Abstract
BACKGROUND
Care for patients with advanced heart failure (HF) has traditionally focused on managing HF alone; however, little is known about the prevalence and contribution of comorbidity to mortality among this population. We compared the impact of comorbidity on mortality in older adults with HF with high mortality risk and those with lower mortality risk, as defined by presence or absence of a prior hospitalization for HF, respectively.
METHODS
This was a retrospective cohort study (2002–2006) of 18,322 age-matched and gender-matched Medicare beneficiaries. We used the baseline year of 2002 to ascertain HF hospitalization history, in order to identify beneficiaries at either high or low risk of future HF mortality. We calculated the prevalence of 19 comorbidities and overall comorbidity burden, defined as a count of conditions, among both high and low risk beneficiaries, in 2002. Proportional hazards regressions were used to determine the effect of individual comorbidity and comorbidity burden on mortality between 2002 and 2006 among both groups.
RESULTS
Most comorbidities were significantly more prevalent among hospitalized versus non-hospitalized beneficiaries; myocardial infarction, atrial fibrillation, kidney disease (CKD), chronic obstructive pulmonary disease (COPD), and hip fracture were more than twice as prevalent in the hospitalized group. Among hospitalized beneficiaries, myocardial infarction, diabetes, COPD, CKD, dementia, depression, hip fracture, stroke, colorectal cancer and lung cancer were each significantly associated with increased hazard of dying (hazard ratios [HRs]: 1.16-1.93), adjusting for age, gender and race. The mortality risk associated with most comorbidities was higher among non-hospitalized beneficiaries (HRs: 1.32-3.78).
CONCLUSIONS
Comorbidity confers a significantly increased mortality risk even among older adults with an overall high mortality risk due to HF. Clinicians who routinely care for this population should consider the impact of comorbidity on outcomes in their overall management of HF. Such information may also be useful when considering the risks and benefits of aggressive, high-intensity life-prolonging interventions.
KEY WORDS: heart failure, mortality, comorbidity
INTRODUCTION
Heart failure (HF) is a serious progressive illness that is highly prevalent among older adults1. Care for patients with advanced HF has primarily centered around treating and managing the HF2,3; however, there is growing recognition of the need to consider comorbidity in treatment decision making3. Prior research has demonstrated the high prevalence of comorbidity among patients with HF overall4 and its association with increased mortality5–8. Less is known about the comorbidity profile of patients who survive to the advanced stages of HF and its contribution to mortality. While several comorbidities have been shown to increase the risk of mortality among patients with a range of HF severity5–8, the effect of comorbidity on mortality among patients with advanced HF has not been characterized. In cancer, a disease with high prognostic certainty, the impact of comorbidity on mortality decreases as the cancer progresses to advanced stages and becomes the primary determinant of mortality9–11. Whether there is a similar decrease in the effect of comorbidity on mortality in the advanced stages of diseases with more variable and uncertain trajectories, such as HF, is unknown.
We used a Medicare database that identifies a wide range of chronic conditions to examine the relationship between comorbidity and mortality among subgroups of older adults with HF. Because this administrative database lacked physiologic markers to identify older adults with advanced HF, we used a history of HF hospitalization during 2002 to identify beneficiaries at high risk for future HF mortality. The purpose of the study was to 1) describe the prevalence of comorbidity and 2) examine the associations between comorbidity and mortality among older adults with HF with high mortality risk as compared to those with lower mortality risk.
METHODS
Design Overview
We conducted a retrospective longitudinal cohort study of Medicare beneficiaries with HF. Participants were selected from the Center for Medicare and Medicaid Services’ Chronic Conditions Warehouse (CCW), a dataset containing beneficiary, assessment, and claims data for a 5% random sample of Medicare beneficiaries. The CCW includes 21 chronic condition variables updated annually indicating the presence of a condition as defined by evidence-based algorithms that specify 1) a minimum number and type of diagnoses/procedure codes, 2) occurring within a specific look-back period and 3) within certain care settings. Identification of HF is derived from the occurrence of at least one of the following International Classification of Diseases, Ninth revision (ICD-9) codes: 398.91, 402.01, 402.11, 402.91, 404.01, 404.11, 404.91, 404.03, 404.13, 404.9, or 428.xx, occurring within a 2-year period in either the inpatient or outpatient setting.
