National Trends in Heart Failure
Hospital Stay Rates, 2001 to 2009
Jersey Chen, MD, MPH,*† Kumar Dharmarajan, MD, MBA,†‡ Yongfei Wang, MS,†
Harlan M. Krumholz, MD, SM†§
New Haven, Connecticut; and New York, New York
This study sought to analyze recent trends over time in heart failure (HF) hospital stay rates, length of stay (LOS),
and in-hospital mortality by age groups with a large national dataset of U.S. hospital discharges.
Heart failure hospital stay rates, LOS, and mortality have fallen over the past decade for older Medicare benefi-
ciaries, but whether this holds true for younger adults is unknown.
From the National Inpatient Sample, we calculated HF hospital stay rates, LOS, and in-hospital mortality from
2001 to 2009 with survey data analysis techniques.
Hospital stays (n ? 1,686,089) with a primary discharge diagnosis of HF were identified from National Inpatient
Sample data between 2001 and 2009. The overall national hospital stay rate decreased from 633 to 463 hospi-
tal stays/100,000 persons, (?26.9%, p-for-trend ?0.001). However, statistically significant declines (p ? 0.001)
were only observed for patients 55 to 64 years of age (?36.5%) 65 to 74 years (?37.4%), and ?75 years
(?28.3%) but not for patients 18 to 44 years of age (?12.8%, p ? 0.57) or 45 to 55 years (?16.2%, p ? 0.04).
Statistically significant declines in LOS were only observed for patients 65 years of age and older. Overall
in-hospital mortality fell from 4.5% to 3.3%, a relative decline of ?27.4%, (p-for-trend ?0.001), but patients 18
to 44 years of age did not exhibit a significant decline (?8.1%, p-for-trend ? 0.18). In secondary analyses signifi-
cant declines in HF hospital stay rate over time were observed for white men, white women, and black women
but not for black men (?9.5%, p-for-trend ? 0.43).
Younger patients have not experienced comparable declines in HF hospital stay, LOS, and in-hospital mortality
as older patients. Black men remain a vulnerable population for HF hospital stay.
1078–88) © 2013 by the American College of Cardiology Foundation
(J Am Coll Cardiol 2013;61:
The heart failure (HF) hospital stay rate has dropped
substantially over the past decade in the Medicare popula-
tion (1)—nearly 30% from 1998 to 2008—implying some
success in preventative efforts. However, the etiology of HF
often differs between younger adults and older Medicare
beneficiaries. Hypertension is the most common etiology of
HF in younger adults (2), whereas coronary artery disease
becomes a more common risk factor for HF among middle-
aged and older patients (3,4). In addition, as age increases,
the prevalence of HF with preserved ejection fraction rises
dramatically, in conjunction with comorbid risk factors such
as hypertension, atrial fibrillation, diabetes mellitus, and
renal insufficiency (3,5,6). As such, declines in the HF
hospital stay rate observed for older Medicare patients
might not necessarily indicate a corresponding decrease for
younger populations. Whether HF hospital stay rates differ
across age groups in the U.S. population is unknown and
the focus of this study.
In addition, we have a limited understanding of how HF
hospital stay rates have changed over time across race-sex
groups. One study of the Medicare population found that
HF hospital stay rates declined at a slower rate for black
From the *Kaiser Permanente, Mid-Atlantic Permanente Research Institute, Rock-
ville, Maryland; †Center for Outcomes Research and Evaluation, Yale-New Haven
Hospital, New Haven, Connecticut; ‡Division of Cardiology, Columbia University
Medical Center, New York, New York; and the §Section of Cardiovascular Medicine,
Department of Medicine; Robert Wood Johnson Clinical Scholars Program, Depart-
ment of Medicine, and the Section of Health Policy and Administration, School of
Public Health, Yale University School of Medicine, New Haven, Connecticut. Dr.
Chen was affiliated with the Section of Cardiovascular Medicine, Department of
Medicine, Yale University School of Medicine, during the time this research was
conducted. This work was supported by an Agency for Healthcare Research and
Quality Career Development Award (1K08HS018781-01) for Dr. Chen. Drs.
Krumholz, Chen, and Wang are supported by Grant U01 HL105270-02 (Center for
Cardiovascular Outcomes Research at Yale University) from the National Heart,
Lung, and Blood Institute. Dr. Dharmarajan is supported by Grant HL007854 from
the National Heart, Lung, and Blood Institute; he is also supported as a Centers of
Excellence Scholar in Geriatric Medicine at Yale by the John A. Hartford Foundation
and the American Federation for Aging Research. Dr. Krumholz chairs a cardiac
scientific advisory board for UnitedHealth and is the recipient of a research grant from
Medtronic, Inc., through Yale University. All other authors have reported that they
have no relationships relevant to the contents of this paper to disclose.
