European Heart Journal (2004) 25, 1614–1619

Clinical research

Quantifying the heart failure epidemic:

prevalence, incidence rate, lifetime risk
and prognosis of heart failure

Downloaded from [Link] by guest on 14 September 2023

The Rotterdam Study

Gysèle S. Bleuminka,b, Anneke M. Knetscha,c, Miriam C.J.M. Sturkenbooma,d,

Sabine M.J.M. Strausa, Albert Hofmana, Jaap W. Deckersa,c,
Jacqueline C.M. Wittemana, Bruno [Link]. Strickera,b,*
a
Department of Epidemiology and Biostatistics, Erasmus Medical Centre, PO Box 1738, 3000, Rotterdam,
The Netherlands
b
Inspectorate for Healthcare, The Hague, The Netherlands
c
Department of Cardiology, Erasmus Medical Centre, Rotterdam, The Netherlands
d
Department of Medical Informatics, Erasmus Medical Centre, Rotterdam, The Netherlands
Received 8 March 2004; revised 14 June 2004; accepted 24 June 2004
Available online 21 August 2004

KEYWORDS Aims To determine the prevalence, incidence rate, lifetime risk and prognosis of
Heart failure; heart failure.
Epidemiology; Methods and Results The Rotterdam Study is a prospective population-based cohort
Prevalence; study in 7983 participants aged P55. Heart failure was defined according to criteria
Incidence;
of the European Society of Cardiology. Prevalence was higher in men and increased with
Lifetime risk;
age from 0.9% in subjects aged 55–64 to 17.4% in those aged P85. Incidence rate of
Prognosis
heart failure was 14.4/1000 person-years (95% CI 13.4–15.5) and was higher in men
(17.6/1000 man-years, 95% CI 15.8–19.5) than in women (12.5/1000 woman-years,
95% CI 11.3–13.8). Incidence rate increased with age from 1.4/1000 person-years in
those aged 55–59 to 47.4/1000 person-years in those aged P90. Lifetime risk was
33% for men and 29% for women at the age of 55. Survival after incident heart failure
was 86% at 30 days, 63% at 1 year, 51% at 2 years and 35% at 5 years of follow-up.
Conclusion Prevalence and incidence rates of heart failure are high. In individuals aged
55, almost 1 in 3 will develop heart failure during their remaining lifespan. Heart failure
continues to be a fatal disease, with only 35% surviving 5 years after the first diagnosis.
c 2004 Published by Elsevier Ltd on behalf of The European Society of Cardiology.

Introduction

Heart failure constitutes a major public health burden in

the western world. Since incidence rates appear to re-
* Corresponding author. Tel.: +31 10 4088294; fax: +31 10 4089382. main stable over the years, at least in men,1 prevalence
E-mail address: [Link]@[Link] ([Link]. Stricker). estimates of heart failure are bound to increase as the


0195-668X/$ - see front matter c 2004 Published by Elsevier Ltd on behalf of The European Society of Cardiology.
doi:10.1016/[Link].2004.06.038
Quantifying the heart failure epidemic 1615

population ages. Hospitalisation rates for heart failure To obtain recent estimates, the point prevalence of heart
have increased considerably.2 The proportion of patients failure was determined at the 1st of January of 1997, 1998
having multiple hospital admissions is rising. In addition, and 1999. Calculations were performed in all participants of
large observational studies have failed to show any sub- the Rotterdam Study who were alive and present at January 1
of each of these years. Four participants were excluded because
stantial change in the prognosis of heart failure in the
of missing medical records. For estimation of incidence rates
general population, despite evidence-based advances in and lifetime risks, the study population comprised 7734 subjects
treatment.3 Hospitalisation rates do not necessarily re- who were free from heart failure at baseline. Subjects were fol-
flect the true incidence and prevalence of heart failure lowed from baseline until the first of one of the following: a di-
in the general population, as only the more serious stages agnosis of incident heart failure, death, loss to follow-up (<1%),
of this syndrome require in-hospital evaluation and treat- date of last collection of information for determination of heart
ment. Although data regarding heart failure incidence, failure, or January 1, 2000. The date of last information on heart
prevalence and prognosis in the community are vital, failure status preceded January 1, 2000 for 14.4% of partici-
few large prospective population-based studies have pants. For the calculation of survival estimates, incident heart

