Incidence and hospital death rates associated with heart failure: A community-wide perspective
Article Outline
Abstract
Purpose
Despite often stated references to the emerging epidemic of heart failure in the United States, relatively little data are available describing the incidence and short-term death rates associated with this clinical syndrome. The objectives of this study were to describe the hospital incidence and death rates associated with acute heart failure and factors associated with an adverse hospital prognosis in residents of the Worcester, Mass, metropolitan area hospitalized at all greater Worcester medical centers with new onset heart failure in 2000.
Subjects and methods
We reviewed the medical records of patients hospitalized for acute heart failure at all 11 area medical centers during 2000. New onset heart failure was diagnosed using standardized criteria. Regression analyses were performed to examine demographic and clinical factors associated with hospital death rates.
Results
During 2000, 2604 men and women from greater Worcester were diagnosed with new onset heart failure; 637 (24.5%) of these cases were initial events. The incidence and attack rates (per 100
000) of heart failure were 219 and 897, respectively. Occurrence of heart failure increased with advancing age, and women were at greater risk for heart failure than men (incidence rates [per 100 000] = 250 and 194, respectively). Hospital case-fatality rates were 5.1%. Hospital death rates were associated with several demographic and clinical characteristics.
Conclusions
The results of this study suggest that heart failure is an important clinical syndrome affecting residents of this large northeast community. Several groups at high risk for developing or dying from heart failure can be identified and targeted for preventive efforts as well as for the receipt of effective treatment modalities.
Keywords: Heart failure , Incidence rates , Hospital death rates
Although heart failure is often referred to as one of the new “epidemics” of cardiovascular disease in the 21st century,1, 2, 3 a paucity of data exists about the magnitude and short-term outcomes associated with this clinical syndrome, particularly from the more generalizable perspective of a population-based investigation.
The purpose of the present study was to describe the incidence and attack rates of heart failure among residents from a large northeast metropolitan area, hospital death rates, and factors associated with a poor short-term prognosis in patients hospitalized with new onset heart failure during the year 2000.
Methods
Residents of the Worcester, Mass, metropolitan area (2000 census estimate = 478 000) hospitalized for possible heart failure at all 11 greater Worcester medical centers during 2000 comprised the study sample. The medical records of patients with primary or secondary (any position but first) discharge diagnoses consistent with the possible presence of heart failure were reviewed. The primary International Classification of Diseases, Ninth Edition (ICD)-9 code reviewed for the identification of cases of acute heart failure was ICD code 428. Additional ICD-9 diagnostic rubrics reviewed for the possible presence of heart failure included rheumatic heart failure (398.91), hypertensive heart disease (402), hypertensive heart and renal disease (404), acute cor pulmonale (415), other diseases of the endocardium (424), primary cardiomyopathies (425.4), pulmonary heart disease and congestion (416.9 and 514), acute lung edema (518.4), edema (782.3), and dyspnea and respiratory abnormalities (786). Confirmation of the diagnosis of heart failure, based on use of the Framingham criteria, included the presence of 2 major criteria or 1 major and 2 minor criteria.4, 5, 6
Because 1 of the objectives of this study was to identify incident cases of acute heart failure occurring in greater Worcester residents, medical records of previous hospitalizations or outpatient visits for heart failure were reviewed. An incident event of heart failure was defined as the absence of a prior hospitalization for heart failure, a physician diagnosis of heart failure, or treatment for heart failure in the past based on the review of data contained in hospital medical records. Patients who developed heart failure secondary to admission for another illness (eg, acute myocardial infarction) or after an interventional procedure (eg, coronary artery bypass surgery), were not included.
Data collection
Information was collected about demographic, medical history, laboratory and physiologic measures, and clinical characteristics of the study sample through the review of information contained in hospital medical records. This included information about patients’ age, sex, race, prior comorbidities (including heart failure), body mass index, admission levels of blood pressure, serum sodium, potassium, creatinine, and blood urea nitrogen, and presenting symptoms.
Data analysis
Incidence and attack rates (initial and prior episodes of acute heart failure) of heart failure were calculated in a standard manner using available published census data. Age- as well as sex-specific incidence and attack rates were calculated based on US census estimates of the greater Worcester adult population (≥25 years) in 2000.