Study Population
We began with all living CCW beneficiaries identified as having HF as of January 1st, 2001 (n = 241,254). We excluded 1) beneficiaries age ≤ 65 years (n = 20,938) to ensure that all included beneficiaries were at least 65 years of age during the lookback period; 2) beneficiaries who received care outside traditional fee-for-service Medicare at any point between January 1st, 2000 and December 31st, 2006 (n = 18,852) and 3) beneficiaries with missing data during the study period (n = 334). The final sample included 201,130 Medicare beneficiaries with HF.
In the absence of physiologic measures to identify patients with advanced HF12, we used a single hospitalization for HF, identified as the primary discharge diagnosis, to identify beneficiaries at high risk for HF mortality. Considerable evidence points to the high mortality risk associated with hospitalization in this population13–17, which is driven by both the stage of HF and also by factors independent of HF status, such as preferences for treatment, access to care, and psychosocial support. In order to maximize the likelihood of identifying beneficiaries with an initial decompensation of their HF, we defined the hospitalized group as beneficiaries with a HF diagnosis and at least one HF hospitalization in 2002, but no HF hospitalizations in 2001. We matched hospitalized beneficiaries on the date of their live discharge from their first hospitalization for HF in 2002 to non-hospitalized beneficiaries - those who had not had a HF hospitalization as of that date. Once matched, beneficiaries in the non-hospitalized group retained their classification throughout the study period regardless of subsequent hospitalizations. We matched 1:1 according to age and gender. If there were multiple matches, one was chosen at random. This process resulted in a final study population of 9,166 matched pairs (n = 18,322). Matched pairs were followed from their match date in 2002 until either death or censorship as of December 31st, 2006. Date of death was ascertained through the CMS Denominator File.
To validate the ability of hospitalization to identify beneficiaries in our study at higher risk of mortality, we calculated 1-year Kaplan–Meier (KM) survival curves for the matched population of hospitalized and non-hospitalized beneficiaries (n = 18,332) (Appendix Fig. 2). At the end of one year, approximately 34% of the hospitalized beneficiaries had died, compared to only 10% of non-hospitalized beneficiaries. Using Cox regression analysis we found a large increase in mortality risk associated with hospitalization (HR: 3.04; p < 0.0001). To address the concern that the HF hospitalization may represent complications from or the progression of a comorbidity, we compared 1-year KM survival curves (Appendix Fig. 3) within a subset of matched hospitalized and non-hospitalized beneficiaries who had no comorbidities (n = 332) and found a similarly increased mortality risk with hospitalization (HR: 3.94; p < 0.0001).
Figure 2.
The 1-year Kaplan–Meier survival curves of hospitalized and non-Hospitalized beneficiaries (n = 18,332).
Figure 3.
The 1-year Kaplan–Meier survival curves of hospitalized and non-Hospitalized beneficiaries with no comorbid conditions (n = 332).
Chronic Condition Measures
We examined 19 cardiac (myocardial infarction, atrial fibrillation, and ischemic heart disease) and noncardiac (dementia, cataracts, chronic kidney disease [CKD], chronic obstructive pulmonary disease [COPD], diabetes, depression, glaucoma, hip/pelvic fracture, osteoporosis, rheumatoid arthritis/osteoarthritis [arthritis], stroke, and cancer [female breast, colorectal, prostate, lung, endometrial]) comorbidities18. We included in our analyses all conditions that met the respective algorithm criteria for identification by the baseline study year 2002. Because these are chronic conditions, we considered them to be present from the time they were identified through the study period.
We examined both individual comorbidities and comorbidity burden, defined as the count of individual conditions. We grouped beneficiaries according to the distribution of comorbidities in the study sample: ≤2, 3–4, and 5+ comorbidities.
Analysis
We used descriptive statistics to summarize demographic characteristics and number of comorbidities present among the hospitalized and non-hospitalized groups. Because of the matched pairs design, we compared the percentages of individual comorbidities at baseline in the two groups using McNemar’s test for equality of paired data.
We conducted longitudinal survival analyses using Cox regression models to determine the association of both individual comorbidities and comorbidity burden with mortality. In the models that used matched data, the Lin and Wei19 robust covariance matrix estimator was used to account for potential clustering induced by matching19. The proportional hazards assumption was tested in all regression models using the log-negative-log survival curve method.
We estimated the effect of comorbidity burden on mortality within the hospitalized and non-hospitalized groups in separate multivariable analyses adjusting for age, gender and race. We tested for the equality of the comorbidity burden hazard ratios for beneficiaries in each group by examining the entire cohort using a model that included an interaction term crossing comorbidity burden with hospitalization status.