Downloaded from [Link] by guest on 14 September 2023

been published that provide recent estimates, especially failure patients were followed from the date of incident heart
failure until the earliest of death, removal from the study area,
in European populations. Furthermore, most recent pop-
or January 1, 2000.
ulation-based estimates originate from relatively short-
term studies,4–6 except for the Framingham Heart Study1
and the Cardiovascular Health Study,7 both of which Heart failure assessment
were performed in the United States. The diagnosis of
heart failure is complex. Signs and symptoms are not spe- Assessment of prevalent heart failure at the baseline examina-
cific and a gold standard to assess the presence of this tion in the Rotterdam Study has been described in detail previ-
disease is lacking. Previously published studies have used ously.10 Briefly, a validated score was used, similar to the
various criteria to assess the presence of heart failure. definition of heart failure of the European Society of Cardiol-
The European Society of Cardiology has therefore pro- ogy.8 This score was based on the presence of at least two signs
vided guidelines for the diagnosis of heart failure, for or symptoms suggestive of heart failure (shortness of breath, an-
kle swelling and pulmonary crepitations) or use of medication
use in clinical practice and epidemiological surveys.8
for the indication of heart failure, in combination with objective
According to these guidelines, objective evidence of car-
evidence of cardiovascular disease. Questions on indication of
diac dysfunction has to be present to establish the pres- cardiovascular medication and breathlessness were lacking at
ence of heart failure, in addition to typical symptoms the start of the Rotterdam Study, but were subsequently added.
(e.g., breathlessness) suggestive of the diagnosis. Consequently, this information was obtained in only 5540 partic-
This study was designed to calculate the prevalence, ipants. In addition, prevalent heart failure cases were obtained
incidence, and lifetime risk of heart failure in partici- through a database containing hospital discharge diagnoses from
pants of the Rotterdam Study, a large prospective popu- all hospitals in the Rotterdam area as of January 1, 1991. Re-
lation-based cohort study with more than 10 years of cords from this database were linked to the Rotterdam Study da-
follow-up. In addition, we studied the prognosis of cases tabase. For potential cases of heart failure identified in this
way, copies of discharge letters were requested. Furthermore,
of incident heart failure.
all medical records were screened in retrospect for the occur-
rence of heart failure in the majority (97%) of participants of
the Rotterdam Study. With these three methods, information
Methods on the presence of heart failure at baseline was available for
all participants.
Setting and study population Cases of incident heart failure were obtained by continuously
monitoring participants of the Rotterdam Study for the occur-
The Rotterdam Study was a population-based prospective cohort rence of heart failure during follow-up through automated link-
study of cardiovascular, locomotor, neurologic and ophthalmo- age with files from general practitioners. All available data on
logic diseases in the elderly.9 All inhabitants of Ommoord, a sub- these events, such as hospital discharge letters and notes from
urb of Rotterdam in the Netherlands, who were 55 years of age general practitioners, were copied from the medical records.
or older were invited to participate. Of all 10 275 subjects in this Apart from this systematic follow-up procedure, we used veri-
age group, 7983 agreed to participate (78%). The Medical Ethics fied hospital discharge diagnoses for case finding, gathered from
Committee of the Erasmus Medical Centre approved the study. all hospitals in the Rotterdam area as described above. The date
The baseline examination was conducted between July 1989 of incident heart failure was defined as the day of the first oc-
and 1993. Participants were visited at home for a standardized currence of symptoms suggestive of heart failure, obtained from
questionnaire and were subsequently examined at the research the medical records, or the day of receipt of a first prescription
centre. Since the start of the Rotterdam Study, cross-sectional for a loop diuretic or an ACE-inhibitor, whichever came first.
surveys have been carried out periodically. In addition, partici- The diagnosis of heart failure was classified as definite, prob-
pants are continuously monitored for major events that occur able, possible or unlikely. Definite heart failure was defined as a
during follow-up, including heart failure, through automated combination of the presence of at least one of the typical signs or
linkage with files from general practitioners. Information on vi- symptoms of heart failure, such as breathlessness at rest or
tal status is obtained regularly from municipal health authorities during exertion, ankle oedema and pulmonary crepitations, and
in Rotterdam and from the general practitioners working in the confirmation by objective evidence of cardiac dysfunction (chest
study district of Ommoord, and was complete for all participants X-ray, echocardiography). This definition is in accordance with
until January 1, 2000. Furthermore, all drug prescriptions dis- the criteria of the European Society of Cardiology.8 Also, for def-
pensed to participants by all pharmacies in the study area are inite heart failure, the diagnosis had to have been made by a
routinely stored in the database. medical specialist. Heart failure was classified as probable when
1616 G.S. Bleumink et al.