Results
Descriptive characteristics
A total of 2604 adult men and women of all ages from the Worcester metropolitan area with independently confirmed decompensated heart failure requiring hospital admissions comprised the study population. Of these, 637 patients (24.5%) experienced a first episode of heart failure; the remainder of the study sample consisted of patients with prior heart failure in whom a new episode of decompensated heart failure occurred during hospitalization in 2000. The median age of the study sample was 79 years. The majority of patients were women, white, and had a history of heart failure, coronary heart disease, and hypertension. Dyspnea and edema were the most commonly reported symptoms by patients at the time of hospital presentation (Table 1). Similar characteristics were noted in the subset of patients experiencing their first episode of heart failure.
Table 1. Descriptive characteristics of study population: Worcester, Mass, 2000
| Characteristic | All patients (n = 2604) | Incident events (n = 637) |
|---|---|---|
| Age (median, years) | 79.0 | 78.0 |
| Women (%) | 56.7 | 57.5 |
| White race (%) | 92.6 | 91.8 |
| Body mass index (%) | ||
| 25–29.9 | 27.4 | 30.5 |
| ≥30 | 29.7 | 28.1 |
| Duration of prehospital delay (median, h) | 2.0 | 4.0 |
| Medical history (%) | ||
| Anemia | 24.5 | 15.5 |
| Chronic lung disease | 34.3 | 23.2 |
| Coronary heart disease | 56.6 | 33.9 |
| Diabetes mellitus | 26.3 | 18.2 |
| Heart failure | 75.5 | - |
| Hypertension | 67.4 | 62.5 |
| Renal failure/disease | 24.5 | 11.6 |
| Stroke | 14.9 | 10.8 |
| Laboratory/physiologic measures | ||
| Serum creatinine (mean, mg/dL) | 1.6 | 1.4 |
| Serum blood urea nitrogen (mean, mg/dL) | 35.2 | 27.6 |
| Serum sodium (mean) | 136.9 | 136.7 |
| Systolic blood pressure (mean, mm Hg) | 143.3 | 150.6 |
| Diastolic blood pressure (mean, mm Hg) | 74.9 | 78.8 |
| Heart rate (mean, beats per minute) | 89.4 | 92.4 |
| Length of hospital stay (median, days) | 4.0 | 4.0 |
| Symptoms (at time of hospital presentation) (%) | ||
| Chest pain/discomfort | 30.3 | 30.6 |
| Dyspnea | 93.1 | 93.1 |
| Edema | 70.2 | 62.6 |
| Weight gain | 8.9 | 6.1 |
| Generalized weakness | 28.4 | 27.6 |
| Nausea | 15.0 | 15.4 |
| Orthopnea | 36.5 | 33.8 |
Attack rates of heart failure
The overall, age-, and sex-specific incidence rates of heart failure are shown in Table 2. The incidence and attack rates (per 100000 population) of heart failure were 219 and 897, respectively. The occurrence of heart failure in the greater Worcester population increased markedly with the aging of this population. Women experienced higher incidence (250 vs 194) as well as attack rates (976 vs 811) of heart failure than men (Table 2).
Table 2. Attack rates⁎ of heart failure: Worcester, Mass, 2000
| Rate⁎ | Incidence rates | Attack rates |
|---|---|---|
| Overall | 219(203,237)† | 897(863,932)† |
| Age-specific (years) | ||
| <55 | 31(24,39) | 90(77,103) |
| 55–64 | 181(136,225) | 653(569,738) |
| 65–74 | 423(347,500) | 1871(1709,2032) |
| 75–84 | 1100(961,1238) | 4543(4262,4824) |
| ≥85 | 1820(1529,2110) | 8158(7543,8772) |
| Sex-specific | ||
| Men | 194(172,219) | 811(765,860) |
| Women | 250(226,278) | 976(927,1027) |
⁎
Per 100 |
† 95% confidence intervals. |
In each of age strata examined, men experienced higher incidence and attack rates of heart failure than women (Figure). The overall incidence and attack rates of heart failure, however, were higher in women than in men because of the greater number and proportion of women in the oldest age groupings.
Figure.
Age- and sex-specific incidence rates of heart failure: Worcester, Mass, 2000.
