Summary
Background
National population-based management and outcome data for patients of all ages hospitalized for heart failure have rarely been reported.
Aim
National population-based management and outcome of patients of all ages hospitalized for heart failure have rarely been reported. The present study reports these results, based on 77% of the French population, for patients hospitalized for the first time for heart failure in 2009.
Methods
The study population comprised French national health insurance general scheme beneficiaries hospitalized in 2009 with a principal diagnosis of heart failure, after exclusion of those hospitalized for heart failure between 2006 and 2008 or with a chronic disease status for heart failure. Data were collected from the national health insurance information system (SNIIRAM).
Results
A total of 69,958 patients (mean age, 78 years; 48% men) were studied. The hospital mortality rate was 6.4%, with 1-month, 1-year and 2-year survival rates of 89%, 71% and 60%, respectively. Heart failure and all-cause readmission-free rates were 55% and 43% at 1 year and 27% and 17% at 2 years, respectively. Compared with a reference sample of 600,000 subjects, the age- and sex-standardized relative risk of death was 29 (95% confidence interval [CI] 28–29) at 2 years, 82 (95% CI 72–94) in subjects aged < 50 years and 3 (95% CI 3–3) in subjects aged ≥ 90 years. For subjects aged < 70 years who survived 1 month after discharge, factors associated with a reduction in the 2-year mortality rate were: female sex; age < 55 years; absence of co-morbidities; and use of angiotensin-converting enzyme inhibitors or angiotensin receptor blockers, beta-blockers, lipid-lowering agents or oral anticoagulants during the month following discharge. Poor prognostic factors were treatment with a loop diuretic before or after hospitalization and readmission for heart failure within 1 month after discharge.
Conclusions
This large population-based study confirms the severe prognosis of heart failure and the need to promote the use of effective medications and management designed to improve survival.
Résumé
Contexte
La prise en charge et le devenir de patients hospitalisés pour insuffisance cardiaque (IC) à un niveau national et tous âges confondus, est rarement rapporté.
Objectif
C’est le cas de cette étude sur 77 % de la population française pour des patients avec une première hospitalisation pour IC en 2009.
Méthodes
Parmi les bénéficiaires du régime général de l’Assurance maladie hospitalisés en 2009 avec un diagnostic principal d’IC ont été exclus ceux hospitalisés pour IC entre 2006 et 2008 ou avec une affection de longue durée pour IC. Les données utilisées étaient celles présentes dans le système d’information de l’Assurance maladie (SNIIRAM).
Résultats
Au total, 69 958 patients ont été inclus (âge moyen 78 ans, 48 % d’hommes). Leur taux de décès hospitalier était de 6,4 %. Leurs taux de survie étaient de 89 % à un mois, 71 % à un an et de 60 % à deux ans. Ceux sans réhospitalisation pour IC étaient de 55 % à un an et de 43 % à deux ans et sans réhospitalisation toutes causes respectivement de 27 % et 17 %. Comparativement à un échantillon permanent par tirage au sort de 600 000 assurés, leur risque relatif de décès à 2 ans, standardisé sur l’âge et le sexe, était de 29 (95 % CI 28–29) et chez les moins de 50 ans de 82 (95 % CI 72–94) et de 3 (95 % CI 3–3) chez ceux de 90 ans et plus. Pour les moins de 70 ans ayant survécu un mois après leur sortie, les facteurs positifs associés au décès à deux ans étaient le sexe féminin, l’âge inférieur à 55 ans, l’absence de comorbidités, une consommation le mois de sortie d’IEC-sartan, de bêtabloquant, d’hypolipémiant, d’anticoagulant oral. Parmi les facteurs péjoratifs, il était retrouvé un traitement par diurétique de l’anse avant ou après hospitalisation et une réhospitalisation pour IC le mois de sortie.
Conclusion
Cette étude sur une large population confirme la sévérité du pronostic de l’IC, la nécessité de favoriser l’utilisation de médicaments bénéfiques et de prises en charge limitant les réhospitalisations.