at least two typical symptoms suggestive of heart failure were Study. In the remaining study population (n = 7734), we
present, and at least 1 of the following: history of cardiovascular identified 725 incident cases of heart failure (335 men,
disease (e.g., myocardial infarction, hypertension), response to 390 women), of whom 673 were classified as definite,
treatment for heart failure, or objective evidence of cardiac dys- and 52 as probable cases. The median follow-up time in
function, while symptoms could not be attributed to another un-
this population was 7.1 years (interquartile range: 5.7–
derlying disease, such as chronic obstructive pulmonary disease.
Heart failure was classified as possible when one of the criteria
8.0) and we had 50 268 person-years of observation in to-
for probable heart failure could not be met. For both probable tal. The majority of our study population was female
and possible heart failure, a diagnosis of a general practitioner (61%) and mean age at baseline was 70.4 years (standard
sufficed. Heart failure was considered unlikely if signs or symp- deviation 9.7 years). Mean age at the onset of heart fail-
toms were present, but when objective evidence failed to show ure was significantly higher in women than in men (82.5
cardiac dysfunction, and if signs or symptoms could be attributed and 77.5 years, respectively).
to another underlying disease. Two research physicians inde-
pendently classified all information on potential heart failure
Prevalence

Downloaded from [Link] by guest on 14 September 2023

events. If there was disagreement, a consensus was reached in
a separate session. Finally, a cardiologist verified all probable
and possible cases, and all cases in which the two physicians Point prevalence of heart failure was determined at Jan-
could not reach consensus. If the cardiologist disagreed with uary 1 of 1997, 1998 and 1999 and was 6.4% (95% CI 5.8–
the research physicians, the cardiologist’s judgment was consid- 7.0), 6.7% (95% CI 6.1–7.4) and 7.0% (95% CI 6.4–7.7),
ered decisive. The research physicians and the cardiologist based respectively. Mean age of the study population increased
their decisions on the same data. Only definite and probable
from 73.3 years in 1997 to 74.5 years in 1999. Prevalence
cases were included in the analyses.
was higher in men than in women (e.g., 1998: men 8.0%,
women 6.0%). There was a sharp rise of prevalence esti-
Statistical analysis mates with age. For example, in 1998 point prevalence
increased from 0.9% (95% CI 0.5–1.6) in subjects aged
Prevalence of heart failure per calendar year was calculated by
55–64 years, 4.0% (95% CI 3.3–4.8) in subjects aged
dividing the number of persons with prevalent heart failure by
the number of subjects present in the study population at Janu-
65–74 years, 9.7% (95% CI 8.4–11.1) in those aged 75–
ary 1 of each calendar year. 95% confidence intervals (CI) were 84 years to 17.4% (95% CI 14.8–20.4) in those aged 85
calculated with Wilson’s (score) method for a binomial propor- years or over.
tion. Prevalence estimates were calculated for men and women
separately and in 10-year age categories. The incidence rate of
heart failure was determined by dividing the number of cases of Incidence rate
incident heart failure by the total number of person-years accu-
mulated in the study population without heart failure at base- The overall incidence rate of heart failure was 14.4/1000
line. The 95% CI around the estimates were calculated based person-years (95% CI 13.4–15.5) and was significantly
