CIRCULATIONAHA/2010/948364
Trends in heart failure care: Has the incident diagnosis of heart failure shifted from the hospital to the
emergency department and outpatient clinics?
Authors: Justin A. Ezekowitz, MBBCh MSc, Padma Kaul, PhD, Jeffrey A. Bakal, PhD, Hude Quan, and
Finlay A. McAlister, MD MSc.
Affiliations: From the Mazankowski Alberta Heart Institute, University of Alberta (J.A.E, P.K., F.A.M.),
the Canadian VIGOUR Center (J.A.E, J.A.B., P.K.), Edmonton, Alberta, Canada; and the University of
Calgary (H.Q), Calgary, Alberta, Canada.
Word Count: 4575, 2 figures, 3 tables
Correspondence:
Justin A. Ezekowitz, MBBCh MSc
2C2 Cardiology WMC
8440-112 street
Edmonton, Alberta, Canada
T6G 2B7
780-407-8719 phone
780-407-6452 fax
justin.ezekowitz@ualberta.ca
1
CIRCULATIONAHA/2010/948364
Abstract
Background: Studies of heart failure (HF) incidence and prevalence frequently rely on hospitalization
records to identify patients. However, little is known about incidence rates and outcomes for HF
patients diagnosed in the outpatient or emergency department (ED) setting and, if over time, HF
incidence, prevalence, and outcomes for these patients have changed.
Methods and Results: In a population-based study of 82323 consecutive patients in a single-payer
health care system in Alberta, Canada from 1999 to 2007, we examined the HF trends over time and
one-year clinical outcomes. HF patients were first diagnosed in a general outpatient clinic (45.7%), in a
specialty outpatient clinic (4.0%), in an ED (13.7%) and in a hospital (36.6%). From year 2000 to year
2006, the age-standardized incidence (/100000 population) decreased from 538 to 403, while the
overall prevalence increased from 1585 to 2510. One-year mortality was significantly different
between patients first diagnosed in a general outpatient clinic (6.6%), a specialty outpatient clinic
(7.5%), ED (19.1%) and hospital (29.8%). Patients initially diagnosed at the time of hospitalization had
the fewest median days alive and out of hospital (347, IQR 136 to 363) over the next year compared
with patients in the ED (354, IQR 313 to 365), specialty outpatient clinic (365, IQR 355 to 365), and
general outpatient clinics (365, IQR 359 to 365, p<0.0001). Patients in the ED had the highest rate of
subsequent ED visits, all-cause hospitalization, cardiovascular hospitalization or HF hospitalization.
Conclusions
The trends observed in incidence, prevalence, and outcomes for patients with HF differ substantially
depending on whether the location of initial diagnosis includes just hospitalized cases or also includes
patients who are diagnosed in ED or outpatient clinics. Patients seen in the ED had the highest rate of
subsequent ED use or hospitalization and efforts to study these patients should be a priority.
Keywords: Heart failure, epidemiology, population
2
CIRCULATIONAHA/2010/948364
Introduction
While heart failure (HF) is known to be a major public health issue, recent reports have provided
conflicting data on whether HF incidence is increasing or decreasing and whether outcomes are
improving or static.1-7 Some of these conflicting results may be attributable to different HF case
definitions. For example, hospitalization data alone is commonly used to examine trends in HF
incidence and outcomes.1-6 Thus, in Scotland, a review of 17 years of hospitalization data showed an
increase in age-standardized incidence of HF until 1994 and declining thereafter until 2003 with a
corresponding improvement in early (1-year) and long-term (5-year) mortality over time.1 On the
other hand, some studies have included outpatients with a new diagnosis of HF when defining their
cohorts and have come to different conclusions.3, 6 For example, in Olmsted county, 42% of a
community based HF cohort were diagnosed at an outpatient visit; however, outcomes were not
reported separately for patients diagnosed in hospital vs. outpatient clinic.6 To date, no studies have
reported trends in frequency and outcomes for a broader cohort of HF patients diagnosed at any
point across the full spectrum of care: outpatient, emergency department and hospital.
By reliance on HF hospitalization alone to define a patient as having HF, the incidence and
prevalence of HF in the community may be substantially underestimated. This underestimate of HF
frequency may in fact be worsening over time as increasing availability of imaging (e.g.
echocardiography) and biomarkers (e.g. natriuretic peptides) make the diagnosis of milder cases of HF
in the outpatient setting more likely. In addition, changes in financial reimbursement for episodes of
HF care and decreasing availability of acute-care hospital beds can be expected to increase the
frequency with which patients newly diagnosed with HF are managed without hospitalization.