Background
Heart failure is a disabling and complex syndrome, with a high prevalence (2% of the Western population, 10% after the age of 75 years) that is still increasing, together with the associated healthcare consumption . Nevertheless, standardized hospitalization rates for heart failure are globally tending to decline in several countries, such as France, but the readmission rate remains high . In the USA, the proportion of Veterans readmitted to hospital within 30 days following discharge was 5.6% for heart failure and 22% for all causes (2002–2006) – similar to the 23% readmission rate for Medicare beneficiaries (2004–2006) . In France, 21% of all patients hospitalized for heart failure in 2008 were readmitted to hospital at least once for heart failure during the same calendar year. Early readmissions for heart failure have multiple causes, and therapeutic patient education and management programmes are recommended at the time of discharge from hospital . An improvement in survival has been reported, particularly after a first hospitalization, but the prognosis nevertheless remains poor . Reported survival rates vary according to the period and the mode of patient selection, but also to sociodemographic and clinical case mixes . Patients with heart failure have a high mortality rate compared with the general population, but this has rarely been compared by sex and age group in large populations . Clinical trials have demonstrated in specific age groups the benefit of a number of drugs in terms of survival, such as angiotensin-converting enzyme inhibitors (ACEIs), angiotensin receptor blockers (ARBs) and beta-blockers, even if their use is suboptimal .
This observational study in patients of all ages derived from a large comprehensive population (77% of the French population) hospitalized for the first time for heart failure in 2009 and followed for 2 years, aimed to determine the readmission and mortality rates compared with a reference population and to assess factors associated with mortality among survivors after hospital discharge.
Methods
Information system and population
In France, the Système national d’information inter-régimes de l’assurance maladie (SNIIRAM; French national interscheme health insurance information system) comprises an individual and anonymous database concerning the beneficiaries of the various schemes . The database comprehensively records all outpatient prescriptions, services and procedures performed and reimbursed, together with their dates. Historical data are limited to a period of 3 years plus the current year. Medications are identified according to Anatomical Therapeutic Classification (ATC) code and laboratory tests are identified from the French national laboratory test coding table. Although the database does not contain any clinical information concerning the results related to prescriptions or examinations, it nevertheless indicates the presence of any “ affections de longue durée ” (ALD; chronic diseases), eligible for 100% reimbursement of healthcare expenditure, such as cardiovascular diseases, after application by the attending physician and approval by the national health insurance consultant physician. An anonymous unique identification number for each subject links this information to data collected by the national hospital discharge database (PMSI: Programme de médicalisation des systèmes d’information ) in public and private healthcare institutions. During the patient’s stay, principal diagnoses (PDs), associated diagnoses (ADs) and any ALD chronic diseases are coded according to the tenth International Classification of Diseases (ICD-10). In parallel, this database also provides a permanent open randomized sample ( Échantillon Permanent des Bénéficiaires ) representing 1/97th of all national health insurance schemes beneficiaries (600,000 subjects) and one of the main advantages of this sample is the exhaustiveness of survival data.
The study population was limited to beneficiaries of the national health insurance general scheme (77% of the 64 million inhabitants in France in 2009), excluding those covered by local mutualist sections (students, civil servants, etc.), the Mutualité Sociale Agricole , the Régime Social des Indépendants and other specific health insurance. The reasons for limiting the study to this population were the availability of vital status and date of death from the National Institute for Statistics and Economic Studies (INSEE) for the general scheme in the SNIIRAM and the lack of completeness for some SNIIRAM characteristics in some other schemes. To optimally target first hospitalizations for heart failure, patients hospitalized with a PD of heart failure (ICD-10 code I50) in 2009, but not between 2006 and 2008, and also not presenting an ALD status for heart failure were selected for this study. Patients hospitalized in 2008 with an AD of heart failure were also excluded. Hospitalizations with the following PDs were not included: I11.0 hypertensive heart disease; I13.0 hypertensive heart and renal disease with (congestive) heart failure; I13.2 hypertensive heart and renal disease with (congestive) heart failure and renal failure; I13.9 hypertensive heart and renal disease, unspecified; K76.1 chronic passive congestion of liver; and J81 pulmonary oedema.
Definitions and statistical analysis
Patient characteristics and management
The search for co-morbidities before the first hospitalization for heart failure took into account the presence of specific chronic diseases identified by ICD-10 codes. Drug treatments were identified by the presence of at least three reimbursements during the 6 months preceding the index hospitalization and by a single reimbursement during the 30 days following discharge among patients still alive. The absence of all-class medications reimbursements during these two periods was also studied in order to identify, especially among the oldest patients, those living in an institution that manages medicinal products directly and for whom no information on drug consumption is therefore recorded in the SNIIRAM database. Reimbursements for hospital outpatient clinic and office visits were used to identify cardiology consultations. As echocardiography is not systematically coded during a cardiology visit, because the levels of reimbursement may be similar for a visit with or without echocardiography, a variable was constructed by combining the existence of at least one visit to the cardiologist or echocardiography.