on the Poisson distribution. Incidence rate estimates were cal- higher in men (17.6/1000 man-years, 95% CI 15.8–19.5)
culated by gender and age (5-year categories).
than in women (12.5/1000 woman-years, 95% CI 11.3–
To calculate the risk to develop heart failure over time, com-
peting risk of death was taken into account. First, we calculated
13.8). The incidence rate increased with age from 1.4/
heart failure free survival at different ages with the Kaplan– 1000 person-years in those aged 55–59 years to 47.4/
Meier method, using incident heart failure and mortality data 1000 person-years in those aged 90 years or older (Table
from the study cohort. Age at baseline was used as the entry time 1). This increase with age was evident for both genders
variable and age at the end of follow-up, incident heart failure, or (Figs. 1(a) and (b)). Incidence rates were on average ap-
death, as failure time variable. Both death and incident heart proximately two times higher in men than in women in
failure were classified as failures. Second, the cumulative abso- each age category, except for the youngest (55–59
lute risk of heart failure over a period was calculated as the inte- years), in which no male cases occurred.
grated product of the age-specific heart failure incidence rates
and heart failure free survival.11 The risks of heart failure over
time were calculated for the total population and for men and Period and lifetime risk
women separately at the ages of 55, 65, 75, and 85 years.
The prognosis of heart failure was determined in 725 sub- The period and lifetime risks for all subjects, and for men
jects with incident heart failure during follow-up. Survival after and women separately, at the ages of 55, 65, 75, and 85
incident heart failure (30-day, 1-year, 2-year and 5-year) was
years are shown in Table 2. All estimates account for the
calculated using the Kaplan–Meier method. Survival was also
determined after exclusion of patients who died in the first 30
risk of competing causes of death. The lifetime risk of
days, thereby excluding those with a first diagnosis of heart fail- heart failure for a person aged 55 was 30.2%. For a
ure on the day of their death and the most severe cases of heart man aged 55 years the lifetime risk was 33.0% and for a
failure. We used Cox proportional hazards regression analysis to woman of the same age it was 28.5%. Lifetime risk of
study gender differences in survival, adjusted for age. heart failure decreased with age in both sexes to approx-
imately 23% in persons who reached 85 years of age with-
out having heart failure. Stratification by gender showed
Results that lifetime risks were higher in men than women at
ages 55–75. In subjects aged 85 years, however, lifetime
A total of 245 prevalent heart failure cases (88 men, 157 risks for developing heart failure were comparable and
women) were identified at baseline in the Rotterdam slightly higher in women (Table 2 and Fig. 2). Cumulative
Quantifying the heart failure epidemic 1617

Table 1 Incidence rates for heart failure per 5-year age category
Age category (years) Number of incident cases Person-years Incidence ratea (95% CI)
55–59 4 2888.6 1.4 (0.5–3.3)
60–64 27 8713.6 3.1 (2.1–4.4)
65–69 56 10392.1 5.4 (4.1–6.9)
70–74 113 9665.6 11.7 (9.7–14.0)
75–79 136 8012.8 17.0 (14.3–20.0)
80–84 166 5513.5 30.1 (25.8–35.0)
85–89 137 3269.0 41.9 (35.3–49.4)
P 90 86 1813.5 47.4 (38.6–58.2)
a
Per 1000 person-years.