Although we have previously shown that patients seen in the emergency department with a principal
diagnosis of HF have equally high morbidity rates as those hospitalized at the time of initial diagnosis, 8
the outcomes for patients diagnosed in other settings such as primary care or specialty outpatient
clinics is unclear. In many countries, including Canada, the bulk of outpatient heart failure care is
delivered via primary care physicians and hence, the overall burden of HF in the community is
important to distinguish from that of incident, hospitalized HF who may be cared for by a range of
specialist physicians.9 HF care is often complex and requires repetitive office visits to ensure
adherence to self-care, medication and physical activity recommendations, and to detect worsening
symptoms early and intervene. Since specialty multidisciplinary HF clinics are available only to a
3
CIRCULATIONAHA/2010/948364
limited number of patients, estimates derived from individual or networked specialty sites may
provide only a limited and potentially biased estimate of comorbidity burden, therapy and survival.10,
11
Accordingly, we examined trends over time in the incidence, prevalence, and outcomes for
patients with HF diagnosed in any setting (outpatient, emergency department, or hospitalization) in a
single-payer system of 3.3 million people in Alberta, Canada.
Methods
Databases
We used previously described methods8, 12 utilizing data created by linking databases maintained by
the Ministry of Health and Wellness in Alberta, Canada: (1) the Discharge Abstract Database, which
records information (including dates, most responsible diagnosis, and up to 15 other diagnoses,
comorbidities, and procedures in ICD-9-CM [1994-2002]; ICD-10 [2002-2006]) on all admissions to
acute care facilities; (2) the Ambulatory Care Database, which tracks all visits to hospital-based
physicians' offices and all emergency departments (ICD-9-CM [1997-2002]; ICD-10 [2002-2006]); and
includes up to 6 diagnosis fields, (3) the Physician Claims Database, which tracks all physician claims
for outpatient services (by ICD-9-CM [1994-2006] diagnostic code and includes up to 3 codes per
encounter); and (4) the Alberta Health Care Insurance Registry, which tracks the vital status of all 3.3
million Albertans.9, 13 Each individual has a scrambled unique personal identifier by which patient
information can be tracked through each of the databases creating a continuous link of events
occurring to an individual patient.
Data Elements and Variable Definitions
The database was initially used to identify patients with a principal or secondary diagnosis of HF
between April 1, 1999 and December 31, 2006 at any one of the acute care facilities, emergency
departments, or outpatient clinics in Alberta, Canada. Specifically, patients were identified using
International Classification of Diseases, 9th revision clinical modification (ICD-9-CM) or ICD-10 codes
for heart failure (428.x or ICD-10 code I50). For the purposes of our analyses, if patients had multiple
contacts in different settings in the same 24 hour period, we used a hierarchy to define the index
locale of diagnosis – i.e. Inpatient claims superseded those from an emergency department, and these
superseded claims from outpatient specialty and outpatient primary care (general) physician claims.
4
CIRCULATIONAHA/2010/948364
Comorbidities were defined on the basis of the ICD codes at the incident or during subsequent
ED visits or hospitalization and supplemented by a search in the prior year for comorbidity diagnoses
in the hospitalization, ambulatory care or physician claims databases.14
Socioeconomic status was examined by assigning a median Statistics Canada neighborhood
household income in Canadian dollars to patients based on their recorded place of residence.
Incidence and Prevalence
Incidence was established based on the date of the earliest visit (examining records for each patient
in emergency department, inpatient, outpatient specialty clinics and general outpatient clinic) in
which HF was listed as a primary or most responsible diagnosis. Patients who did not have a diagnosis
in any of the databases prior to April 1, 1999 were considered incident. Patients were considered as
prevalent if they had a diagnosis of HF prior to this date, and once incident, until death.
Outcomes
The primary outcome of interest was one-year survival. We also examined all-cause,
cardiovascular or HF hospitalizations or ED visits. Cardiovascular ED visits or hospitalizations were any
of ICD-9-CM codes 410, 413, 417 or 428 or ICD-10 codes I2 or I47 to I50 as the most responsible
diagnosis. Resource utilization was also examined by length of stay (LOS) in hospital, and number of
days alive and out of hospital in the first year after diagnosis. In order to allow comparable survival
times, the subset of patients with incidence between April 1, 1999 and December 31, 2005 were
followed to death or one year, whichever came first.
