Since the advent of prospective payment, patients are discharged sicker and quicker, and acute care delivery has shifted from the hospital to the home setting. Despite the increased use of home care services, very little is known about how skilled home care and patients' illness characteristics affect hospital readmission.

The purpose of this study was to compare readmission patterns among chronically ill people discharged with and without skilled home care and determine if these patterns are amenable to change. A retrospective design was used for this study. A convenience sample (n = 922) was obtained from discharges at a regional referral hospital located in northern Michigan during 1990 and 1991. Home care and self care patients were matched for age and diagnostic group. Disease Staging was used to measure illness severity.

Results indicate that the 30-day readmission rate was the same for self care (13%) and home care patients (14%). The 30-day readmission rate was similar for self care and home care patients with low, medium and high mortality risk. However, the 30-day readmission rate among the very high risk group was lower for home care patients (20%) than for self care patients (24%). Hierarchical log-linear results indicate no interactive effect between home care use and readmission.

The results suggest that skilled home care may have controlled hospital readmission within 30 days for this study sample. Further studies should apply a behavior model, a longer time series, and measures of illness characteristics that include functional status, self care complexity, caregiver status and the presence of major symptoms.