Manuscript received August 3, 2012; revised manuscript received November 8,
2012, accepted November 28, 2012.
Journal of the American College of Cardiology
© 2013 by the American College of Cardiology Foundation
Published by Elsevier Inc.
Vol. 61, No. 10, 2013
men compared with other groups (1). However, black
patients are more likely to develop HF at younger ages than
white patients (2), and whether this affects differences across
race with respect to declines in HF hospital stays is
unknown. Furthermore, black patients represent a higher
proportion of the uninsured and Medicaid enrollees (7),
which might potentially lead to differences in HF hospi-
tal stays across race compared with studies examining
Medicare data alone. As such, confirming whether black
men had slower declines in HF hospital stay rate in the
general population is important, because it might indicate
that this group is a particularly vulnerable population that
would benefit from targeted preventative efforts against
HF risk factors (8,9).
Accordingly, we analyzed data from the National Inpa-
tient Sample (NIS), a large national dataset of acute care
hospital stays that includes all age groups and all types of
health insurance coverage to examine changes across patient
age categories in HF hospital stay rates, length of stay
(LOS), and in-hospital mortality between 2001 and 2009.
Secondary analyses examined trends in HF hospital stay by
Data. The NIS, collected by the Agency for Healthcare
Research and Quality Healthcare Cost and Utilization
Project, is the largest all-payer inpatient database publicly
available in the United States. Consisting of discharge data
from over 1,000 hospitals across 44 states, the NIS was
designed to approximate a 20% stratified sample of U.S.
community hospitals (10). Statistical sampling weights pro-
vided by the NIS allow extrapolation to calculate expected
hospital stay rates for the nation (11). The NIS data were
collected on all patients, regardless of health insurance
provider. The following NIS fields were used for this
analysis: patient age, sex, race, principal and secondary
diagnosis codes, admission date, discharge date, in-hospital
death, insurance status, and state of hospital stay. Secondary
analyses stratified by race-sex categories were conducted in
a subset of patients hospitalized in states that reported
complete data on patient race across all years to the NIS.
Study cohort. A total of 71,371,439 hospital discharges
were reported to the NIS from 2001 to 2009 from 44 states
reporting data to NIS. We excluded the following hospital
stays: discharges in which patient age was ?18 years (n ?
12,091,363); those with missing data on patient age, sex,
admission date, discharge date, or in-hospital death (n ?
252,617); discharges in which patients were admitted and
discharged alive the same day, because such events might
not have truly represented hospital stays for acute conditions
(n ? 1,0373,079); and discharges from states that did not
report data for each year of the study period (n ?
9,648,462). The HF hospital stays were classified as those
with a principal discharge diagnosis of HF on the basis of
the following International Classification of Diseases-9th
(ICD-9-CM) codes: 402.01;
402.11; 402.91; 404.01; 404.03;
404.11; 404.13; 404.91; 404.93;
a pre-specified analysis calculating
population-based HF hospital stay
rates/100,000 persons for each cal-
endar year, with the numerator
representing the number of HF
representing the population 18 years of age and older from U.S.
Census estimates for each state (12). Survey analysis methods
were employed that used hospital-level discharge weights
provided by the NIS to estimate the number of HF hospital
stays on a national level (13). The HF hospital stay rates
were calculated for the overall cohort and for subgroups of
age (18 to 45, 45 to 54, 55 to 64, 65 to 74, and ?75 years),
sex, and, insurance status (Medicare, Medicaid, private
insurance [including health maintenance organizations],
self-pay, no charge, and other). Because a denominator
population of individuals specifying particular types of
health insurance could not be constructed for each year,
differences in HF hospital stay rate by insurance status were
not estimated. Differences in LOS and in-hospital mortality
rates were able to be evaluated by insurance status, because
calculation of these rates used the hospital stay as the unit of
analysis. Comorbidities were identified from ICD-9-CM
secondary diagnosis codes and were classified according to
hierarchical condition categories, similar to those used by
the Centers for Medicare and Medicaid Services for calcu-
lating their 30-day HF mortality measure (14). To evaluate
whether HF hospital stay rates declined faster than the
overall any-cause hospital stay rate, we calculated the
hospital stay rate/100,000 persons for all principal diagnoses
over the study period.
