Detecting early clinical deterioration in chronic heart failure patients post-acute hospitalisation - a critical component of multidisciplinary, home-based intervention?

Abstract

Aims: To firstly describe the prevalence, characteristics and consequences of early clinical deterioration (ECD) in chronic heart failure (CHF) patients discharged from acute hospital care and, secondly, to examine the potential benefits of a multidisciplinary, home-based intervention (HBI) in limiting the common sequelae of such deterioration. Methods: This phenomenon was studied in 90 CHF patients assigned to the intervention arm of a randomised study of HBI. ECD was defined as death, unplanned re-admission or clinical instability (detected at a home visit) within 14 days of hospital discharge. Multivariate analysis was used to determine the independent correlates of ECD. Using these data, a 1:1 case–control ratio of patients assigned to the usual care arm of the study was selected to match those patients exhibiting non-fatal ECD and subject to HBI. Subsequent morbidity and mortality rates were then compared on the basis of the presence or absence of non-fatal ECD and/or HBI. Results: Of the 90 patients assigned to HBI, two died suddenly, five required an unplanned re-admission to hospital and 28 were found to be clinically unstable at a planned home visit, within 14 days of discharge from the index admission. The combined prevalence of ECD this cohort was therefore 39% (35 of 90 patients) and was independently correlated with greater age (OR=1.1 per yearly increment; P<0.001) and comorbidity (OR=2.0 per incremental Charlson index of comorbidity score; P<0.001). Patients who exhibited clinical instability at the home visit were significantly more likely to be non-adherent to prescribed treatment (10 of 28 vs. 9 of 55; P<0.05). Compared to the remainder of the cohort also subject to HBI, despite remedial intervention, patients who exhibited non-fatal ECD had reduced event-free survival (11 of 33 vs. 38 of 55; P<0.001), more frequent unplanned re-admission (0.2 vs. 0.1 admissions/patient/month; P<0.01), and more prolonged hospital stay (1.6 vs. 0.5 days/patient/month; P<0.001) in the subsequent 6-month period. However, compared to case–controls, these patients (n=33 in both groups) had fewer days of hospitalisation (1.6 vs. 3.6 days/patient/month; P=0.05) and, most significantly, were more likely to survive to 6 months (6 vs. 13 died; P<0.05). Conclusion: ECD is a common phenomenon in older CHF patients discharged from acute hospital care and is associated with poorer health outcomes in the longer-term. Post-discharge HBI is an important means for identifying and addressing ECD. Although HBI conveys benefits incremental to usual care, these data also provide a sound basis for increasing its effectiveness by applying earlier home visits in selected ‘high-risk’ patients.