Hospital death rates associated with acute heart failure
In the total study sample, 5.1% of patients died during the acute hospitalization. Among patients with a first documented episode of heart failure, 5.3% of patients died, whereas 5.0% of patients with a previous episode of heart failure died.
In examining factors associated with hospital case-fatality rates, patients dying during the acute hospitalization were more likely to be older, of desirable body weight, and have a history of anemia or stroke (Table 3). Patients with higher concentrations of blood urea nitrogen, serum creatinine, and admission heart rates were more likely to die than respective comparison groups. On the other hand, patients with lower serum sodium concentrations and lower systolic and diastolic blood pressure values were at greater risk for dying during hospitalization than patients with higher levels of these clinical variables (Table 3).
Table 3. Hospital case-fatality rates (CFR) in patients with heart failure according to selected characteristics
| Characteristic | Total study sample | ||
|---|---|---|---|
| n | CFR (%) | Adjusted OR (95% CI) | |
| Age (years) | |||
| <55 | 175 | 1.7 | 1.0 |
| 55–64 | 228 | 2.6 | 1.54(0.37,6.42) |
| 65–74 | 515 | 2.5 | 1.14(0.31,4.17) |
| 75–84 | 1003 | 7.0 | 3.19(0.95,10.66) |
| ≥85 | 674 | 6.1 | 2.38(0.69,8.20) |
| Sex | |||
| Male | 1125 | 4.9 | 0.94(0.63,1.40) |
| Female | 1471 | 5.3 | 1.0 |
| Body mass index | |||
| <25 | 926 | 5.5 | 1.0 |
| 25–29.9 | 592 | 4.6 | 0.82(0.52,1.32) |
| ≥30 | 638 | 3.3 | 0.73(0.42,1.25) |
| Blood urea nitrogen (mg/dL) | |||
| <20 | 715 | 3.4 | 1.0 |
| 20–24 | 385 | 3.1 | 0.75(0.36,1.56) |
| 25–44 | 837 | 5.4 | 1.25(0.69,2.25) |
| ≥45 | 633 | 7.7 | 2.13(1.05,4.34) |
| Creatinine (mg/dL) | |||
| 1.0 | 584 | 3.9 | 1.0 |
| 1.0–1.1 | 509 | 3.7 | 0.84(0.44,1.59) |
| 1.2–1.7 | 782 | 6.0 | 1.16(0.63,2.15) |
| ≥1.8 | 695 | 5.9 | 0.96(0.44,2.10) |
| Sodium (mg/dL) | |||
| <135 | 648 | 6.8 | 1.0 |
| 135–136 | 419 | 4.8 | 0.68(0.39,1.20) |
| 137–139 | 749 | 3.6 | 0.49(0.30,0.81) |
| ≥140 | 763 | 5.1 | 0.68(0.43,1.07) |
| Systolic blood pressure (mm Hg) | |||
| <120 | 591 | 7.3 | 1.0 |
| 120–139 | 630 | 4.9 | 0.91(0.55,1.52) |
| 140–159 | 636 | 5.4 | 1.03(0.61,1.75) |
| ≥160 | 732 | 3.4 | 0.75(0.40,1.40) |
| Diastolic blood pressure (mm Hg) | |||
| <60 | 496 | 9.3 | 1.0 |
| 60–74 | 824 | 5.2 | 0.73(0.46,1.15) |
| 75–84 | 490 | 2.9 | 0.38(0.20,0.73) |
| ≥85 | 705 | 4.0 | 0.61(0.34,1.12) |
| Heart rate (beats per minute) | |||
| <70 | 475 | 4.2 | 1.0 |
| 70–84 | 711 | 5.5 | 1.48(0.83,2.62) |
| 85–99 | 614 | 3.9 | 1.07(0.57,2.02) |
| ≥100 | 794 | 6.3 | 1.84(1.03,3.29) |
| Medical history | |||
| Anemia | 635 | 5.8 | 1.00(0.66,1.54) |
| Stroke | 386 | 5.7 | 1.07(0.65,1.76) |
| Renal failure | 635 | 5.0 | 0.89(0.53,1.52) |
| Coronary heart disease | 1468 | 4.1 | 0.66(0.46,0.97) |
| Chronic lung disease | 891 | 5.1 | 1.03(0.70,1.52) |
| Diabetes | 686 | 3.8 | 0.81(0.51,1.29) |
| Hypertension | 1751 | 4.7 | 0.86(0.58,1.27) |
| Presenting symptoms | |||
| Chest pain/discomfort | 786 | 2.8 | 0.58(0.36,0.95) |
| Weight gain | 232 | 1.7 | 0.35(0.13,0.98) |
| Dyspnea | 2418 | 5.1 | 1.59(0.79,3.21) |
| Nausea | 389 | 3.3 | 0.62(0.34,1.14) |
| Orthopnea | 949 | 3.2 | 0.58(0.38,0.90) |
| Edema | 1822 | 5.1 | 1.19(0.79,1.80) |
| Weakness | 735 | 6.5 | 1.10(0.74,1.62) |
| C statistic | 0.76 | ||
We carried out a logistic regression analysis to more systematically examine factors associated with an increased risk of dying during hospitalization while controlling for a variety of potentially confounding prognostic factors (Table 3). The results of this analysis suggested that patients 75 years and older were more likely to die during hospitalization than younger patients. Patients who were overweight, patients with prior coronary disease, and patients complaining of selected symptoms (eg, chest pain, weight gain, nausea, orthopnea) were less likely to die during hospitalization than respective comparison groups. Trends similar to the univariable findings were observed when we examined the association between various physiologic (eg, blood pressure) and laboratory variables (eg, serum creatinine and