Statistical Analysis
Results are presented with means (standardized deviation [SD]) or medians (interquartile
range [IQR]) compared using t-tests or Wilcoxon rank sum scores; counts and percentages were
analyzed using chi-square and Cochran-Armitage tests for trends. Trends in continuous variables were
tested using ANOVA. Age and sex standardized rates were calculated by grouping the ages in five year
increments from 20 to 85 and then over 85 and comparing the relevant proportion of patients from
the census of midpoint of that year. We compared medication use before and after the index visit
using McNemar’s Test. Kaplan-Meier curves were plotted and compared using log-rank statistics. A
Cox proportional hazards model was used with a stepwise variable selection (entry criteria p = 0.10)
to generate adjusted hazard ratios (HR), and 95% confidence intervals (95% CI). All tests were two-
5
CIRCULATIONAHA/2010/948364
sided, with the level of significance set at p < 0.05 unless otherwise indicated, and performed using
the SAS V9.2 (Cary, NC).
Results
Between 1999 and 2006, a total of 82323 patients were seen for an incident diagnosis of heart
failure. Patients were first diagnosed in a hospital (n=30124, 36.6%), in an emergency department
without being hospitalized (n=11252, 13.7%), in a specialty outpatient clinic (n=3306, 4.0%), or in a
general outpatient clinic (n=37641, 45.7%, Table 1). Overall, the age-standardized HF incidence
decreased from 538/100000 population in 2000 to 403/100000 population in 2006 (p<0.001 for
trend, Figure 1). The age-standardized incidence decreased over time for diagnoses made in Specialty
Outpatient Clinics (p <0.01), General Outpatient Clinics (p <0.05), the emergency department (p
<0.01), and the hospital (p <0.01). However, the declining crude incidence rates varied greatly
depending on location of diagnosis: thus, while incidence decreased by 41% between 2000 and 2006
if cases were defined only on the basis of hospitalizations (p<0.001), the decline was only 16% for
emergency department diagnoses (p<0.01) and incidence did not statistically change for other
locations of diagnosis over time. As a result, a greater proportion of all HF cases were first identified in
the outpatient setting over time. Prevalent cases of HF increased steadily from 1585 cases/100000
population in 1999 to 2510 cases/100000 patient population in 2005.
Baseline characteristics differed significantly by location of where the initial diagnosis of HF
was made (Table 2). Patients admitted to hospital or diagnosed but discharged from the emergency
department were older than those seen at an outpatient clinic (p<0.0001). Patients seen at specialty
outpatient clinics were 13% less likely to be female than at other index locations (p<0.0001). Other
vascular disease (prior myocardial infarction or revascularization, peripheral arterial disease, atrial
fibrillation), diabetes, chronic renal disease, prior malignancy and dementia were more common in
patients seen in the hospital. Patients diagnosed in the emergency department or hospital has lower
socioeconomic status compared to patients in outpatient clinics.
One-year mortality of patients after an incident diagnosis of HF was significantly different
between patients diagnosed in a General Outpatient Clinic (6.6%), a Specialty Outpatient Clinic (7.5%),
Emergency Department (19.1%), or Hospital (29.8%; p<0.0001, Table 3 and Figure 2). Hospitalized
patients had the fewest days alive and out of hospital after the index hospitalization (median days
6
CIRCULATIONAHA/2010/948364
347, IQR 136 to 363; p <0.001) compared with patients diagnosed in the Emergency Department
(median 354 days, IQR 313 to 365), Specialty Outpatient Clinic (median 365 days, IQR 355 to 365), and
General Outpatient Clinics (median 365 days, IQR 359 to 365). Of note, 67% of patients diagnosed in
the outpatient setting were never hospitalized in the first year after diagnosis (Table 3). Patients in
the ED had the highest rate of subsequent ED visits, all-cause hospitalization, cardiovascular
hospitalization or HF hospitalization (Table 3). Mean values for days alive and out of hospital were
also significant in the same pattern. One-year mortality trended lower for patients diagnosed in 2006
than in 1999 – a decrease of 6.8% over time (p=0.06). This trend was different for men (7.5% decline,
p=0.07) than women (4.7% decline, p=0.5). Predictors of 1- year mortality were examined using a Cox
proportional hazards model. Compared to patients seen initially at a General Outpatient Clinic, and
after adjusting for baseline imbalances in demographics and comorbidities, incident HF patients first
diagnosed at a Specialty Outpatient Clinic (HR 1.2, 95%CI 1.04 to 1.39), Emergency Department (HR
2.53, 95%CI 2.37 to 2.69) and Hospital (HR 3.72, 95%CI 3.54 to 3.91) had higher mortality risks.
