Summary
Background
Acute heart failure (AHF) is a life-threatening medical emergency for which no new effective therapies have emerged in recent decades. No previous study has exhaustively described the entire course of care of AHF patients from first medical contact to hospital discharge or assessed its impact on prognosis.
Aim
To fully describe the course of care and analyze its influence on outcomes in patients hospitalized with an AHF syndrome in an academic university center.
Methods
One hundred and nineteen adults with AHF from three public academic university hospitals were consecutively enrolled in a multicenter prospective observational cohort study. All of the emergency departments, intensive care units, coronary care units, cardiology wards and other medical wards participated in the study.
Results
The composite primary outcome (6-month rate of cardiovascular death, readmission for acute heart failure, acute coronary syndrome or stroke) occurred in 59% of patients. This rate was high and similar regardless of first medical contact, type of transport, first medical department of admission and number of medical departments involved in the course of care. A cardiologist was involved in management in 80% of cases. The global median hospital stay was shorter with cardiology vs non-cardiology management (7 days [interquartile range 4–11] vs 10 days [interquartile range 7–18]; P = 0.003). History of hypertension ( P = 0.004), need for non-invasive ventilation ( P = 0.023) and Lee prognostic score ( P = 0.028) were independently associated with the primary outcome.
Conclusions
Morbimortality and readmissions were high regardless of the course of care in patients admitted for AHF in real life. The reduction in hospital stay when cardiologists were involved in management encourages the creation of “mobile AHF cardiology teams”.
Résumé
Contexte
L’insuffisance cardiaque aiguë est une urgence médicale pour laquelle aucune thérapie efficace n’a émergé au cours des dix dernières années. Aucune étude n’a décrit de manière exhaustive l’ensemble du parcours de soins des patients admis pour insuffisance cardiaque aiguë ni évalué son impact sur le pronostic.
Objectif
Décrire de façon exhaustive le parcours de soins et analyser son influence sur le pronostic des patients hospitalisés pour un syndrome d’insuffisance cardiaque aiguë dans un centre universitaire académique.
Méthodes
Cent dix-neuf adultes en insuffisance cardiaque aiguë ont été inclus consécutivement dans une étude observationnelle prospective de cohorte multicentrique. Tous les services d’urgence, les unités de réanimation, les unités de soins coronaires, les services de cardiologie et tous les autres services médicaux ont participé à l’étude.
Résultats
Le critère de jugement principal composite, incluant les décès à 6 mois d’origine cardiovasculaire, les réadmissions pour insuffisance cardiaque aiguë, les syndromes coronaires aigus et les accidents vasculaires cérébraux, a été observé chez 59 % des patients. Ce taux restait élevé et similaire quel que soit le type de premier contact médical, le type de transport, le premier service d’admission, et le nombre de départements médicaux impliqués dans le parcours de soins. Un cardiologue a été impliqué dans la prise en charge des patients dans 80 % des cas. Dans ce cas la durée médiane d’hospitalisation était alors plus courte (7 jours [IQR 4–11] vs 10 jours [IQR 7–18] ; p = 0,003). Un antécédent d’hypertension artérielle ( p = 0,004), le besoin de ventilation non invasive ( p = 0,023), et le score pronostique de Lee ( p = 0,028) étaient indépendamment associés avec le critère de jugement principal.
Conclusions
La morbi-mortalité et le taux de réadmissions restent élevés quel que soit le parcours de soins des patients admis pour insuffisance cardiaque aiguë. La réduction de la durée d’hospitalisation lorsque des cardiologues sont impliqués dans la prise en charge des patients encourage la création d’« équipes mobiles d’insuffisance cardiaque aiguë ».
Background
Heart failure is a life-threatening complication of all cardiac diseases. This syndrome has a chronic evolution, but is also characterized by acute onset and exacerbation, which impair survival and quality of life. In the USA and Europe, the number of patients hospitalized at least once per year for an acute heart failure (AHF) syndrome is increasing, leading to an annual incidence of 2.5 per thousand inhabitants . AHF is a life-threatening medical emergency with high in-hospital and long-term mortality rates. One in 10 patients with AHF dies during hospitalization , and one in three dies within the year following an episode . Moreover, AHF is a major burden on the medical system and healthcare costs because it is the leading cause of admission to hospitals in industrialized countries, and the rate of readmissions for AHF is currently increasing, ranging from 30% to 45% per year . Despite advances in long-term care for heart failure, no new effective therapies for AHF have emerged in recent decades. Indeed, the management of AHF is usually difficult because of numerous clinical scenarios and the presence of several other co-morbidities, which also require non-cardiologic care. Sometimes AHF is not the main reason for admission, and can be an associated diagnosis. In this context, physicians from different specialties can manage patients with AHF, and the course of patient care can be variable ; this variability may influence outcome and hospital stay. Furthermore, recent guidelines emphasize that the “time-to-treatment” concept may be important in patients with AHF and acute coronary syndromes (ACS) . Thus, the development of an optimal course of care may be critical to improving the management of patients with AHF. Many studies have assessed the prognosis of patients with AHF . Some of these studies demonstrated that predefined care pathways – systematic plans for the care of particular patients over a particular time period – could improve outcome . However, none of these studies has exhaustively described the entire course of care, from the first medical contact to hospital discharge, or assessed its impact on prognosis.