Readmission, survival and risk factors associated with mortality
Survival rates were calculated for the overall patient population (including hospital deaths). The 2-year mortality was compared with that of the Échantillon Permanent des Bénéficiaires sample, with standardization according to sex and age, and calculation of relative risk (RR) and 95% confidence intervals (CIs). Readmissions were taken into account only when they occurred at least 7 days after discharge, in order to avoid considering early transfers. Readmissions were studied according to the presence of a PD or AD of heart failure or an all-cause PD. Readmission-free survival rates (all-cause or heart failure) were also calculated. Readmission rates for heart failure were also calculated with censoring of deaths, i.e. based on all patients still alive at the end of the period considered. The presence of at least one readmission – or the first readmission – was reported according to the principal diagnoses coded and was grouped according to the main chapters of ICD-10 for patients still alive at 2 years.
To evaluate the available factors possibly associated with survival, the first step in the study consisted of descriptive analysis of the patients surviving 30 days after discharge and presenting at least one all-class medicinal product reimbursement 6 months before and 1 month after hospitalization. Univariate and then multivariable Cox models were used to identify factors independently associated with mortality between 1 month and 24 months after discharge, with calculation of the hazard ratio (HR). Two groups were considered (patients aged < 70 years and patients aged ≥ 70 years) in view of the age-related differences in patient characteristics, especially the proportions of the two types of heart failure. The standardized survival of these patients was also compared with that of patients excluded because of the absence of medicinal product reimbursement, who were likely to be institutionalized. SAS version 4.3 software was used (SAS Institute Inc, Cary, NC, USA).
Results
For the overall population of general scheme beneficiaries, 152,601 hospitalizations for heart failure were identified in 2009, corresponding to 130,333 patients; 69,958 (53%) of these patients were hospitalized for heart failure for the first time (48% men; mean age, 78 ± 3 years).
Survival
One and 2-year survival rates were 71% and 60%, respectively. One and 2-year survival rates without readmission for heart failure after hospital discharge were 55% and 43%, respectively, and without all-cause readmission were lower, at 27% and 17%, respectively ( Table 1 ). First readmission rates for heart failure among those patients still alive at the end of the period considered were 5% at 1 month and 14% at 6 months and then stabilized to reach 16% at 2 years. Readmission rates for heart failure were higher in youngest patients, who also had the highest survival rates.
| Time since discharge (months) | ||||
|---|---|---|---|---|
| 1 | 6 | 12 | 24 | |
| Survival a (%) | ||||
| All ages | 89.2 | 78.3 | 70.8 | 59.7 |
| < 55 years | 96.3 | 91.9 | 88.8 | 84.6 |
| 55–69 years | 94.7 | 88.6 | 84.0 | 77.1 |
| 70–79 years | 92.6 | 83.7 | 77.4 | 68.0 |
| 80–89 years | 87.2 | 74.6 | 65.7 | 52.3 |
| ≥ 90 years | 78.5 | 60.6 | 49.7 | 34.3 |
| Survival without readmission for heart failure (%) | ||||
| Total | 84.6 | 64.7 | 54.8 | 42.6 |
| < 55 years | 91.5 | 77.5 | 70.0 | 64.5 |
| 55–69 years | 89.4 | 73.9 | 64.9 | 56.4 |
| 70–79 years | 87.8 | 68.9 | 59.5 | 48.5 |
| 80–89 years | 82.8 | 61.6 | 51.0 | 36.9 |
| ≥ 90 years | 75.0 | 49.6 | 38.8 | 24.4 |
| Survival without readmission for all causes (%) | ||||
| Total | 73.1 | 38.7 | 26.8 | 16.7 |
| < 55 years | 75.4 | 40.3 | 32.2 | 26.2 |
| 55–69 years | 74.1 | 39.5 | 28.5 | 19.9 |
| 70–79 years | 74.2 | 39.2 | 27.3 | 17.0 |
| 80–89 years | 73.0 | 39.1 | 26.4 | 15.2 |
| ≥ 90 years | 68.6 | 34.3 | 22.5 | 11.6 |
| At least one readmission for heart failure without death (%) | ||||
| Total | 4.6 | 13.6 | 15.8 | 16.5 |
| < 55 years | 4.8 | 14.4 | 16.4 | 18.5 |
| 55–69 years | 5.3 | 14.7 | 17.5 | 19.4 |
| 70–79 years | 4.9 | 14.8 | 17.5 | 19.5 |
| 80–89 years | 4.4 | 13.0 | 15.2 | 15.2 |
| ≥ 90 years | 3.5 | 11.1 | 12.0 | 12.4 |
Mortality
The 2-year mortality rate was higher than that in a random sample of about 600,000 beneficiaries: 40% for all ages combined and 45.5% for patients aged ≥ 75 years ( Fig. 1 ). The age- and sex-standardized RR of death was 29 (95% CI 28–29) for all patients hospitalized for heart failure, 31 (95% CI 30–32) for women and 27 (95% CI 26–28) for men. The RR was higher among the youngest patients and decreased with age: < 50 years, RR 82 (95% CI 72–94); 50–59 years, RR 17 (95% CI 15–19), 60–69 years, RR 12 (95% CI 11–13); 70–79 years, RR 7 (95% CI 7–7); 80–89 years, RR 4 (95% CI 4–4); ≥ 90 years, RR 3 (95% CI 3–3).
