Summary
Background
No clinical practice guidelines are available for the treatment of heart failure (HF) in patients with preserved left ventricular ejection fraction (LVEF).
Aims
To determine how cardiologists manage medical treatment in HF patients after hospital discharge, according to LVEF.
Methods
The FUTURE study was a cross-sectional survey conducted in HF outpatients by French private cardiologists between September 2007 and August 2008. Patients had to have been hospitalized within the previous 18 months with a diagnosis of HF. Clinical data and HF treatments (angiotensin-converting enzyme inhibitors [ACEIs], angiotensin receptor blockers [ARBs], beta-blockers, diuretics and aldosterone antagonists) were recorded retrospectively, with precise information on drug doses, at two successive time points (at hospital discharge and at the index consultation). HF treatment was compared in patients with reduced (less than or equal to 40%) versus preserved (more than 40%) LVEF.
Results
Completed data were available for 1137 HF patients enrolled by 424 cardiologists. Mean patient age was 72 ± 11 years; LVEF was reduced in 56% and preserved in 44%. The therapeutic approach was similar in the two groups, both at hospital discharge and at the index consultation. At the index consultation, HF treatment was: beta-blocker (74%); ACEI/ARB (83%); loop diuretic (86%); aldosterone antagonist (31%). The majority of patients (62%) received a beta-blocker plus an ACEI or an ARB; 56% reached more than or equal to 50% of the target dose for each treatment. There were no major differences in treatments and dosages between the groups with low and preserved LVEF. In 15% of cases where the drug dose was not increased, fear of adverse events was reported as the reason.
Conclusion
The FUTURE survey showed a similar approach to HF treatment irrespective of LVEF. Compared with previous studies, we saw an improvement in the use of recommended HF drugs, especially beta-blockers. However, achievement of target doses could be improved.
Résumé
Buts
Il n’y a pas de recommandations scientifiques bien décrites auxquelles le cardiologue peut se référer en ce qui concerne le traitement de l’insuffisance cardiaque à fraction d’éjection préservée. Le but de cette étude a été de déterminer la façon dont les cardiologues prennent en charge le traitement médical des patients en fonction de leur FEVG après la sortie de l’hôpital.
Méthodes
FUTURE est un registre transversal réalisé sur des patients insuffisants cardiaques ambulatoires suivis par des cardiologues libéraux entre septembre 2007 et Août 2008. Les patients devaient avoir été hospitalisés au cours des 18 mois précédents, avec un diagnostic d’insuffisance cardiaque. Les données cliniques et le traitement médical (IEC, ARA2, bêtabloquants, diurétiques et antagonistes de l’aldostérone) ont été relevés à deux moments successifs (à la sortie de l’hôpital et à la consultation d’entrée dans l’étude) avec des informations précises sur les doses de médicaments.
Résultats
Des données complètes ont été obtenues chez 1137 patients insuffisants cardiaques par 424 cardiologues. L’âge moyen était de 72 ± 11 ans. La FEVG était abaissée (inférieure ou égale à 40 %) chez 56 % et préservée (supérieure à 40 %) chez 44 % des patients. À la sortie de l’hôpital aussi bien qu’à la consultation d’entrée dans l’étude, l’approche thérapeutique est apparue similaire dans les deux groupes. À la consultation d’entrée, le traitement était : bêtabloquants (74 %), IEC ou ARA2 (83 %), diurétique de l’anse (86 %) et antagonistes de l’aldostérone (31 %). La majorité des patients (62 %) recevait une combinaison de bêtabloquant et d’IEC ou ARA2 et 56 % avait atteint au moins 50 % de la dose cible pour chacun des traitements. Il n’y avait pas de différence majeure dans les traitements et dans les doses entre les deux groupes de patients avec FEVG basse ou préservée. Dans 15 % des cas où la dose de médicament n’a pas été augmentée, l’explication rapportée était une crainte des effets secondaires.
Conclusion
Quelle que soit la FEVG, FUTURE montre que l’approche du traitement de l’insuffisance cardiaque est similaire. Par rapport aux études antérieures, il est noté une amélioration dans l’utilisation des médicaments recommandés dans l’insuffisance cardiaque, en particulier les bêtabloquants. Toutefois, une amélioration est encore possible en ce qui concerne les doses administrées.