We examined the effect of individual comorbidities on mortality within the hospitalized and non-hospitalized groups using the same approach as for comorbidity burden. Comorbidities significantly associated with mortality in either group were included in a second multivariable regression that also adjusted for the presence of other comorbidities. Finally, we calculated age-stratified median survival for specific combinations of comorbidities that were the most prevalent and conferred the highest mortality risk among hospitalized beneficiaries in the multivariable analyses. Statistical significance was set at P < 0.05 for two-sided tests except for results reported in Table 2 for which a Bonferroni-corrected p-value of < 0.003 was considered statistically significant. All analyses were performed using SAS® statistical software (Version 9.2, SAS Institute, Cary, North Carolina).
Table 2.
Prevalence of Individual Comorbid Conditions Among Hospitalized and Non-hospitalized Beneficiaries
| Individual conditions, n (%) | Hospitalized Beneficiaries (n = 9166) | Non-hospitalized Beneficiaries (n = 9166) | p-value | ||
|---|---|---|---|---|---|
| Ischemic Heart Disease* | 8,056 | (87.9) | 5,910 | (64.5) | < 0.0001 |
| Diabetes * | 4,617 | (50.4) | 3,122 | (34.1) | < 0.0001 |
| COPD*† | 4,263 | (46.5) | 1,866 | (20.4) | < 0.0001 |
| Atrial fibrillation* | 4,110 | (44.9) | 1,884 | (20.6) | < 0.0001 |
| CKD*† | 3,670 | (40.0) | 1,365 | (14.9) | < 0.0001 |
| Arthritis* | 2,826 | (30.8) | 2,573 | (28.1) | < 0.0001 |
| Dementia* | 2,266 | (24.7) | 2,000 | (21.8) | < 0.0001 |
| Cataracts* | 2,007 | (21.9) | 2,520 | (27.5) | < 0.0001 |
| Depression* | 1,960 | (21.4) | 1,213 | (13.2) | < 0.0001 |
| Osteoporosis | 1,259 | (13.7) | 1,195 | (13.0) | 0.1559 |
| Stroke * | 1,224 | (13.4) | 845 | (9.2) | < 0.0001 |
| Glaucoma | 906 | (9.9) | 982 | (10.7) | 0.0648 |
| Myocardial infarction* | 803 | (8.8) | 161 | (1.8) | < 0.0001 |
| Prostate Cancer | 312 | (3.4) | 325 | (3.6) | 0.5903 |
| Hip Fracture* | 260 | (2.8) | 145 | (1.6) | < 0.0001 |
| Breast Cancer | 131 | (1.4) | 167 | (1.8) | 0.0345 |
| Lung Cancer* | 131 | (1.4) | 83 | (0.9) | 0.0010 |
| Colorectal Cancer | 121 | (1.3) | 113 | (1.2) | 0.5978 |
| Endometrial Cancer | 11 | (0.1) | 9 | (0.1) | 0.6547 |
*Statistically significant at Bonferroni-corrected p < 0.003;
†CKD = Chronic kidney disease; COPD = Chronic obstructive pulmonary disease
RESULTS
Hospitalized beneficiaries had a median age of 81 years (IQR: 75–86) and were predominantly female (59%) (Table 1). Because of the matched study design, these characteristics were similar for non-hospitalized beneficiaries. At baseline, 16.5% of non-hospitalized beneficiaries and 42% of hospitalized beneficiaries had 5+ comorbidities. Median survival was 1.9 (IQR: 0.54, 4.4) years for hospitalized beneficiaries and 4.6 (IQR: 1.9, >5) years for non-hospitalized beneficiaries.
Table 1.
Baseline Characteristics of Study Cohort (n = 18,332)
| Characteristic | Hospitalized Beneficiaries (n = 9166) | Non-hospitalized Beneficiaries (n = 9166) | ||
|---|---|---|---|---|
| Age, years, median (IQR) | 81 | (75–86) | 81 | (75–86) |
| Female, n (%) | 5416 | (59.1) | 5416 | (59.1) |
| Race/Ethnicity, n (%) | ||||
| White | 7783 | (84.9) | 7915 | (86.4) |
| Black | 1026 | (11.2) | 825 | (9.0) |
| Hispanic | 179 | (2.0) | 204 | (2.2) |
| Number of chronic conditions, median (IQR) | 4 | (3–5) | 3 | (2 – 4) |
| <= 2 conditions, n (%) | 1,342 | (14.6) | 4070 | (44.4) |
| 3-4 conditions | 3,975 | (43.4) | 3,587 | (39.1) |
| 5+ conditions | 3,849 | (42.0) | 1,509 | (16.5) |
| Median Survival Time, years (IQR) | 1.8 | (0.54, 4.4) | 4.6 | (1.9, >5) |
The majority of individual comorbidities, both cardiac and noncardiac, were significantly more prevalent among hospitalized beneficiaries than among non-hospitalized beneficiaries (Table 2). Burden of comorbidity was associated with a significantly increased hazard of dying among both groups. Having 3–4 comorbidities was associated with a 22% higher hazard of dying among AHF hospitalized beneficiaries and a 49% higher hazard of dying among NHF non-hospitalized beneficiaries, compared to having ≤2 comorbidities (Table 3).