Readmissions
The rate of at least one all-cause readmission among survivors was 69% and remained relatively stable according to age ( Table 2 ). A diagnosis of heart failure was reported for 24% of first readmissions and another cardiovascular disease was reported for 22%, i.e. 46% overall for all cardiovascular diseases. The proportion of first readmissions for heart failure increased with age, whereas it decreased for other cardiovascular diseases. The proportion of readmissions for diseases of the respiratory system, symptoms and injury also increased according to age.
| Age (years) | All | |||||
|---|---|---|---|---|---|---|
| < 55 | 55–69 | 70–79 | 80–89 | ≥ 90 | ||
| Events | ||||||
| Total number | 4005 | 10 731 | 17 600 | 28 888 | 8734 | 69 958 |
| Readmission a with or without death (%) | 68.0 | 72.1 | 73.0 | 69.0 | 60.0 | 69.0 |
| Death without readmission a (%) | 5.8 | 8.0 | 10.0 | 15.8 | 28.4 | 14.3 |
| Readmission a or death (%) | 73.8 | 80.1 | 83.0 | 84.8 | 88.4 | 83.3 |
| Principal diagnosis at the first readmission | ||||||
| Total number | 2721 | 7733 | 12 841 | 19 918 | 5237 | 48 306 |
| Cardiovascular diseases (%) | 55.2 | 52.8 | 46.9 | 43.4 | 42.9 | 46.3 |
| Heart failure | 20.9 | 20.5 | 22.6 | 26.2 | 30.2 | 24.5 |
| Angina pectoris | 1.5 | 2.4 | 2.2 | 1.4 | 1.1 | 1.7 |
| Acute myocardial infarction | 0.7 | 0.8 | 0.8 | 0.9 | 1.1 | 0.9 |
| Chronic ischaemic heart disease | 3.2 | 4.7 | 3.4 | 1.5 | 0.4 | 2.4 |
| Cardiomyopathy | 8.8 | 4.2 | 1.4 | 0.5 | 0.2 | 1.7 |
| Atrial fibrillation and atrial flutter | 6.0 | 7.3 | 4.6 | 2.7 | 1.4 | 4.0 |
| Cerebral infarction | 0.4 | 0.5 | 0.8 | 1.2 | 1.4 | 0.9 |
| Other cardiovascular diseases | 13.7 | 12.4 | 11.1 | 9.0 | 7.1 | 10.2 |
| Tumours (%) | 4.2 | 6.9 | 7.5 | 7.0 | 5.5 | 6.8 |
| Endocrine diseases (%) | 4.4 | 4.4 | 3.5 | 2.6 | 2.2 | 3.1 |
| Diseases of the nervous system (%) | 2.6 | 2.8 | 2.5 | 1.9 | 1.4 | 2.2 |
| Diseases of the eye and adnexa (%) | 0.4 | 2.3 | 3.7 | 3.4 | 1.5 | 2.9 |
| Diseases of the respiratory system (%) | 5.0 | 6.2 | 8.1 | 9.9 | 13.1 | 9.0 |
| Diseases of the digestive system (%) | 4.5 | 4.1 | 5.1 | 4.8 | 4.7 | 4.7 |
| Diseases of the genitourinary system (%) | 4.0 | 3.9 | 4.3 | 4.0 | 3.2 | 4.0 |
| Symptoms, signs and abnormal clinical and laboratory findings (%) | 5.7 | 5.4 | 6.6 | 8.0 | 9.2 | 7.3 |
| Injury (%) | 2.2 | 2.2 | 3.4 | 5.9 | 7.7 | 4.7 |
| Factors influencing health status (%) | 3.7 | 3.8 | 2.7 | 2.1 | 1.5 | 2.6 |
| Other (%) | 8.1 | 5.2 | 5.7 | 7.0 | 7.1 | 6.4 |
| Total (%) | 100.0 | 100.0 | 100.0 | 100.0 | 100.0 | 100.0 |
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