sodium) to the risk of dying during hospitalization for heart failure. Higher admission serum blood urea nitrogen levels (adjusted odds ratio [OR] 1.02; 95% confidence interval [CI] 1.01, 1.03), lower levels of diastolic blood pressure (OR 0.98; 95% CI 0.97, 0.99), and higher admission heart rate findings (OR 1.01; 95% CI 1.01, 1.02) were also significantly related to hospital death rates when expressed as continuous variables. Relatively similar findings of the prognostic importance of most previously noted factors in the total study sample were observed when we restricted our analysis to patients with a first episode of heart failure. In these patients, however, more attenuated associations were observed due to the smaller number of patients included and fewer hospital deaths that occurred.
Discussion
The results of the present study, carried out in residents from a large northeast metropolitan area hospitalized with a confirmed new episode of acute heart failure during 2000, suggest considerable population burden from this clinical syndrome. Our findings also provide insights to the short-term mortality associated with this clinical syndrome and to the profile of individuals at increased risk of dying from heart failure during their acute hospitalization. The lack of representative and contemporary data in this area, using standardized ascertainment procedures and validation of possible cases of heart failure, argue for the importance of population-based registries for the more systematic study of this clinical syndrome.7 These community-wide surveillance projects need to be developed and maintained for long periods of time to provide timely and ongoing insights to the effective primary and secondary prevention of this serious clinical problem.
Attack rates of heart failure
Despite the considerable population burden from heart failure and the availability of several effective disease-modifying medical therapies based on the results of secondary prevention trials, limited population-based data exist, particularly during recent years, describing the incidence rates of heart failure and whether differences in the attack rates of heart failure exist in different population groups. The few published rates vary. This is partially due to differences in the demographic and clinical characteristics of the samples under study, methodologic and case ascertainment approaches used, and definitions of heart failure utilized. The attack rates of heart failure are highly linked with advancing age.
The initial study of the descriptive epidemiology of heart failure was carried out nearly 40 years ago by Gibson and colleagues, who examined the prevalence of heart failure in whites residing in two rural communities in North Carolina and Vermont.8 Data from the Framingham Heart Study provide the most widely quoted estimates about the attack rates of heart failure. In this study, the annual incidence rates (per 1000 person years) of heart failure were 1.4 and 2.3 for women and men, respectively.4, 5, 6 Although both sexes showed an increased risk of developing heart failure with advancing age in the Framingham Study, the incidence rates of heart failure were generally higher for men than for women. More recent findings from the Framingham Heart Study suggest that, in 1075 men and women who developed heart failure over a nearly 50-year period of study (1950–1999), age-adjusted incidence rates of heart failure were essentially unchanged in men over time, whereas the incidence rate of heart failure decreased considerably among women during the most recent decade under study (1990–1999).9 Using diagnostic criteria similar to those of the Framingham Study, the age- and sex-adjusted incidence rates of heart failure did not change in residents of Olmsted County, Minn, between 1981 (3.0 per 1000 person years) and 1991 (2.8 per 1000 person years).10, 11 These estimates, however, were based on a total of only 248 cases of new onset heart failure that occurred over the 2 study years in this primarily white population. Similar to other studies, the incidence rates of heart failure increased with advancing age in male and female residents of Olmsted County.