Downloaded from [Link] by guest on 14 September 2023

90 gender-adjusted survival was significantly lower in sub-
80 jects with incident heart failure than in the remainder

Incidence rate
70
of our study cohort (hazard ratio 4.3, 95% CI 3.8–4.8).
60
50
40
30
20
10 Discussion
0
55-59 60-64 65-697 0-74 75-798 0-84 85-89 ≥90 In this long-term prospective population-based cohort
(a) age category (years) study, we found that heart failure prevalence, incidence
and risk are high. The incidence rate was significantly
90
80
higher in men than in women and increased with age
from 1.4/1000 person-years in subjects aged 55–59 years
Incidence rate

70
60 to 47.4/1000 person-years in those aged 90 years or
50 older. Our study showed that the probability for an indi-
40 vidual aged 55 years to develop heart failure during his or
30
20
her remaining lifetime is 30.2%. As expected, lifetime
10 risk decreased at older ages, probably because of deple-
0 tion of susceptibles and a shorter remaining lifespan. In
55-59 60-64 65-69 70-74 75-79 80-84 85-89 ≥90 our study, lifetime risk of heart failure was higher in
(b) age category (years) young men than in young women. In the older individu-
als, however, lifetime risks were practically the same
Fig. 1 (a) Age-specific male incidence rates (/1000 man years) and 95%
confidence band. (b) Age-specific female incidence rates (/1000 woman-
in men and women. Heart failure remains a deadly dis-
years) and 95% confidence band. ease for both genders, with a 5-year survival of only 35%.
Our age-specific incidence rate estimates are similar
to the results from an investigation in a general practi-
tioner’s database in the United Kingdom,12 but differ
risks in shorter time intervals (5–25 years) increased somewhat from other recent population-based studies.
with age and were higher in men at all ages, reflecting Estimates in the Cardiovascular Health Study were higher
the higher incidence rates in men. in all age categories. Although this study also used clini-
cal criteria for the assessment of heart failure, the inves-
Prognosis tigators selected their participants through a Medicare
eligibility list.7 This may explain some of the differences
Of the 725 persons with incident heart failure, 445 sub- with our study, which was performed in an unselected
jects died following the diagnosis (198 men and 247 population. Besides differences in selection criteria and
women). Median survival was 2.1 years (range: 1 day– population characteristics, comparison between investi-
9.0 years). Cumulative survival was 86% at 30 days after gations is further complicated because studies have used
the onset of heart failure (95% CI 83–88%), 63% at 1 year different criteria to assess the presence of heart failure.
(95% CI 59–66%), 51% at 2 years (95% CI 47–55%) and 35% For example, in the Framingham Heart Study, clinical cri-
at 5 years (95% CI 31–39%). There was no significant dif- teria were used that do not include evidence of cardiac
ference in cumulative survival after incident heart fail- dysfunction on echocardiography, which is an important
ure between men and women (Fig. 3, log rank test: tool in heart failure diagnosis in clinical practice.1 There-
p = 0.15). Age-adjusted survival in Cox proportional haz- fore, in the Framingham Heart Study, the true incidence
ards analysis was similar in men and women (hazard ratio of heart failure may have been underestimated. In the
female gender: 0.88, 95% CI 0.72–1.07). After exclusion Hillingdon heart failure study, potential cases were iden-
of patients who died in the first 30 days, 1-, 2- and 5-year tified on the basis of referrals by general practitioners of
survival were 73%, 59% and 41%, respectively. Age- and patients with suspected heart failure.6 Although similar
1618 G.S. Bleumink et al.