Table 3.
Effect of Burden of Comorbidity on Mortality among Beneficiaries with Heart Failure*
| Burden of Comorbidity | Adjusted Hazard Ratio (95% C.I.) | Interaction between Hospitalization status and comorbidity burden | |
|---|---|---|---|
| Hospitalized Beneficiaries | Non-hospitalized Beneficiaries | p-value | |
| Low (<=2 conditions) | 1.0 (Reference) | 1.0 (Reference) | - |
| Medium (3–4 conditions) | 1.22 (1.13, 1.31) | 1.49 (1.39, 1.59) | 0.0003 |
| High (5+ conditions) | 1.57 (1.44, 1.68) | 2.35 (2.17, 2.54) | < 0.0001 |
*Adjusted for age, gender and race
Most comorbidities were associated with a significantly increased mortality risk among hospitalized beneficiaries (Table 4). For example, among cardiac conditions, myocardial infarction was associated with a 52% higher hazard of dying. Among noncardiac conditions, CKD was associated with a 62% and dementia with a 50% higher hazard of dying. The majority of comorbidities were associated with a significantly higher relative risk of mortality among non-hospitalized beneficiaries as compared to hospitalized beneficiaries. However, there was no significant difference in the mortality risk associated with ischemic heart disease, diabetes, osteoporosis, or any of the cancers other than lung cancer, between the two groups. When adjusting for the presence of other conditions, the majority of comorbidities continued to confer a significant, albeit decreased, mortality risk among both groups of beneficiaries (Table 5).
Table 4.
Hazard Ratios for Mortality by Comorbidity among Beneficiaries with Heart Failure*
| COMORBIDITY | Adjusted Hazard Ratio (95% C.I.) | Interaction between hospitalization status and condition | |
|---|---|---|---|
| Cardiac Conditions | Model 1: Hospitalized Beneficiaries | Model 2: Non-hospitalized Beneficiaries | Model 3: P-Value |
| Myocardial infarction | 1.52 (1.40, 1.66) | 2.23 (1.82, 2.73) | 0.0025 |
| Ischemic Heart Disease | 1.08 (1.00, 1.17) | 1.10 (1.04, 1.17) | 0.5891 |
| Atrial fibrillation | 0.99 (0.94, 1.04) | 1.34 (1.25, 1.43) | < 0.0001 |
| Noncardiac Conditions | Model 1: Hospitalized Beneficiaries | Model 2: Non-hospitalized Beneficiaries | Model 3: P-Value |
| Lung Cancer | 1.93 (1.57, 2.37) | 3.78 (2.86, 4.98) | 0.0075 |
| Endometrial Cancer | 1.73 (0.67, 4.52) | 2.11 (1.11, 4.00) | 0.7282 |
| CKD† | 1.62 (1.54, 1.70) | 1.98 (1.83, 2.14) | 0.0021 |
| Dementia | 1.50 (1.43, 1.60) | 2.19 (2.05, 2.34) | < 0.0001 |
| Hip Fracture | 1.33 (1.16, 1.51) | 1.93 (1.60, 2.33) | 0.0003 |
| Colorectal Cancer | 1.29 (1.06, 1.56) | 1.27 (0.98, 1.65) | 0.978 |
| COPD† | 1.28 (1.22, 1.34) | 1.90 (1.77, 2.03) | < 0.0001 |
| Stroke | 1.21 (1.13, 1.29) | 1.64 (1.50, 1.80) | < 0.0001 |
| Depression | 1.18 (1.11, 1.25) | 1.51 (1.40, 1.64) | < 0.0001 |
| Diabetes | 1.16 (1.11, 1.22) | 1.32 (1.24, 1.41) | 0.25 |
| Breast Cancer | 1.12 (0.90, 1.41) | 1.17 (0.94, 1.46) | 0.8548 |
| Prostate Cancer | 1.01 (0.89, 1.15) | 1.06 (0.92, 1.24) | 0.8823 |
| Osteoporosis | 0.97 (0.90, 1.03) | 1.02 (0.94, 1.12) | 0.068 |
| Arthritis | 0.88 (0.84, 0.93) | 0.97 (0.91, 1.04) | 0.0008 |
| Glaucoma | 0.79 (0.73, 0.86) | 0.84 (0.77, 0.92) | 0.0793 |
| Cataracts | 0.73 (0.69, 0.77) | 0.78 (0.73, 0.83) | 0.1456 |
*Adjusted for age, gender and race
†CKD = Chronic kidney disease; COPD = Chronic obstructive pulmonary disease
Table 5.
