of energy, shortness of breath, difficulty sleeping, pain other than chest pain, worrying and feeling sad experienced a level of symptom-associated distress described as “quite a bit” or “very much.” The degree of global symptom distress was relatively high and was comparable to prior studies of patients with advanced cancer (32,33).

These data are consistent with prior studies of symptom prevalence and distress in populations with CHF. In a large survey of CHF patients in their last six months of life, patients experienced worsening pain, confusion and dyspnea as death approached (19). Of the 92 patients in this study who died during an index hospitalization, 35% had severe pain and 43% had dyspnea; of the 865 patients alive one year after enrollment, 18% had severe pain and 32% had dyspnea. Over time, anxiety and depression increased, functional status declined, and hospitalizations increased, causing severe financial burden on families (19). In a retrospective study of 80 Swedish CHF patients in the last six months of life, breathlessness (88%), pain (75%), and fatigue (69%) were the most prevalent symptoms (20), and limitation in physical activity (49%) and anxiety (49%) also were common. Another study that compared seriously ill hospitalized non-cancer patients (including patients with advanced CHF) with hospitalized cancer patients noted many distressing symptoms in both groups, with a higher prevalence of shortness of breath and cough among non-cancer patients (21). Finally, the recently published Cardiovascular Health Study found that patients who die from coronary heart disease in the presence of CHF, as compared to those with no history of CHF, were more likely to report a wide array of physical symptoms, use benzodiazepines, have activity restrictions and rate their health as fair or poor (31).

Overall quality of life was moderately compromised in this population with advanced CHF, and the multivariate analyses indicated that symptom distress was a major contributor to the variability in this phenomenon. In addition, specific symptoms -- lack of energy, feeling irritable, and feeling drowsy -- were independently predictive of poor quality of life.

These findings are similar to those from a recent study that used both the MSAS and the Minnesota Living with Heart Failure Questionnaire to measure the impact of symptom prevalence and symptom burden on health-related quality of life (HRQOL) in a convenience sample of 53 heart failure patients (34). These patients had a mean of 15.1±8.0 symptoms and high symptom burden. Shortness of breath and lack of energy were most prevalent, and difficulty sleeping was the most burdensome symptom. In this study, lower age, worse functional status, total symptom prevalence and total symptom burden predicted 67% of the variance in HRQOL. Together, the data from these studies suggest that interventions that favorably affect patients' ability to achieve greater mobility, improve symptom distress, and augment psychological well-being in this population may have the greatest impact in terms of improved quality of life.

Several methodological limitations may affect the interpretation of the data from our study. Patients were recruited from two academic urban medical centers and may not be representative of the total population with CHF residing in community settings. Demographic data of patients who refused consent are not available for comparison so the possibility of selection bias is also possible. The sample included a disproportionate number of white men, and thus the generalizability of the findings to different racial and ethnic groups, and more broadly, to women, remains tentative. The impact of treatment for CHF is potentially a very important mediator of outcomes in the quality-of-life domains explored in this study, and although all of the patients surveyed were under the care of a cardiologist, we were not able to characterize treatment effects. Similarly, we did not assess the extent to which pharmacologic and nonpharmacologic therapies were used to address physical and psychological symptoms, or other factors related to quality of life. These factors should be more fully explored in future studies.

In summary, this study shows that community-dwelling patients with advanced CHF experience numerous symptoms, significant symptom distress, and a moderately compromised quality of life. The latter assessment is complex and independently associated with a range of factors. Although correlational data do not allow attributions of causality, it is likely that efforts to improve symptom distress, emotional well-being, and functional impairment will eventuate in an improved quality of life. This, in turn, underscores the value of effective palliative care in the population with advanced CHF. Additional epidemiologic studies and treatment trials are needed to evaluate the types of palliative interventions that are likely to have the greatest impact in the care of these patients.