Discussion
Our study demonstrates three principal findings: (1) the incidence and prevalence of HF are grossly
underestimated if one relies on hospitalization data alone to define cases; (2) trends over time in HF
incidence, prevalence, and outcomes differ substantially depending on location of initial diagnosis and
studies that rely on hospitalization data alone are likely to substantially overestimate the magnitude
of such secular trends; and (3) patients seen in the emergency department had the highest proportion
of subsequent HF, CV and all-cause hospitalizations. These findings are of substantial importance for
the use of administrative data, and specifically, the use of hospitalization data alone, for HF research.
To our knowledge, this is the first time this has been comprehensively described using data spanning
the full continuum of care including the emergency department.
Patients may vary in symptoms, access to health care, propensity to seek medical care, and
health systems may vary in the ability to provide equitable access. In our study within a single
healthcare system with universal access to physicians, imaging, outpatient and emergency services,
we identified that only 36.6% of newly diagnosed HF patients had an incident diagnosis from hospital,
and 13.7% from the emergency department. To date, we and others have focused predominantly on
hospitalized HF patients1, 2, 9 and recently patients seen in the emergency department8. This may
7
CIRCULATIONAHA/2010/948364
vastly underestimate the incident population, while overestimating the risk for poor outcomes for all
patients with heart failure. For example, while the one-year mortality rate was 29.8% for hospitalized
patients, it was only 6.6% and 7.5% for those initially diagnosed in general outpatient or specialty
clinic offices, respectively. These patients had a different risk profile with fewer comorbid features so
caution should be exercised in attempting to draw conclusions about severity of disease seen in
different health care settings.
Overall age-standardized incidence of HF in our study of 403 patients per 100000 population in
2006 is higher than reported from other locales. Studies such as those by Jhund et al and
Schaufelberger et al using hospitalizations with a most responsible diagnosis of HF to define their
cohorts reported much lower rates than we observed. Roger and colleagues used hospital and
outpatient data from Olmsted County and demonstrated a small but non-significant increase in the
incidence of HF over time from 1979 to 2000 with an incidence of 383 (for men) and 315 (for women)
per 100000 population new diagnoses for HF between 1996 and 2000.6 This study importantly
included 42% of patients who were initially identified as an outpatient but only 26% were never
hospitalized – the majority of our cohort (69%) initially diagnosed as an outpatient were not
hospitalized in the next year. Thus, our data demonstrates the need to incorporate a broader
spectrum of data for a closer approximation of the true incidence of HF.
Survival data from first presentation highlights the importance of intervening early with life
prolonging medications and devices in an optimal system of care. While Jhund et al. report a median
increase in survival post-hospital discharge of 1.01 years over the 16 years they studied, the median
survival after hospitalization remained a dismal 2.34 years for men and 1.79 years for women in
Scotland.1 In our study, patients initially seen as an outpatient had a one-year mortality of 6.6%
compared with those whose incident visit was a hospitalization in which one-year mortality rate was
29.8%. The mortality rate in the recent GISSI-HF trial (which was largely conducted in the outpatient
setting and enrolled patients of similar age and baseline characteristics as our primary care clinic
diagnosed patients) was ~7%, similar to that of our patients first diagnosed in the outpatient clinic
setting.15 In contrast, a recent clinical trial enrolling hospitalized HF patients reported a similar allcause mortality rate at 1 year (26.0% in the placebo arm) to that of our hospital-diagnosed patients.16
Improvements in survival of patients with HF is likely multifactorial. Recognition of HF and risk
factors for HF likely plays a role, as does availability and application of evidence-based medications
8
CIRCULATIONAHA/2010/948364
with a known mortality benefit (e.g. ACE inhibitors, beta-blockers, spironolactone), improved systems
of care at the specialty and family physician level, availability of publically accessible HF guidelines,
and improved recognition and treatment of risk factors for cardiovascular disease in general. As an
example, antihypertensive therapy prevents the development of heart failure.17 In Canada, a recent
survey found that 66% of Canadian adults with hypertension were being treated with medication and
controlled to levels recommended by national and international guidelines.18 This is reassuring, as
prevention of HF via diagnosis and treatment of hypertension and other risk factors for coronary
artery disease will be the mainstay for reducing the incidence of HF in the coming decades.
