Predicting risk of hospitalization among adult home care clients: An international perspective

The primary objective of this project was to evaluate hospital admissions among older adult clients following entry into formal home care services and to develop a risk scale to predict hospitalization. Study design was a secondary analysis of data generated from the interRAI Home Care tool using near national home care data from Belgium, Canada, Italy, the Netherlands, New Zealand and the US. Baseline sample of (n = 2,928,674) was used to describe the cohort and establish profiles. The follow-up assessments (n = 1,831,766) were used to analyze change and develop the prediction model. Dependent measure was hospitalization within three months prior to follow-up. In the 90-day period prior to baseline assessment, 40.5 % of older adult home care clients had a hospitalization. Within 90 days prior to follow-up, 29.0 % were hospitalized. Final model (HospRisk-HC) based on scale distributions and prior hospitalization, comprised nine categories. Categories 1-5 represented persons without baseline hospital stay; categories 6-8 with baseline hospitalization. Average follow-up hospitalization rates ranged from 8.9 % to 33.5 % for categories1-5, and 50.9 % to 71.5 % for categories 6-9. The eta value was .432. HospRisk-HC scale differentiates between those with significantly lower and higher likelihood of hospitalization. Early identification of older adults most at risk for hospitalization helps direct attention and allocation of resources for efficient and effective home care service use.