EFFECTS OF HOME-BASED INTERVENTION ON UNPLANNED READMISSIONS AND OUT-OF-HOSPITAL DEATHS

OBJECTIVE: To determine the effect of a home-based intervention (HEI) on the frequency of unplanned readmission and out-of-hospital death am ong patients discharged home from acute hospital care. DESIGN: A rando mized controlled trial comparing HBI with usual care (UC). SETTING: A tertiary referral hospital servicing the north-western region oi Adela ide, South Australia. PARTICIPANTS: Medical and surgical patients (n = 762) discharged home after hospitalization, INTERVENTION: Home-based intervention (n = 381) consisted of counseling of all patients before discharge followed by a single home visit (by a nurse and pharmacist! to those patients considered to be at high risk of readmission (n = 31 4) in order to optimize compliance with and knowledge of the treatment regimen, identify early clinical deterioration, and intensify follow- up of such patients where appropriate, MEASUREMENTS: The primary endpo int was the number of unplanned readmissions plus out-of-hospital deat hs over a 6-month follow-up period. RESULTS: During the study follow-u p, the major endpoint occurred most commonly in the UC group (217 vs 1 55 episodes: P < .001). Overall, the HBI group demonstrated fewer unpl anned readmissions (154 vs 197: P = .022), out-of-hospital deaths II v s. 20: P < .001), total deaths (12 vs. 29: P = .006), emergency depart ment attendances (236 vs 314: P < .001), and total days of hospitaliza tion (1452 vs 1766: P < .001), There was a disproportionate reduction in multiple events among HBI patients (P = .035). Hospital-based costs of health care during study follow-up tended to be lower in the HBI g roup ($A2190 vs $A2680 per patient: P = .102). Mean cost oi HBI was $A 190 per patient visited, whereas other community-based health care cos ts were similar for both groups. CONCLUSIONS: Among high-risk patients discharged from acute hospital care, HBI is beneficial in limiting un planned readmissions and reducing risk of out-of-hospital death. It ma y be particularly cost-effective ii applied selectively to patients wi th a history of frequent unplanned hospital admission.