Background
HF is a costly public health problem that increases in prevalence with age in the populations of western countries . Despite advances in medical strategies in the treatment of myocardial infarction or hypertension, the proportion of patients presenting with HF is increasing, with an estimated 500 to 750,000 HF patients in France , at least 10% of whom are elderly subjects (aged over 80 years). HF is a serious and potentially life-threatening disease that impairs quality of life. Median survival after diagnosis of HF does not exceed 4 years.
Although guidelines and recommendations for the treatment of chronic HF are based on clinical trial results , under “real-life” conditions, affected populations are often very different from patients enrolled in clinical trials : they are generally older and more often have a preserved LVEF. The mean age of patients admitted to hospital for HF in France in 2000 was 77 years . The “OFICA registry” , which records data on patients hospitalized for HF in France, shows that one-half of affected patients are aged over 80 years. In addition, a sizable number of hospitalized patients have preserved LVEF. The threshold used to define impaired and preserved LV function (35%, 40%, 45% or 50%) is not clearly defined. Nevertheless, the prognosis of HFpEF seems to be slightly better than that of HFrEF . Whether treatment differs between patients with HFpEF and those with HFrEF remains unclear. The extent to which cardiologists continue and/or improve the treatment strategy initiated in hospital is unknown. It would also be interesting to know if cardiologists are more reluctant to optimize treatment in patients with HFpEF. These are some of the questions that the FUTURE survey set out to answer.
The primary objective of the FUTURE survey was to better describe the medical treatment of HF after hospital discharge, according to LVEF. Several studies of HF have been conducted previously in France (IMPROVEMENT , the National French Cardiologists survey , IMPACT-RECO 1 and 2 and the National General Hospitals registry ), showing increasing adherence to guidelines in current practice. However, none of these studies compared groups of HFrEF and HFpEF patients or analysed the prescriptions made at the time of hospital discharge and afterwards in HF outpatients.
Methods
The FUTURE study was a cross-sectional survey with retrospective data collection, conducted in HF outpatients by French cardiologists in private practice, between September 2007 and August 2008. Physicians were instructed to enroll the first four patients who satisfied the inclusion criteria. To fulfill these criteria, patients had to have been hospitalized within the previous 18 months with a diagnosis of worsening HF and not to be currently followed by the participating cardiologist. Patients who were participating in any other clinical trial during the study period were not authorized to participate in the survey. Neither age nor LVEF was an inclusion criterion.
Cardiologists were asked to list comorbidities, clinical features of HF at the time of the last hospitalization, the treatment prescribed at hospital discharge, the patient’s current clinical status and ongoing treatment. Clinical history and medical treatment(s) were recorded using the patient files from the cardiologists’ offices. Data concerning the last hospitalization for HF were collected retrospectively.
The CNIL (the French Commission on Data Processing, Data Files and Individual Liberties) gave its consent to the collection of patient data.
As the aim of the present study was to describe the influence of private cardiologists’ practice on treatment strategy, the analysis of treatments was restricted to HF outpatients who visited their cardiologist at least once since hospital discharge and before the date of baseline data collection (index consultation). Patients were divided into two groups (HFrEF and HFpEF) according to LVEF at hospital discharge (less than or equal to 40%, reduced LVEF; more than 40%, preserved LVEF). Treatment of HF (drug[s], dosage and target dose) was described according to the 2008 ESC guidelines for HF with reduced LVEF (ACEIs [except for perindopril, which is not recommended for HF but was accepted because it is widely used in France at a target dose of 4 mg/day] or ARBs, beta-blockers, loop diuretics and aldosterone antagonists, either alone or in combination).
Statistical analysis
Statistical analysis was performed with SAS 9.1 software (SAS Institute, Cary, NC, USA). Results are presented as means ± standard deviations for continuous variables and as numbers and percentages for dichotomous variables. All tests were two-sided and the alpha risk was set at 0.05. The influence of demographic and clinical variables, as well as the characteristics of each cardiologist in terms of the probability of prescribing each therapeutic class, was modelled by logistic regression. The prevalence of HF with preserved LVEF was estimated with its two-sided 95% confidence interval, assuming a binomial distribution. Between-group comparisons were performed using an unpaired Student’s t test or the Mann–Whitney U test for continuous data and a Chi 2 test or Fisher’s exact test for categorical variables. Some comparisons were adjusted for age using two-way analysis of variance.