Adjusted Hazard Ratios for Mortality by Comorbidity among Beneficiaries with Heart Failure*
| COMORBIDITY | Adjusted HR (95% C.I.) | |
|---|---|---|
| Cardiac Conditions | Model 1: Hospitalized Beneficiaries | Model 2:Non-hospitalized Beneficiaries |
| Myocardial Infarction | 1.44 (1.32, 1.57) | 1.64 (1.31, 2.04) |
| Ischemic Heart Disease | 1.00 (0.93, 1.09) | 0.94 (0.88, 1.00) |
| Atrial Fibrillation | - | 1.30 (1.22, 1.40) |
| Noncardiac Conditions | Model 1: Hospitalized Beneficiaries | Model 2:Non-hospitalized Beneficiaries |
| Lung Cancer | 1.86 (1.52, 2.28) | 3.58 (2.63, 4.89) |
| CKD† | 1.57 (1.49, 1.65) | 1.77 (1.63, 1.92) |
| Dementia | 1.44 (1.36, 1.52) | 1.91 (1.77, 2.05) |
| Colorectal Cancer | 1.39 (1.16, 1.67) | - |
| COPD† | 1.24 (1.19, 1.31) | 1.70 (1.58, 1.82) |
| Hip Fracture | 1.24 (1.08, 1.42) | 1.44 (1.17, 1.76) |
| Stroke | 1.10 (1.02, 1.18) | 1.23 (1.11, 1.36) |
| Depression | 1.06 (0.99, 1.12) | 1.13 (1.04, 1.24) |
| Diabetes | 1.04 (0.99, 1.09) | 1.22 (1.15, 1.31) |
| Arthritis | 0.87 (0,83, 0.92) | - |
| Glaucoma | 0.79 (0.73, 0.86) | 0.87 (0.79, 0.95) |
| Cataracts | 0.74 (0.70, 0.78) | 0.81 (0.75, 0.87) |
| Endometrial Cancer | - | 1.07 (0.45, 2.55) |
*Adjusted for age, gender, race, and other comorbid conditions. Cells with no results indicate non-significant association with mortality in analyses unadjusted for the presence of other conditions (See Table 4)
†CKD = Chronic kidney disease; COPD = Chronic obstructive pulmonary disease
Among hospitalized beneficiaries, CKD, COPD, and dementia conferred the highest mortality risk and were highly prevalent. Median survival was <1 year for hospitalized beneficiaries with CKD and dementia and <2 years for beneficiaries with other combinations of conditions (Fig. 1). In contrast, median survival for hospitalized beneficiaries without any of the three comorbidities was as high as 4.8 years for beneficiaries aged 66–75 years.
Figure 1.
Age-stratified median survival times by combinations of comorbid conditions among hospitalized beneficiaries.
DISCUSSION
In this study of Medicare beneficiaries with HF, we used a single HF hospitalization to identify beneficiaries at a high risk of mortality compared to non-hospitalized beneficiaries. Both cardiac and noncardiac comorbidities were significantly more prevalent among hospitalized beneficiaries than among non-hospitalized beneficiaries. While the relative mortality risk associated with comorbidity was even higher among non-hospitalized beneficiaries, both the overall burden of comorbidity and the majority of individual comorbidities were associated with increased mortality risk among hospitalized beneficiaries. The presence of CKD, dementia, myocardial infarction, and lung cancer each conferred ≥ 50% increase in risk among hospitalized beneficiaries. Certain combinations of highly prevalent noncardiac comorbidities (e.g.: CKD and dementia) were associated with markedly reduced survival.