Some strengths and limitations of our data deserve recognition. Administrative data has been
validated in multiple populations for patients hospitalized with heart failure 19, 20 and for
comorbidities.14 Prior work with Medicare data has used 3 claims for heart failure in 20 months to
define incident outpatient HF,3 however, the outpatient codes for heart failure used in this analysis
have undergone validation in our jurisdiction.14 Chart re-abstraction data from 3300 physician charts
in our healthcare region did not show a difference in sensitivity, specificity or predictive value when
different case definition algorithms employing single vs. multiple outpatient claims were used (H.
Quan, personal communication). Data used for our analyses has significant strengths over that drawn
from selected registry patients, patients in randomized trials, or those in whom inclusion is dependent
on insurance or access to care since we were able to include information on all patients seen within a
geographic region with a single payer health care system and universal access to care. However, we
lack information on laboratory values, ejection fraction, and functional status and as such cannot
draw conclusions about trends in diagnosis or outcomes for systolic versus diastolic HF.
Conclusions
The incidence of heart failure is declining over time, however, the case definition (and specifically
location of first encounter) is critically important when interpreting administrative or registry data.
One-year mortality for patients with heart failure is declining, and patients diagnosed initially as an
outpatient have a more favorable outcome than those first diagnosed in the emergency department
or hospital. Patients seen in the emergency department had the highest rate of subsequent ED use or
hospitalization and efforts to study these patients should be a priority.
9
CIRCULATIONAHA/2010/948364
Acknowledgements
Drs. Ezekowitz and Kaul are supported by the New Investigator program of the Canadian Institutes of
Health Research (CIHR) and the Alberta Heritage Foundation for Medical Research (AHFMR).
Drs. McAlister and Quan are supported by AHFMR. Dr. McAlister also holds the Patient Health
Management Chair at the University of Alberta. Additional funding was provided from AHFMR via the
Alberta HEART Team Grant.
This study is based, in part, on de-identified data provided by Alberta Health and Wellness through
the Alberta Cardiac Access Collaborative; however, the interpretation and conclusions contained
herein do not necessarily represent the views of the Government of Alberta nor Alberta Health and
Wellness.
Disclosures
No conflicts of interest to disclose.
10
CIRCULATIONAHA/2010/948364
Figures.
Figure 1. Age-adjusted trends over time of outpatient, emergency department and hospitalization for
incident heart failure.
Figure 2. Survival curves after index visit for heart failure at Specialty Outpatient Clinic, General
Outpatient Clinic, Emergency Department or hospital.
11
CIRCULATIONAHA/2010/948364
Table 1. Location of incident diagnosis of heart failure, by year.
Year of index visit
1999*
2000
2001
2002
2003
2004
2005
2006
8518
11117
11055
10470
10350
10131
10402
10282
3807
4980
4899
4665
4491
4725
4934
5140
(44.7)
(44.8)
(44.3)
(44.6)
(43.4)
(46.6)
(47.4)
(50.0)
210
368
370
420
497
499
492
450
(2.5)
(3.3)
(3.4)
(4.0)
(4.8)
(4.9)
(4.7)
(4.4)
Emergency
1111
1455
1496
1488
1425
1336
1430
1512
Department
(13.0)
(13.1)
(16.5)
(14.2)
(13.8)
(13.2)
(13.8)
(14.7)
Hospital
3390
4314
4290
3897
3937
3571
3546
3180
(39.8)
(38.8)
(38.8)
(37.2)
(38.0)
(35.3)
(34.1)
(30.9)
N
General
Outpatient
Clinic
Specialty
Outpatient
Clinic
Values are n (%). *1999 is partial year (April-December)
12
CIRCULATIONAHA/2010/948364
Table 2. Baseline characteristics of incident patients (1999-2005) with heart failure by location of
incident diagnosis.