Table 2 Cumulative risk of heart failure in different time periods for participants aged 55, 65, 75, and 85 years old; total and
stratified by gender
Period riska (years)

Age 5 10 15 20 25 30 35 Lifetime

Total
55 0.6 2.1 4.5 9.2 14.7 21.8 27.2 30.2
65 2.6 7.6 13.6 21.3 27.1 30.3
75 7.5 17.2 24.6 28.7
85 14.8 23.1

Men
55 0 2.8 6.8 13.4 19.6 27.9 31.6 33.0

Downloaded from [Link] by guest on 14 September 2023

65 4.2 11.4 18.2 27.1 31.2 32.7
75 9.5 22.0 27.7 29.8
85 16.2 22.4

Women
55 1.0 1.8 3.0 6.2 11.2 17.5 24.3 28.5
65 1.2 4.6 10.0 16.7 24.0 28.5
75 6.2 14.1 22.6 27.9
85 14.3 23.3
a
Numbers are percentages. Competing risk of death is taken into account.

35 criteria were used in this study, age-specific incidence

was somewhat lower, possibly because not all potential
30
cases were referred. Age-specific prevalence estimates
25 of heart failure were also somewhat higher in the Cardi-
ovascular Health Study, especially at younger ages.13
risk (%)

20 Slightly lower prevalence estimates per age-category

15 were found in a study in residents of Olmsted County,
men Minnesota and in the Framingham Heart Study.4,14 Both
10 women used Framingham criteria for case ascertainment. The
Echocardiographic Heart of England Screening study used
5
criteria based on the guidelines of the European Society
0 of Cardiology and found age-specific prevalence esti-
55 65 75 85 mates of heart failure that were similar to ours.5
age (years) Only one other study, the Framingham Heart Study,
calculated lifetime risks for heart failure.15 Lifetime risk
Fig. 2 Age-specific lifetime risk of heart failure startified by gender. for the development of heart failure in this study was
approximately 20% and was, in contrast to our findings,
independent of age and gender. The investigators did
not find a decrease in lifetime risk at older ages, which
was attributed to an increasing incidence with advancing
Cumulative
survival age, outpacing the increasing mortality from competing
1.0 causes. However, no age-limit was set for the calcula-
tion of cumulative risks in the Rotterdam Study, while
0.8 in the Framingham Heart Study cumulative risks were
calculated until the age of 94 years. Furthermore, life-
0.6 time risks in the Framingham Study were calculated
from 1971 to 1996, while in the Rotterdam Study they
0.4 were calculated from 1989 to 2000. Therefore, changes
male in mortality from competing causes over calendar time
0.2 may explain some of the differences between the two
female
studies. Furthermore, although questioned by some,16
0.0 improvements in myocardial infarction treatment over
0 2 4 6 8 time might account for the higher incidence rates of
years
heart failure that we found.
Fig. 3 Kaplan–Meier survival curve for incident heart failure cases Heart failure is a fatal disease, despite advances
stratified by gender. in treatment over the past 15 years.3 We found no
Quantifying the heart failure epidemic 1619

differences between men and women in heart failure 2. Stewart S, MacIntyre K, MacLeod MMC et al. Trends in hospitalisation
for heart failure in Scotland, 1990–1996. Eur Heart J
prognosis. Our survival estimates are very similar to
2001;22:209–17.
those found in three other recent population-based stud- 3. Jessup M, Brozena S. Heart failure. N Engl J Med 2003;348:2007–18.
ies.1,12,17 However, compared to heart failure mortality 4. Redfield MM, Jacobsen SJ, Burnett Jr JC et al. Burden of systolic
in hospital-based studies,18–20 prognosis in our popula- and diastolic ventricular dysfunction in the community. Appreciat-
tion-based study was better, probably as less severe ing the scope of the heart failure epidemic. JAMA 2003;
289:194–202.
cases of heart failure were also included. As the diagno- 5. Davies MK, Hobbs FDR, Davis RC et al. Prevalence of left-ventricular
sis of heart failure is difficult, some studies applied systolic dysfunction and heart failure in the Echocardiographic Heart
scores for the classification of heart failure, while other of England Screening study: a population based study. Lancet
studies used clinical definitions or relied on hospital dis- 2001;358:439–44.
6. Cowie MR, Wood DA, Coats AJS et al. Incidence and aetiology of
charge codes. Therefore, a large part of the differences
heart failure. A population-based study. Eur Heart J 1999;20:421–8.
between studies may be explained by varying criteria. 7. Gottdiener JS, Arnold AM, Aurigemma GP et al. Predictors of
Besides a baseline screening in the majority of partici- congestive heart failure in the elderly: the Cardiovascular Health