Therefore, we conducted a prospective observational cohort study of all patients hospitalized with an AHF syndrome in an academic university center. This study aimed to fully describe the patients’ course of care, and to analyze the influence of course of care on patient outcomes.
Methods
Patients
For 3 weeks from February to March 2014, we prospectively and consecutively enrolled in a cohort all patients hospitalized with AHF at one of the three public academic university hospitals in Marseille (University Health Centre of Marseille, France). All of the emergency departments (EDs), intensive care units (ICUs), coronary care units (CCUs), cardiology wards and other medical wards agreed to participate in the study. Patients were selected if they had rapid onset or acute worsening of symptoms and signs of heart failure, associated with elevated plasma concentrations of natriuretic peptides, regardless of severity, according to the opinion of the attending investigator. The exclusion criteria were age < 18 years and case report forms (CRFs) containing too many missing data. Two local investigators (L.C. and F.T.) had to confirm the diagnosis of AHF, using all available medical records during the hospital stay, for patients to be definitively included. To ensure that a consecutive sample of all definite diagnoses was obtained, a control of quality, based on the French medico-administrative database, was performed after the inclusion period. The following ICD-10 diagnosis codes were used: I501 (pulmonary edema/left heart failure), R570 (cardiogenic shock) and I500 (right or global heart failure). All patients provided informed consent as required by the institutional review board.
Data collection
Clinical, biological and imaging data were collected at admission and during hospitalization, using a dedicated CRF. Co-morbidities were assessed using the Charlson index ( Supplementary file 1 ). The course of care before and after admission was fully described, and included first medical contact, mode of transport, medical site of first admission and the other medical sites involved thereafter. Risk stratification was assessed using the prognostic scale recommended by Lee ( Supplementary file 2 ). Cardiology management was considered when a senior cardiologist was involved in the decision regarding the diagnostic and therapeutic strategy. In each of the three EDs in the university center, an electronic CRF was created. During the coding at the end of the emergency care, as soon as the following diagnosis codes appeared as the primary or secondary diagnosis (I501, R570, I500 or R060 [dyspnea]), a warning message was displayed to remind the physician that he had to complete the CRF if the inclusion criteria were met. For all other medical sites in the center, a written CRF was completed by the physicians who managed the patient on site, and was controlled thereafter by one of the investigators (L.C.) by consulting medical records.
Follow-up and outcomes
The primary outcome was a composite, including cardiovascular death, readmission for AHF, ACS or stroke occurring within the 6 months after inclusion. Prespecified secondary outcomes included cardiovascular death, readmission for AHF within the 6 months after inclusion, and hospital stay. The outcome at 6 months was obtained by contacting the patients and the patients’ physicians.
Statistical analysis
A sample size of 114 participants was planned to have 80% power to detect a difference in change in the primary outcome between the different course of care groups, assuming a basal rate of 60%, an absolute difference of 25% and a type I error rate of 5% (two-sided).
Categorical variables are expressed as percentages and were compared using the χ 2 test or Fisher’s exact test. Quantitative variables are expressed as medians and interquartile ranges (IQRs) and were compared using the Mann-Whitney test. Pearson’s correlation coefficient was used to study the association between two quantitative variables.
Univariate analyses were performed first, to study the association between patient characteristics and outcomes. The following variables were tested: age; sex; hypertension; cigarette smoking; hyperlipidemia; diabetes; obesity; history of chronic heart failure; ischemic heart disease; atrial fibrillation; chronic obstructive pulmonary disease; dementia; stroke history; precipitating factor; heart rate; systolic arterial blood pressure; B-type natriuretic peptide; troponin I; creatinine; urea; serum sodium concentrations; arterial blood pH; cardiogenic shock; Lee prognostic scale; and type of treatment. The following variables describing the course of care were also tested: first medical contact; type of transport; medical site of first admission; number of medical sites involved during the hospitalization; hospital stay; and location where the patients were discharged. To study the variables independently associated with the primary outcome, multivariable logistic regression modeling was performed using a backward elimination procedure. All of the variables that were found to be associated with a P -value < 0.20 in the univariate analysis were tested in this model. All statistical analyses were two-tailed, and the results were considered to be statistically significant when P -values < 0.05 were obtained. The analyses were performed using IBM SPSS Statistics 20.0 (IBM Inc., Armonk, NY, USA).