Our findings confirm the results of prior studies13–17 that hospitalization for HF identifies patients at high mortality risk. Multiple factors may contribute both to hospitalization and mortality risk in addition to severity of HF, including, socioeconomic status, psychosocial support, and preferences for treatment. If these factors were completely independent of HF status, it might be expected that the relative effect of comorbidity would be similar among both groups of beneficiaries. However, the finding that the relative association between comorbidity and mortality was lower among hospitalized beneficiaries than non-hospitalized beneficiaries suggests that a more advanced stage of HF was driving mortality in this group. Nonetheless, comorbidity continued to contribute to mortality risk among hospitalized beneficiaries.
Our findings support the conclusion that comorbidity continues to confer a significant risk of mortality among persons with advanced HF. This conclusion stands in contrast to the relationship between comorbidity and mortality among patients with cancer, whereby the mortality risk associated with comorbidity in the early stages of cancer drops markedly and becomes non-significant in the advanced stages9–11. Several mechanisms may be responsible for the persistent deleterious effects of comorbidities in advanced HF. Conditions that share pathophysiology or risk factors with HF may contribute to mortality by interacting with and exacerbating the course of HF20. Conditions unrelated to HF may confer a competing effect on mortality because of the variable trajectory of HF. Alternatively, there is growing evidence that conditions traditionally considered to be unrelated to HF, such as depression21 and cognitive impairment22, may be pathophysiologically related to HF. Further examination of the effect of comorbidity on mortality in the advanced stages of these diseases, and the mechanisms by which comorbidity affects mortality, is warranted.
The finding that comorbidity confers a significant mortality risk among older adults hospitalized with HF highlights the importance of using comorbidity information to guide treatment decision-making for this population. Patients in the advanced stages of HF are candidates for intensive interventions aimed at reducing HF mortality3. The increased mortality risk associated with comorbidity suggests that patients with particular comorbidities may derive only a limited survival benefit from such interventions. Although HF guidelines acknowledge the need to consider comorbidity in the treatment decision-making process for persons with advanced disease, little empiric evidence exists to help guide providers in using comorbidity information to inform decision-making. A recent study modeling the maximum benefit of implantable cardiac defibrillators (ICDs) found that CKD and dementia identified patients with particularly poor survival even when assuming all sudden cardiac deaths to be preventable by an ICD23. We extend these findings among a population-based cohort by identifying comorbidities that are highly prevalent, and when present in combination, contribute to markedly reduced survival among older adults with advanced HF. Future work examining if and how certain combinations of comorbidities might provide prognostically useful information to help guide care will be relevant.
Our analyses have several limitations. Because claims databases typically do not contain clinical information, we lacked data to characterize the severity and impact of comorbidity. While it is possible that beneficiaries with advanced HF also have more advanced comorbidities, an association between comorbidity severity and mortality would only strengthen our conclusion that comorbidity continues to confer a significant mortality risk in advanced HF. Although we were unable to measure psychosocial factors that may have influenced hospitalization, the finding that the relative mortality attributable to comorbidity was lower among hospitalized beneficiaries provides support that the stage of HF was driving mortality in this group rather than these unmeasured factors. Overall health status is not available through administrative data, and it is possible that comorbidity may have served as a proxy for better health. Specifically, we found that cataracts conferred a protective effect among both groups of beneficiaries; however, cataracts may have been coded when they were removed, and beneficiaries with better overall health may be more likely to be deemed suitable for cataract surgery. The use of administrative data for estimating the prevalence of conditions is at risk for underestimating the true burden of disease24. We believe that our use of the CCW, specifically designed to mitigate the risk of underestimation through the use of evidence-based algorithms to indicate the presence of a disease, facilitated the accurate reporting of disease prevalence in our sample. Because we defined conditions as present once the algorithm criteria were met we may have over-reported the prevalence of cataracts since cataracts can be cured. We were unable to distinguish between systolic and diastolic HF. As prior evidence indicates that some comorbidities confer greater mortality risk among patients with diastolic versus systolic HF25, our mortality estimates might be impacted by the proportions of patients in our study cohort who had either type of HF.
Our findings highlight the significant impact that comorbidity, particularly noncardiac disease, has on increasing mortality even among those who are hospitalized with HF and are at overall high mortality risk. To the extent that these patients represent individuals with advanced heart failure, these findings challenge the assumption that as HF progresses into advanced stages, it becomes the primary determinant of outcomes. A model of care that focuses primarily on managing HF may not be adequate for this population; instead, our findings support a shift towards an expanded care model that considers the impact of coexisting illnesses in increasing mortality, addressing quality of life, and determining appropriate treatment interventions, even among patients with advanced disease. Future research should examine the potential mechanisms by which comorbidity affects mortality in patients with HF – by contributing to the severity of HF or by providing a competing mortality risk – to help clinicians better target their care and treatment approach.