Reflecting the ever-increasing clinical and public health burden of this condition, the number of hospital discharges associated with heart failure in the United States has increased substantially over time. These estimates, in part, reflect the aging of the US population, declining death rates from coronary heart disease, and increasing prevalence of persons with coronary heart disease. Between 1971 and 1998, nonvalidated hospitalizations for heart failure increased more than 3-fold with hospitalization rates increasing nearly 5-fold over this period for persons 65 years and older.1, 2 In 1979, there were a total of 377000 hospital discharges for heart failure, with this estimate increasing to 978000 in the late 1990s.1, 2 Data from the National Hospital Discharge Survey also indicate substantial increases in the number of hospitalizations for a first listed diagnosis of heart failure between 1985 (577000) and 1995 (871000) and an increase in the number of hospitalizations for individuals with any diagnosis of heart failure (1.7 to 2.6 million) over this period.12 On the other hand, hospital admission rates for patients with a primary discharge diagnosis of heart failure in all Oregon hospitals between 1991 and 1995 were stable.13 In elderly patients (aged ≥65 years) hospitalized for heart failure in Ontario between 1992 and 2000, the number of yearly admissions for heart failure was also relatively constant over time.14
Death rates associated with heart failure
Although patients discharged after hospitalization for heart failure have been shown to be at high risk for recurrent hospitalization, impaired quality of life, and long-term mortality, few data are available, particularly from the more generalizable perspective of a community-based study, about the hospital death rates associated with heart failure and factors associated with an increased short-term risk of dying. Most of the published literature describes patient’s long-term survival experience after heart failure, suggesting a continuing poor long-term prognosis in these high risk patients.10, 11
Despite major advances in the management of cardiovascular disease during the past decade and encouraging results of intervention trials in patients with heart failure treated with vasodilator therapy, angiotensin-converting enzyme inhibitors, and beta-blockers, limited data exist to determine whether the prognosis associated with heart failure has changed over time. A population-based study in Scotland suggested trends of improving 30-day prognosis in more than 66000 patients hospitalized with a principal (albeit unvalidated) diagnosis of heart failure throughout Scotland between 1986 and 1995.15 Among patients with a primary discharge diagnosis of heart failure in Oregon hospitals over the period 1991 through 1995, in-hospital death rates decreased from 6.9% in 1991 to 4.7% in 1995.13 In a study of nearly 6700 patients with a primary discharge diagnosis of heart failure receiving care at a single academic tertiary care center over a 10-year period, hospital death rates decreased from 8.4% in 1985–1987 to 6.1% in 1994–1996.16 In the ADHERE (Acute Decompensated Heart Failure National Registry) study of patients hospitalized with a primary diagnosis of heart failure in 263 hospitals throughout the United States between 2001 and 2003, hospital death rates were 4.2%.17 These short-term death rates closely mirror our findings in which 5.0% of patients admitted with heart failure died in the hospital. In the study of approximately 88 000 elderly patients hospitalized for heart failure in hospitals throughout Ontario over the period 1992–2000, hospital case-fatality rates were essentially identical in 1992 (12.6%) and 1999 (12.3%).14
Factors associated with hospital death rates after heart failure
The results of the present study suggest that older patients were more likely to die than younger patients. Older patients have consistently been shown to be at higher risk for dying over the short term than younger patients, whereas patients with various symptoms of acute heart failure may reflect those with a more decompensated state of left ventricular function. Heavier patients, those with prior coronary disease, and patients presenting to area hospitals with selected symptoms of heart failure were less likely to die than respective comparison groups. Although the findings between prognosis after heart failure and overweight may seem paradoxical, they are consistent with limited published data.18 In the ADHERE Registry, the risk of dying during hospitalization for acute heart failure was reasonably estimated on the basis of routinely available vital signs (eg, low systolic blood pressure) and laboratory data (eg, high levels of blood urea nitrogen and serum creatinine) obtained at the time of hospital admission.17 In the present study, these factors were also shown to be related to a poorer hospital prognosis after acute heart failure. These data provide insights to the profile of hospitalized patients at increased risk of dying after heart failure in whom targeted treatment efforts might be directed. In particular, patients with impaired renal function, lower levels of blood pressure at the time of hospital presentation, older patients, and those with various symptoms of acute heart failure should have increased surveillance directed toward them and increased use of disease modifying therapies such as angiotensin-converting enzyme inhibitors and beta blockers.