Results
Baseline characteristics
During the inclusion period, 132 patients admitted with AHF were eligible for study entry. Of these patients, 13 were excluded because of CRFs containing too many missing data. Thus, data are presented for 119 consecutive patients. The baseline characteristics are summarized in Table 1 . The median age of the patients was 81 years (IQR 72–87), and 50% of the patients were men. A previous hospitalization for AHF was reported in 56% of patients. Evidence of overt underlying heart disease was reported in 86 patients (72%). Ischemic heart disease was the most frequent etiology. Infection was the most frequent precipitating factor (29%). Transthoracic echocardiography was performed during hospitalization in 89% of patients. The median left ventricular ejection fraction (LVEF) was 40% (IQR 30–55). The LVEF was > 50% in 48 patients (45%). The median B-type natriuretic peptide value was 659 pg/mL (IQR 347–1126). Sixteen (13%) patients had a Lee prognostic score > 150, which defined a group at high risk of serious complications and death. Cardiogenic shock was noted in eight patients (7%). For the treatment initiated at admission and during hospitalization, intravenous diuretics were used in 114 patients (96%), intravenous nitrates in 31 patients (26%) and inotropes in 8 patients (7%). Non-invasive ventilation was needed in 24 patients (20%), and endotracheal intubation with mechanical ventilation was needed in three patients (2.5%). Six patients (5%) required hemodialysis, and one patient required a left ventricular assist device. Finally, coronary angiography was performed in 32 patients (27%), with percutaneous coronary intervention in 20 patients (17%).
| Age (years) | 81 [72–87] |
| Men | 60 (50) |
| Previous hospitalization for AHF | 67 (56) |
| Underlying cardiovascular disease | |
| Ischemic heart disease | 40 (47) |
| Valvular heart disease | 18 (21) |
| History of hypertension | 79 (66) |
| Dilated cardiomyopathy | 13 (15) |
| Rhythm disorder | 10 (12) |
| Diabetes | 49 (41) |
| Atrial fibrillation | 47 (40) |
| COPD | 20 (17) |
| Dementia | 13 (11) |
| Stroke | 15 (13) |
| Obesity | 20 (17) |
| Charlson scale ≥ 4 | 41 (35) |
| Lee scale > 150 | 16 (13) |
| Precipitating factors | |
| Infection | 34 (29) |
| Myocardial ischemia | 18 (15) |
| Arrhythmia | 18 (15) |
| High blood pressure | 13 (11) |
| Other | 22 (18) |
| Not identified | 15 (13) |
| Type of AHF | |
| Global | 57 (48) |
| Left | 55 (46) |
| Right | 7 (6) |
| Cardiogenic shock | 8 (7) |
| LVEF (%) | 40 [30–55] |
| Admission | |
| SBP (mmHg) | 150 [128–171] |
| Heart rate (bpm) | 90 [78–110] |
| BNP (ng/L) | 659 [347–1126] |
| Troponin (μg/L) | 0.07 [0.03–0.3] |
| Creatinine (μmol/L) | 103 [79–137] |
| Urea (mmol/L) | 8.3 [6.6–12.8] |
| Sodium (mmol/L) | 140 [137–143] |
| Arterial blood pH | 7.38 [7.33–7.46] |
| Hospital stay (days) | 7 [4–13] |
| In-hospital mortality | 7 (6) |
Course of care
The first medical contact was the ED for 69 patients (58%), physician-staffed ambulances for 21 patients (18%) and general practitioners or cardiologists for 29 patients (24%). Transportation to the hospital was medical (physician-staffed ambulance) for 33 patients (28%).
After the first medical contact, 16 patients (13%) remained hospitalized for 24 hours in the observation unit of the ED, and were then transferred to another department (93%) or died (6%). None of these patients was discharged to their home or transferred to a rehabilitation care hospital. Ninety-two patients (77%) were hospitalized in CCUs ( n = 60; 50%) or cardiology wards ( n = 20; 17%). Three patients (3%) required the ICU because of respiratory or hemodynamic instability. Twenty patients (17%) were admitted to another medical ward. During their hospitalization, 71 patients (60%) were transferred to at least two different medical departments. Fig. 1 summarizes the entire course of care of the cohort.