Acknowledgements
Dr. Ahluwalia was supported by a training grant from the National Institute on Aging (T32AG1934) and is currently supported by an Office of Academic Affiliation’s VA Associated Health Postdoctoral Fellowship Program at the VA Greater Los Angeles HSR&D Center of Excellence. Dr. Fried is a recipient of a Midcareer Investigator Award in Patient-Oriented Research from the National Institute on Aging (K24 AG028443). Dr. Chaudhry is the recipient of a Paul Beeson/K23 Career Development Award (K23AG030986) from the National Institute on Aging. Supported by the Claude D. Pepper Older Americans Independence Center at Yale University School of Medicine (#2P30AG021342-06 NIH/NIA).
Conflict of Interest
None disclosed.
Appendix
References
- 1.Marengoni A, Rizzuto D, Wang HX, Winblad B, Fratiglioni L. Patterns of chronic multimorbidity in the elderly population. J Am Geriatr Soc. 2009;57:225–30. doi: 10.1111/j.1532-5415.2008.02109.x. [DOI] [PubMed] [Google Scholar]
- 2.Boyd CM, Darer J, Boult C, Fried LP, Boult L, Wu AW. Clinical practice guidelines and quality of care for older patients with multiple comorbid diseases: implications for pay for performance. JAMA. 2005;294:716–24. doi: 10.1001/jama.294.6.716. [DOI] [PubMed] [Google Scholar]
- 3.Jessup M, Abraham WT, Casey DE, et al. 2009 focused update: ACCF/AHA Guidelines for the Diagnosis and Management of Heart Failure in Adults: a report of the American College of Cardiology Foundation/American Heart Association Task Force on Practice Guidelines: developed in collaboration with the International Society for Heart and Lung Transplantation. Circulation. 2009;119:1977–2016. doi: 10.1161/CIRCULATIONAHA.109.192064. [DOI] [PubMed] [Google Scholar]
- 4.Havranek EP, Masoudi FA, Westfall KA, Wolfe P, Ordin DL, Krumholz HM. Spectrum of heart failure in older patients: Results from the National Heart Failure Project. Am Heart J. 2001;143:412–7. doi: 10.1067/mhj.2002.120773. [DOI] [PubMed] [Google Scholar]
- 5.Rusinaru D, Saaidi I, Godard S, Mahjoub H, Battle C, Tribouilloy C. Impact of chronic obstructive pulmonary disease on long-term outcome of patients hospitalized for heart failure. Am J Cardiol. 2008;101:353–8. doi: 10.1016/j.amjcard.2007.08.046. [DOI] [PubMed] [Google Scholar]
- 6.Bruch C, Rothenburger M, Gotzmann M, et al. Chronic kidney disease in patients with chronic heart failure–impact on intracardiac conduction, diastolic function and prognosis. Int J Cardiol. 2007;118:375–80. doi: 10.1016/j.ijcard.2006.06.066. [DOI] [PubMed] [Google Scholar]
- 7.Corell P, Gustafsson F, Schou M, Markenvard J, Nielsen T, Hildebrandt P. Prevalence and prognostic significance of atrial fibrillation in outpatients with heart failure due to left ventricular systolic dysfunction. Eur J Heart Fail. 2007;9:258–65. doi: 10.1016/j.ejheart.2006.08.004. [DOI] [PubMed] [Google Scholar]
- 8.From AM, Leibson CL, Bursi F, et al. Diabetes in heart failure: prevalence and impact on outcome in the population. Am J Med. 2006;119:591–9. doi: 10.1016/j.amjmed.2006.05.024. [DOI] [PubMed] [Google Scholar]
- 9.Alho OP, Hannula K, Luokkala A, Teppo H, Koivunen P, Kantola S. Differential prognostic impact of comorbidity in head and neck cancer. Head Neck. 2007;29:913–8. doi: 10.1002/hed.20608. [DOI] [PubMed] [Google Scholar]
- 10.Tetsche MS, Dethlefsen C, Pedersen L, Sorensen HT, Norgaard M. The impact of comorbidity and stage on ovarian cancer mortality: a nationwide Danish cohort study. BMC Cancer. 2008;8:31. doi: 10.1186/1471-2407-8-31. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Read WL, Tierney RM, Page NC, et al. Differential prognostic impact of comorbidity. J Clin Oncol. 2004;22:3099–103. doi: 10.1200/JCO.2004.08.040. [DOI] [PubMed] [Google Scholar]