In-hospital survival curves were constructed that assumed
that the discharge date recorded in NIS was the date of
death, where the denominator represented the number of
patients still hospitalized on a given hospital day, and the
numerator represented the number of patients who were not
recorded as having an in-hospital death on that hospital day.
Survival curves were generated for 3-year periods (2001 to
2003, 2004 to 2006, and 2007 to 2009) to examine how
aggregate in-hospital mortality changed over the study
Statistical significance of the annual changes in HF
hospital stay rate and in-hospital mortality were assessed
with Poisson regression that included a variable represent-
ing the linear trend from the baseline year of 2001; a similar
analysis was conducted for LOS with linear regression. All
p values were 2-sided with a significance threshold of p ?
0.001. Statistical analyses were performed with SAS (ver-
sion 9.2, SAS Institute, Cary, North Carolina).
HF ? heart failure
ICD-9-CM ? International
Classification of Diseases,
9th revision, Clinical
LOS ? length of stay
NIS ? National Inpatient
JACC Vol. 61, No. 10, 2013
March 12, 2013:1078–88
Chen et al.
Trends in Heart Failure Hospital Stay
35. Bahrami H, Kronmal R, Bluemke DA, et al. Differences in the
incidence of congestive heart failure by ethnicity: the multi-ethnic
study of atherosclerosis. Arch Intern Med 2008;168:2138–45.
36. Boyle JP, Honeycutt AA, Narayan KM, et al. Projection of diabetes
burden through 2050: impact of changing demography and disease
prevalence in the U.S. Diabetes Care 2001;24:1936–40.
37. Flegal KM, Carroll MD, Ogden CL, Curtin LR. Prevalence and trends
in obesity among US adults, 1999–2008. JAMA 2010;303:235–41.
38. Blanchard J, Lurie N. Preventive care in the United States: are blacks
finally catching up? Ethn Dis 2005;15:498–504.
39. Victor RG, Leonard D, Hess P, et al. Factors associated with
hypertension awareness, treatment, and control in Dallas County,
Texas. Arch Intern Med 2008;168:1285–93.
40. Gordon HS, Nowlin PR, Maynard D, Berbaum ML, Deswal A.
Mortality after hospitalization for heart failure in blacks compared to
whites. Am J Cardiol 2010;105:694–700.
41. Rathore SS, Foody JM, Wang Y, et al. Race, quality of care, and
outcomes of elderly patients hospitalized with heart failure. JAMA
42. Austin PC, Daly PA, Tu JV. A multicenter study of the coding
accuracy of hospital discharge administrative data for patients admitted
to cardiac care units in Ontario. Am Heart J 2002;144:290–6.
43. Birman-Deych E, Waterman AD, Yan Y, Nilasena DS, Radford MJ,
Gage BF. Accuracy of ICD-9-CM codes for identifying cardiovascular
and stroke risk factors. Med Care 2005;43:480–5.
44. Goff DC Jr., Pandey DK, Chan FA, Ortiz C, Nichaman MZ.
Congestive heart failure in the United States: is there more than meets
the I(CD code)? The Corpus Christi Heart Project. Arch Intern Med
45. Lee DS, Donovan L, Austin PC, et al. Comparison of coding of heart
failure and comorbidities in administrative and clinical data for use in
outcomes research. Med Care 2005;43:182–8.
46. Kumler T, Gislason GH, Kirk V, et al. Accuracy of a heart failure
diagnosis in administrative registers. Eur J Heart Fail 2008;10:658–60.
47. Quan H, Li B, Saunders LD, et al. Assessing validity of ICD-9-CM
and ICD-10 administrative data in recording clinical conditions in a
unique dually coded database. Health Serv Res 2008;43:1424–41.
48. Rosamond WD, Chang PP, Baggett C, et al. Classification of heart
failure in the atherosclerosis risk in communities (ARIC) study: a
comparison of diagnostic criteria. Circ Heart Fail 2012;5:152–9.
49. Saczynski JS, Andrade SE, Harrold LR, et al. A systematic review of
validated methods for identifying heart failure using administrative
data. Pharmacoepidemiol Drug Saf 2012;21 Suppl 1:129–40.
Key Words: epidemiology y heart failure y hospital stay y
hospitalization y mortality.
For a supplementary table, please see the online version of this article.
1088Chen et al.
Trends in Heart Failure Hospital Stay
JACC Vol. 61, No. 10, 2013
March 12, 2013:1078–88