Location of index visit
General
Specialty
Outpatient
Outpatient
Clinic
Clinic
32501 (45.1)
2856(4.0)
9741 (13.5)
26945 (37.4)
72 (60-80)
69 (57-77)
75 (65 -82)
77 (67-84)
<0.0001
16397 (50.5)
1248 (43.7)
4810 (49.4)
13452 (50.1)
<0.0001
4273 (13.2)
814 (28.5)
2189 (22.5)
10230 (38.0)
<0.0001
1637 (5.0)
277 (9.7)
538 (5.5)
3121 (11.6)
<0.0001
Hypertension
19197 (59.1)
1898 (66.5)
6418 (65.9)
18431 (68.4)
<0.0001
Atrial fibrillation
4794 (14.8)
574 (20.1)
2223 (22.8)
8357 (31.0)
<0.0001
Diabetes
6324 (19.5)
722 (25.3)
2733 (28.1)
8233 (30.6)
<0.0001
2062 (6.3)
310 (10.9)
989 (10.2)
4586 (17.0)
<0.0001
2953 (9.1)
331 (11.6)
5160 (19.2)
1328 (13.6)
<0.0001
1203 (3.7)
215 (7.5)
1059 (10.9)
4322 (16.0)
<0.0001
COPD
8861 (27.3)
936 (32.8)
4388 (45.1)
11704 (43.4)
<0.0001
Cancer
2985 (9.2)
272 (9.5)
1288 (13.2)
4840 (18.0)
<0.0001
Dementia
1319 (4.1)
128 (4.5)
840 (8.6)
3582 (13.3)
<0.0001
Income,
57138
57746
54787
54785
Canadian dollars,
(44289-
(45597-
(42872-
(42872-
N (%)
Age, years,
median (IQR)
Female
Prior myocardial
infarction
Prior
revascularization
Peripheral
arterial disease
Cerebrovascular
disease
Chronic renal
disease
Emergency
department
13
Hospital
p-value
<0.0001
CIRCULATIONAHA/2010/948364
median (IQR)
75254)
76962)
72857)
71871)
Values are n(%) unless otherwise stated. Prior revascularization includes percutaneous coronary
intervention and coronary artery bypass grafting. COPD = chronic obstructive pulmonary disease. IQR
= interquartile range.
14
CIRCULATIONAHA/2010/948364
Table 3. One-year outcomes of patients with heart failure by location of incident visit.
Location of index visit
General
Specialty
Outpatient
Outpatient
Clinic
Clinic
32501 (45.1)
2856 (4.0)
9741 (13.5)
26945 (37.4)
Any ED visit
10096 (31.1)
1048 (36.7)
4505 (46.3)
9229 (34.3)
<0.0001
Any hospitalization
10183 (31.3)
1150 (40.3)
5809 (59.6)
12116 (45.0)
<0.0001
CV hospitalization
4798 (14.8)
635 (22.2)
3676 (37.7)
7703 (28.6)
<0.0001
HF hospitalization
1211 (3.7)
128 (4.5)
1454 (14.9)
2121 (7.9)
<0.0001
365 (359-
365 (355-
354 (313-
347 (136-
365)
365)
365)
363)
344 (64.6)
336 (75.9)
288 (121.0)
254 (149.1)
<0.0001
2156 (6.6)
214 (7.5)
1859 (19.1)
8036 (29.8)
<0.0001
N (%)
Days alive and out
of hospital,
median (IQR)
Emergency
department
Hospital
p-value
<0.0001
Days alive and out
of hospital, mean
(SD)
Mortality
*Incident population restricted to minimum of one-year complete follow-up from index date
therefore includes patients from April 1, 1999 to December 31, 2005. CV = cardiovascular. ED =
emergency department. HF = heart failure. IQR = interquartile range. SD = standard deviation.
15
CIRCULATIONAHA/2010/948364
Figure 1.
16
CIRCULATIONAHA/2010/948364
0.8
0.7
General Outpatient Clinic n = 32501
Specialty Outpatient Clinic n = 2856
Emergency Department
n = 9741
Hospital
n = 26945
0.6
Survival
0.9
1.0
Figure 2.
0
50
100
150
200
Days
17
250
300
350
CIRCULATIONAHA/2010/948364
Reference List
(1) Jhund PS, MacIntyre K, Simpson CR, Lewsey JD, Stewart S, Redpath A, Chalmers JW, Capewell
S, McMurray JJ. Long-term trends in first hospitalization for heart failure and subsequent
survival between 1986 and 2003: a population study of 5.1 million people. Circulation.
2009;119:515-23.
(2) Tu JV, Nardi L, Fang J, Liu J, Khalid L, Johansen H, for the Canadian Cardiovascular Outcomes
Research Team. National trends in rates of death and hospital admissions related to acute
myocardial infarction, heart failure and stroke, 1994-2004. CMAJ. 2009;180:E118-E125.