Study strengths and limitations
The strengths of the present investigation are its inclusion of all residents from a geographically defined metropolitan area hospitalized with an independently confirmed new episode of heart failure at all area medical centers and standardized review of hospital medical records. The limitations of this study include the small number of minorities hospitalized with heart failure and inability to distinguish patients with systolic from those with diastolic dysfunction.
Conclusions
The results of the present investigation confirm that acute heart failure represents a considerable burden to the health of the community, particularly in older individuals. Given the aging of the US population and declining national death rates due to cardiovascular disease placing these individuals at increased risk for heart failure during their latter years of life, increasing trends in the magnitude and morbidity of heart failure are likely to continue for the foreseeable future. These trends, as well as changes in the management of patients with acute and chronic manifestations of heart failure over time, warrant further monitoring, particularly in representative US communities.7
Acknowledgments
This research was made possible by the cooperation of the medical records, administration, and cardiology departments of participating hospitals in the Worcester metropolitan area and through funding support provided by the National Institutes of Health (RO1 HL69874).
References
- . 2005 Heart and Stroke Statistical Update . Dallas, Texas: American Heart Association; 2004;
- . Morbidity and Mortality Chartbook on Cardiovascular, Lung and Blood Diseases 2004 . Bethesda, Maryland: National Institutes of Health, National Heart, Lung and Blood Institute; 2004;
- . Heart Failure (Evaluation and Care of Patients with Left Ventricular Systolic Dysfunction) . Rockville, Maryland: Agency for Health Care Policy and Research, US Department of Health and Human Services; 1994; (Clinical Practice Guidelines #11, AHCPR Publication No. 94-0612)
- . The natural history of congestive heart failure (the Framingham study) . N Engl J Med. . 1971;285:1441–1446
- . Epidemiology of heart failure . Am Heart J. . 1991;121:951–957
- . The epidemiology of heart failure (the Framingham Study) . J Am Coll Cardiol. . 1993;22:6A–13A
- . Assessing the population burden from heart failure . Arch Intern Med. . 1999;159:15–17
- . The prevalence of congestive heart failure in two rural communities . J Chronic Dis. . 1966;19:141–152
- Long-term trends in the incidence of and survival with heart failure . N Engl J Med. . 2002;347:1397–1402
- The incidence and prevalence of congestive heart failure in Rochester, Minnesota . Mayo Clin Proc. . 1993;68:1143–1150
- Congestive heart failure in the community (trends in incidence and survival in a 10-year period) . Arch Intern Med. . 1999;159:29–34
- . Hospitalization of patients with heart failure (National Hospital Discharge Survey, 1985 to 1995) . Am Heart J. . 1999;137:352–360
- . Analysis of trends in hospitalizations for heart failure . J Card Fail. . 1999;5:79–84
- Trends in heart failure outcomes and pharmacotherapy (1992 to 2000) . Am J Med. . 2004;116:581–589
-
Evidence of improving prognosis in heart failure. Trends in case fatality in 66547 patients hospitalized between 1986 and 1995
.
Circulation
. 2000;102:1126–1131
- . Ten year trends in hospital care for congestive heart failure . Arch Intern Med. . 2000;160:325–332
- Risk stratification for in-hospital mortality in acutely decompensated heart failure. Classification and regression tree analysis . JAMA . 2005;293:572–580
- The Obesity Paradox. Body mass index and outcomes in patients with heart failure . Arch Intern Med. . 2005;165:55–61
PII: S0002-9343(05)00315-3
doi:10.1016/j.amjmed.2005.04.013
© 2005 Elsevier Inc. All rights reserved.
