| This study describes the experience of a new admission cohort (n = 1,008) with initial nursing home stays of at least 28 days over a two year period or until the episode ended with death or discharge to the community. Compared to prior literature, there were fewer discharges to the community (15%) and initial deaths either in the nursing home or hospital (27%) and more permanent placements (58%) including those who experienced hospitalizations (33%). In bed days, those discharged to the community accounted for 3%, initial deaths, 13%, and permanent stayers, 84%. These patterns have implications for organization, delivery of care, and practice.;Duration of use was measured for an episode of care by inclusion of nursing home days following readmission after hospitalization (i.e., total length of stay (TLOS)). Median TLOS was 491 days compared to 170 days for individual stays. Residents experienced as many as seven readmissions over two years. Therefore, TLOS represents an important improvement in measurement of nursing home utilization.;Three diagnoses responsible for the most nursing time initially (i.e., fractures, dementia, and cancer), age, source of admission, and Medicaid eligibility accounted for 92% of the variance between five immediate outcomes--discharged home, died in the nursing home without leaving, hospitalized and died there, hospitalized and returned, and still in the nursing home without discharge. Nursing care variables did not improve this prediction. Predictors of TLOS were different for those discharged to the community and those not. Independence and restorative care facilitated shorter stays for those returning to the community. Longer stays for those not returning were associated with independence, no skilled nursing care, not having certain diagnoses accounting for most nursing time--cancer, circulatory disease, chronic respiratory disease, or diabetes--and being female. These findings suggest that diagnoses related to initial nursing care are especially useful predictors of nursing home utilization. |