Downloaded from [Link] by guest on 14 September 2023

pants using a validated score, we applied clinical criteria Study. J Am Coll Cardiol 2000;35:1628–37.
for heart failure throughout the Rotterdam Study, based 8. Remme WJ, Swedberg K. Guidelines for the diagnosis and treatment
of chronic heart failure. Task Force for the diagnosis and treatment
upon guidelines of the European Society of Cardiology. of chronic heart failure, European Society of Cardiology. Eur Heart J
Apart from hospital discharge letters, medical records 2001;22:1527–60.
from general practitioners were available for assessment 9. Hofman A, Grobbee DE, de Jong PTVM et al. Determinants of disease
of cases. Consequently, less severe cases were also in- and disability in the elderly: the Rotterdam Study. Eur J Epidemiol
1991;7:403–22.
cluded in our study. However, some underestimation of
10. Mosterd A, Hoes AW, de Bruyne MC et al. Prevalence of heart failure
the true prevalence and incidence may have been caused and left ventricular dysfunction in the general population. Eur Heart
by the fact that old and diseased individuals were less J 1999;20:447–55.
likely to participate in the Rotterdam Study. Another lim- 11. Kalbfleisch JD, Prentice RL. The statistical analysis of failure time
itation of our study is that we did not distinguish between data. New York: Wiley; 1980.
12. Johansson S, Wallander M, Ruigómez A et al. Incidence of newly
underlying causes of heart failure. Among elderly diagnosed heart failure in UK general practice. Eur J Heart Fail
patients, systolic hypertension and cardiac hypertrophy 2001;3:225–31.
may be more important than ischaemic heart 13. Kitzman DW, Gardin JM, Gottdiener JS et al. Importance of heart
disease.8 failure with preserved systolic function in patients P 65 years of
age. Am J Cardiol 2001;87:413–9.
In conclusion, heart failure prevalence and incidence
14. Ho KKL, Pinsky JL, Kannel WB et al. The epidemiology of heart
are substantial. As age is an important risk factor for failure: the Framingham Study. J Am Coll Cardiol 1993;22(Suppl. A):
heart failure, the burden of this disease on health care 6A–13A.
systems in western societies increases as these popula- 15. Lloyd-Jones DM, Larson MG, Leip EP et al. Lifetime risk for
tions age. In individuals aged 55 years, 30% will develop developing heart failure: the Framingham Heart Study. Circulation
2002;106:3068–72.
heart failure during their remaining lifespan; i.e., almost 16. Hellermann JP, Goraya TY, Jacobsen SJ et al. Incidence of heart
one out of three individuals. Heart failure continues to failure after myocardial infarction: is it changing over time?. Am J
be a fatal disease, despite advances in treatment, with Epidemiol 2003;157:1101–7.
only 35% surviving 5 years after the first diagnosis. Pre- 17. Cowie MR, Wood DA, Coats AJS et al. Survival of patients with a new
diagnosis of heart failure: a population based study. Heart
vention of the development of heart failure in high-risk
2000;83:505–10.
patients is therefore fundamental. 18. 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.
19. Croft JB, Giles WH, Pollard RA et al. Heart failure survival among
References older adults in the United States. Arch Intern Med 1999;159:505–10.
20. Blackledge HM, Tomlinson J, Squire IB. Prognosis for patients newly
1. Levy D, Kenchaiah S, Larson MG et al. Long-term trends in the admitted to hospital with heart failure: survival trends in 12,220
incidence of and survival with heart failure. N Engl J Med index admissions in Leicestershire 1993–2001. Heart 2003;
2002;347:1397–402. 89:615–20.