- 12.Metra M, Ponikowski P, Dickstein K, et al. Advanced chronic heart failure: A position statement from the Study Group on Advanced Heart Failure of the Heart Failure Association of the European Society of Cardiology. Eur J Heart Fail. 2007;9:684–94. doi: 10.1016/j.ejheart.2007.04.003. [DOI] [PubMed] [Google Scholar]
- 13.Rudiger A, Harjola VP, Muller A, et al. Acute heart failure: clinical presentation, one-year mortality and prognostic factors. Eur J Heart Fail. 2005;7:662–70. doi: 10.1016/j.ejheart.2005.01.014. [DOI] [PubMed] [Google Scholar]
- 14.Setoguchi S, Stevenson LW, Schneeweiss S. Repeated hospitalizations predict mortality in the community population with heart failure. Am Heart J. 2007;154:260–6. doi: 10.1016/j.ahj.2007.01.041. [DOI] [PubMed] [Google Scholar]
- 15.Dickstein K, Gleim GW, Snapinn S, James MK, Kjekshus J. The impact of morbid events on survival following hospitalization for complicated myocardial infarction. Eur J Heart Fail. 2006;8:74–80. doi: 10.1016/j.ejheart.2005.04.010. [DOI] [PubMed] [Google Scholar]
- 16.MacIntyre K, Capewell S, Stewart S, et al. Evidence of improving prognosis in heart failure: trends in case fatality in 66 547 patients hospitalized between 1986 and 1995. Circulation. 2000;102:1126–31. doi: 10.1161/01.cir.102.10.1126. [DOI] [PubMed] [Google Scholar]
- 17.McCullough PA, Philbin EF, Spertus JA, Kaatz S, Sandberg KR, Weaver WD. Confirmation of a heart failure epidemic: findings from the Resource Utilization Among Congestive Heart Failure (REACH) study. J Am Coll Cardiol. 2002;39:60–9. doi: 10.1016/S0735-1097(01)01700-4. [DOI] [PubMed] [Google Scholar]
- 18.Braunstein JB, Anderson GF, Gerstenblith G, et al. Noncardiac comorbidity increases preventable hospitalizations and mortality among Medicare beneficiaries with chronic heart failure. J Am Coll Cardiol. 2003;42:1226–33. doi: 10.1016/S0735-1097(03)00947-1. [DOI] [PubMed] [Google Scholar]
- 19.Lin DY, Wei LJ. The robust inference for the Cox proportional hazards model. J Am Stat Assoc. 1989;84:1074–78. [Google Scholar]
- 20.Silverberg DS, Wexler D, Iaina A, Schwartz D. The interaction between heart failure and other heart diseases, renal failure, and anemia. Semin Nephrol. 2006;26:296–306. doi: 10.1016/j.semnephrol.2006.05.006. [DOI] [PubMed] [Google Scholar]
- 21.Joynt KE, Whellan DJ, O'Connor CM. Why is depression bad for the failing heart?? A review of the mechanistic relationship between depression and heart failure. J Card Fail. 2004;10:258–271. doi: 10.1016/j.cardfail.2003.09.008. [DOI] [PubMed] [Google Scholar]
- 22.Cohen MB, Mather PJ. A review of the association between congestive heart failure and cognitive impairment. Am J Geriatr Cardiol. 2007;16:171–4. doi: 10.1111/j.1076-7460.2007.06563.x. [DOI] [PubMed] [Google Scholar]
- 23.Setoguchi S, Nohria A, Rassen JA, Stevenson LW, Schneeweiss S. Maximum potential benefit of implantable defibrillators in preventing sudden death after hospital admission because of heart failure. CMAJ. 2009;180:611–6. doi: 10.1503/cmaj.080769. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Klabunde CN, Warren JL, Legler JM. Assessing comorbidity using claims data: an overview. Med Care. 2002;40(IV):26–35. doi: 10.1097/00005650-200208001-00004. [DOI] [PubMed] [Google Scholar]
- 25.Ahmed A, Rich MW, Sanders PW, et al. Chronic kidney disease associated mortality in diastolic versus systolic heart failure: a propensity matched study. Am J Cardiol. 2007;99:393–8. doi: 10.1016/j.amjcard.2006.08.042. [DOI] [PMC free article] [PubMed] [Google Scholar]