(3) Curtis LH, Whellan DJ, Hammill BG, Hernandez AF, Anstrom KJ, Shea AM, Schulman KA.
Incidence and Prevalence of Heart Failure in Elderly Persons, 1994-2003. Archives of Internal
Medicine. 2008;168:418-24.
(4) Teng THK, Finn J, Hobbs M, Hung J. Heart Failure: Incidence, Case-Fatality and Hospitalization
Rates in Western Australia between 1990 and 2005. Circ Heart Fail. 2010;CIRCHEARTFAILURE.
(5) Schaufelberger M, Swedberg K, Koster M, Rosen M, Rosengren A. Decreasing one-year
mortality and hospitalization rates for heart failure in Sweden: Data from the Swedish Hospital
Discharge Registry 1988 to 2000. Eur Heart J. 2004;25:300-7.
(6) Roger VL, Weston SA, Redfield MM, Hellermann-Homan JP, Killian J, Yawn BP, Jacobsen SJ.
Trends in heart failure incidence and survival in a community-based population. JAMA.
2004;292:344-50.
(7) Levy D, Kenchaiah S, Larson MG, Benjamin EJ, Kupka MJ, Ho KK, Murabito JM, Vasan RS. Longterm trends in the incidence of and survival with heart failure. N Engl J Med. 2002;347:1397402.
(8) Ezekowitz JA, Bakal JA, Kaul P, Westerhout CM, Armstrong PW. Acute heart failure in the
emergency department: Short and long-term outcomes of elderly patients with heart failure.
Eur J Heart Fail. 2008;10:308-14.
(9) Ezekowitz JA, van WC, McAlister FA, Armstrong PW, Kaul P. Impact of specialist follow-up in
outpatients with congestive heart failure. CMAJ. 2005;172:189-94.
(10) McAlister FA, Teo KK, Taher M, Montague TJ, Humen D, Cheung L, Kiaii M, Yim R, Armstrong
PW. Insights into the contemporary epidemiology and outpatient management of congestive
heart failure. Am Heart J. 1999;138:87-94.
(11) McAlister FA, Tu JV, Newman A, Lee DS, Kimber S, Cujec B, Armstrong PW. How many patients
with heart failure are eligible for cardiac resynchronization? Insights from two prospective
cohorts. Eur Heart J. 2006;27:323-9.
18
CIRCULATIONAHA/2010/948364
(12) Kaul P, Chang WC, Westerhout CM, Graham MM, Armstrong PW. Differences in admission
rates and outcomes between men and women presenting to emergency departments with
coronary syndromes. CMAJ. 2007;177:1193-9.
(13) Chang WC, Kaul P, Westerhout CM, Graham MM, Fu Y, Chowdhury T, Armstrong PW. Impact
of sex on long-term mortality from acute myocardial infarction vs unstable angina. Arch Intern
Med. 2003;163:2476-84.
(14) Quan H, Parsons GA, Ghali WA. Validity of information on comorbidity derived rom ICD-9-CCM
administrative data. Med Care. 2002;40:675-85.
(15) Tavazzi L, Maggioni AP, Marchioli R, Barlera S, Franzosi MG, Latini R, Lucci D, Nicolosi GL, Porcu
M, Tognoni G. Effect of n-3 polyunsaturated fatty acids in patients with chronic heart failure
(the GISSI-HF trial): a randomised, double-blind, placebo-controlled trial. Lancet.
2008;372:1223-30.
(16) Gheorghiade M, Gattis WA, O'Connor CM, Adams KF, Jr., Elkayam U, Barbagelata A, Ghali JK,
Benza RL, McGrew FA, Klapholz M, Ouyang J, Orlandi C. Effects of tolvaptan, a vasopressin
antagonist, in patients hospitalized with worsening heart failure: a randomized controlled trial.
JAMA. 2004;291:1963-71.
(17) Levy D, Larson MG, Vasan RS, Kannel WB, Ho KK. The progression from hypertension to
congestive heart failure. JAMA. 1996;275:1557-62.
(18) Wilkins K, Campbell NRC, Joffres MR, McAlister FA, Nichol M, Quach S, Johansen HL, Tremblay
MS. Blood pressure in Canadian adults. Health Reports (Statistics Canada). 2010;21:82-003XPE.
(19) 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. 2000;160:197-202.
(20) 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.
19