Results
A total of 1673 patients with a diagnosis of HF established during a previous hospitalization were enrolled by 424 cardiologists between September 2007 and August 2008. Complete data were available for 1408 patients at hospital discharge; 221 patients were excluded from the analysis, mainly because they had not been hospitalized within the previous 18 months. LVEF at hospital discharge was not available for 44 patients. A total of 1137 patients attended at least one visit with their cardiologist between hospital discharge and the index consultation. Therefore, patient characteristics according to LVEF were described in 1408 patients, whereas the course and outcome of clinical variables and treatments were analysed on a subset of 1137 patients.
Patients were enrolled in the survey 5.8 ± 4.5 months after hospital discharge after a mean number of 2.3 ± 1.7 visits to the cardiologists. Average time between hospital discharge and the index consultation was less than 3 months for 403 patients, 3–6 months for 291 patients, 6–12 months for 286 patients and more than 12 months for 157 patients.
Characteristics and clinical variables at hospital discharge
For the 1408 patients analysed at hospital discharge, the mean age was 72 ± 11 years, 65% were aged over 70 years, 67% were men and the mean LVEF was 40 ± 13%. The cause of HF was ischaemic heart disease in 727 patients (52%), arterial hypertension in 582 (41%), valvular heart disease in 219 (16%), and was associated with primary dilated cardiomyopathy in 462 (33%) (more than one cause of HF could be reported in a given patient). Comorbidities were frequent and included hypertension (65%), diabetes (31%), renal failure (35%), asthma/COPD (18%) and anaemia (8%). Reasons for hospitalization were a first episode of acute HF (“de novo HF” according to the 2008 ESC classification ) in 44% of patients and worsening of known HF in 55%. In 33% of patients, the initial clinical presentation of HF on admission to hospital was acute pulmonary oedema. At hospital discharge, 69% patients had sinus rhythm as seen on their electrocardiogram.
LVEF was reduced (less than or equal to 40%) in 792 patients (56%) and preserved (more than 40%) in 616 (44%). Some characteristics ( Table 1 ) were more frequent in HFpEF patients than in HFrEF patients: the former were older and more often were female, overweight or obese. HF had been diagnosed more recently and was more often of hypertensive or cardiac valvular origin and anaemia was more frequent. Conversely, plasma concentrations of B-type natriuretic peptide were significantly more elevated and renal failure was more frequent in patients with reduced LVEF. However, no difference was observed between the two groups regarding some comorbidities: ischaemic origin of HF, history of stroke, asthma/COPD and diabetes. Based on LVEF, the condition of HFpEF patients was less serious at hospital discharge ( Table 2 ).
| HFrEF ( n = 792; 56%) | HFpEF ( n = 616; 44%) | P | |
|---|---|---|---|
| Age (years) | 71 ± 12 | 74 ± 11 | < 0.0001 |
| Men | 583 (74) | 355 (58) | < 0.0001 |
| BMI (kg/m 2 ) | 26.5 ± 4.9 | 26.8 ± 4.8 | 0.23 |
| Overweight/obese | 308 (39) | 285 (47) | 0.004 |
| Time since diagnosis of HF (years) | 3.9 ± 4.9 | 2.5 ± 3.5 | < 0.0001 |
| HF of ischaemic origin | 422 (53) | 305 (50) | 0.16 |
| HF of hypertensive origin | 235 (30) | 347 (56) | < 0.0001 |
| HF of valvular origin | 95 (12) | 124 (20) | < 0.0001 |
| Dilated cardiomyopathy | 348 (44) | 114 (19) | < 0.0001 |
| HF of mixed origin | 273 (34) | 259 (42) | 0.004 |
| History of stroke | 66 (8) | 56 (9) | 0.62 |
| Renal failure | 295 (37) | 192 (31) | 0.02 |
| Asthma/COPD | 146 (18) | 112 (18) | 0.90 |
| Anaemia | 56 (7) | 63 (10) | 0.03 |
| Diabetes | 237 (30) | 201 (33) | 0.28 |
| Hypertension | 441 (56) | 472 (77) | < 0.0001 |
| Hypercholesterolaemia | 439 (55) | 355 (58) | 0.41 |
| Current smokers | 286 (36) | 164 (27) | 0.0002 a |
| Excessive drinkers | 148 (19) | 57 (9) | < 0.0001 b |
| Sinus rhythm c | 476 (71) | 309 (66) | 0.06 |
| Plasma creatinine (μmol L −1 ) | 106 ± 29 | 101 ± 31 | 0.016 |
| Plasma haemoglobin (g/dL) | 12.6 ± 1.7 | 12.6 ± 1.7 | 0.86 |
| Plasma BNP (pg/mL) c,d | 370 ± 1006 | 243 ± 591 | < 0.0001 |
| CRT | 110 (14.3) | 24 (4.0) | < 0.0001 |
| ICD | 87 (11.3) | 12 (2.0) | < 0.0001 |
| Cardiac pacemaker | 178 (22.4) | 90 (14.6) | 0.0002 |
a P = 0.04: comparison adjusted for sex.
