The ICREX-94 research was a prospective, non-interventional, observational, transversal, multicentric registry conducted in seven cardiology units and three cardiogeriatrics units in the Val-de-Marne department (zip code 94) in France.
The present study was conducted in the Val-de-Marne department: 245 sq. km, 1.4 million inhabitants, a mix of residential cities and low-incomes cities (mean HDI 0.58, max 0.78, min 0.35). The Val-de-Marne health system comprises 48 hospitals, with a total capacity of 9500 beds. In 2018, in the 10 hospitals participating in this study, 2393 heart failure admissions were recorded by the “Caisse Primaire d’Assurance Maladie” (French health insurance fund), representing 85% of all AHF admissions in Val-de-Marne. In 2016, ten Val-de-Marne cardiology and cardiogeriatrics departments, academic and non-academic, public and private, large and small, interested in HF care, decided to create a network (FINC94) and to collaborate, in order to share their experiences, train healthcare professionals and conduct clinical studies such as this one (20).
There were seven classical cardiology departments, found in both teaching and nonteaching hospitals, public and private, and three cardiogeriatrics departments specialized in heart failure management, with geriatricians who had received specific academic and practical training in cardiovascular medicine. In addition to their geriatrics background, the geriatricians in these units trained for several months in cardiology departments specialized in HF, and have university diplomas in echocardiography and cardiovascular disease of elderly patients. Therefore, they work in close cooperation with the HF team of their departments. There was no specific protocol when patients were hospitalized in cardiology or cardiogeriatrics departments, except to follow 2016 ESC guidelines on HF. Upon admission to cardiogeriatrics departments, an individual and multidisciplinary approach (by geriatricians, physiotherapists, dietitians and social workers) was established, focused on stabilization of comorbidities, return to self-sufficiency and renutrition in addition to specific cardiology follow-up.
There were no specific guidelines to direct patients to a cardiology or cardiogeriatrics department.
Consecutive patients over 18 years of age, hospitalized for acute HF and alive at discharge were eligible for the study. Diagnosis of AHF was based on signs and symptoms of HF— clinical point of view, BNP at admission > 100 pg/ml and heart structure suggesting HF on echocardiograms, as recommended (ESC guidelines). Patients who did not understand the French language were excluded. The study was compliant with Helsinki rules and was approved by the local ethics committee (#20181128163709). All patients gave their informed consent.
Baseline data collection
The following data was collected at inclusion and if patients were rehospitalized: HF type (i.e., right, left, global), etiology of HF, date of diagnosis of HF, clinical characteristics including geriatric comorbidities like dementia and depression, ECG data (sinus rhythm, atrial fibrillation), and biological data such as BNP, haemoglobinemia, and serum creatinine. In addition, the human development index (HDI), which evaluates the progress of a country or a region in the long term, adapted to the Ile-de-France region, was determined by the town of residency. The HDI takes into account three basic dimensions of human development: a long and healthy life (life expectancy), access to knowledge (education) and a decent standard of living (income) (21). We recorded echocardiographic characteristics, such as left ventricular ejection fraction (LVEF), and medical treatments with respective doses and whether the patient had a multi-site and/or defibrillator pacemaker. We defined patients as “well-treated” when they had received more than 50% of the target dose of the treatment by ARB and beta blockers.
Follow-up data collection
Patients were followed over 90 days after discharge from hospital by direct phone calls and correspondence. If the patient did not answer, we called the patient’s family, caregiver, general practitioner or cardiologist. Rehospitalizations within the 90 days were recorded, with medical reports and the same clinical, ECG and biological variables as on first admission. Cause and mode of rehospitalization were analyzed by Clinical Endpoints Committees (CEC) set up to review all medical reports of rehospitalized patients. Each CEC included one cardiologist and one geriatrician trained in endpoint adjudication. All events were reviewed independently by each CEC. Any disagreement between CECs was resolved by a third physician as CEC chairman (EB, KR, LH, CD, TD).
CECs divided hospital admissions in four classes: AHF Planned Rehospitalization, AHF Unplanned Rehospitalization, Non-AHF Planned Rehospitalization, and Non-AHF Unplanned Rehospitalization. For AHF readmissions, the underlying causes were classified by CECs as follows: infection, unstable hypertension, arrhythmia, medical treatment modification, non-adherence, anaemia, myocardial infarction, pulmonary embolism, acute renal failure, very severe chronic heart failure (i.e. "frequent flyer" patients with ≥ 3 hospitalizations in the year or with NTproBNP > 5000 pg/ml).
Statistical analysis
Continuous variables are expressed as median [interquartile range (IQR)], and categorical variables are expressed as number or frequency (percentage). Differences in patient clinical characteristics between cardiology and cardiogeriatrics departments were tested by the ctwo or fisher test for categorical data and by the Wilcoxon test for continuous data.
Differences in clinical characteristics between patients hospitalized for acute heart failure and non hospitalized patients were obtained with univariate logistic regression and the Wald test.
Finally we produced a Kaplan-Meyer curve of readmission for acute heart failure within 90 days depending on the type of department and did a survival analysis using a univariate cox regression.
A two-sided p-value < 0.05 was considered statistically significant. All statistical analyses were performed using R version 4.