b P = 0.0002: comparison adjusted for sex.
| Hospital discharge | Index consultation | |||||
|---|---|---|---|---|---|---|
| HFrEF ( n = 661) | HFpEF ( n = 476) | P | HFrEF ( n = 661) | HFpEF ( n = 476) | P | |
| NYHA I | 35 (5) | 44 (9) | <0.0001 | 84 (13) | 87 (18) | <0.0001 |
| NYHA II | 344 (53) | 311 (66) | 396 (60) | 315 (67) | ||
| NYHA III | 226 (35) | 106 (22) | 158 (24) | 65 (14) | ||
| NYHA IV | 43 (7) | 13 (3) | 19 (3) | 5 (1) | ||
Change between hospital discharge and index consultation
Functional status was analysed in 1137 patients who visited the participating cardiologist between hospital discharge and the index consultation. At hospital discharge, most patients were in class II and III of the NYHA classification (58% in class II, 29% in class III, 7% in class I and 5% in class IV). At the index consultation, functional status was improved in all patients, regardless of ejection fraction (15% in NYHA class I, 63% in class II, 20% in class III, 2% in class IV; P < 0.0001). Based on LVEF, the functional status of HFpEF patients remained significantly better ( Table 2 ). In addition, systolic blood pressure was higher in HFpEF patients compared with in HFrEF patients (134 ± 16 mmHg and 124 ± 17 mmHg, respectively; P < 0.0001). Heart rate was only available at the index consultation and was 72.7 ± 16.3 beats/min and 72.6 ± 15.8 beats/min in the HFrEF and HFpEF groups, respectively (difference not significant).
Therapeutic management
At hospital discharge
At the time of hospital discharge, 71% of 1137 patients were receiving one of the four recommended beta-blockers (carvedilol, bisoprolol, metoprolol CR XL or nebivolol), 82% were receiving an ACEI or an ARB, 87% were receiving a loop diuretic and 29% were receiving an aldosterone antagonist. Half of the patients (52.8%) received treatment combining a beta-blocker, an ACEI or an ARB and a diuretic. About one-third of patients (37%) treated with the combination of a beta-blocker and an ACEI or an ARB received more than or equal to 50% of the target dose for each treatment but only 5% of these patients were at the target dose. The target doses of HF treatment were seldom prescribed at hospital discharge, thus allowing cardiologists to titrate treatment further in the subsequent months. Other drugs were frequently prescribed: cardiac glycosides were given to 17% of patients, vitamin K antagonists to 43%, antiplatelet agents to 57%, statins to 60% and calcium channel blockers to 16%. Only 17% of patients were transferred to a convalescent ward or hospital and 83% returned home.
According to the LVEF, regardless of the HF treatment prescribed (ACEI and/or ARB, beta-blocker, loop diuretic or aldosterone antagonist), these therapies were prescribed significantly less frequently in HFpEF patients but the dose reached (more than or equal to 50% of target dose or target dose) was not different between the two groups, except for beta-blockers ( Table 3 ). Differences were also found for non-drug therapies: a cardiac rehabilitation programme was initiated less often in HFpEF patients (13% vs 24%; P < 0.0001); stays in a convalescent ward were less frequent in HFpEF patients (13% vs 19%; P = 0.003); and specific educational information was given to fewer HFpEF patients (35% vs 48%; P < 0.0001). Only a few patients (8%) joined an ambulatory health network combining private and hospital physicians (11% of HFrEF patients and 3% of HFrEF patients; P < 0.0001).
