The effect of Evercare on hospital use.

OBJECTIVES To examine the use of hospital and related medical care services of a novel managed care program using nurse practitioners (NPs) and directed specifically at long-stay nursing home residents. DESIGN Quasi-experimental posttest design with two control groups to minimize selection bias. SETTING Nursing homes. PARTICIPANTS Evercare enrollees in five sites were compared with two sets of controls: nursing home residents in the same nursing homes who did not enroll in Evercare (control-in) and residents of nursing homes that did not participate in Evercare (control-out). MEASUREMENTS Utilization data from Medicare and United Healthcare (the parent corporation for Evercare) were obtained for slightly more than 2 years. Patterns of use were assessed by calculating the monthly use rate for each group and aggregating to form annual rates. Usages addressed included hospital admissions and days, emergency room visits, therapy services, mental health services, and podiatry. Adjustments were made to correct for age, race, and sex. Because the groups differed in terms of the rate of cognitive impairment, the analysis was stratified on this variable. RESULTS The incidence of hospitalizations was twice as high in control residents as in Evercare residents (4.63 and 4.67 per 100 enrollees per month vs 2.43 in the 15 months after census, P<.001). This difference corresponded to Evercare's use of intensive service days. The same pattern held for preventable hospitalizations (0.80 and 0.86 vs 0.28, P<.001). The pattern held when residents were stratified by cognitive status. On average, using a NP is estimated to save about $103,000 a year in hospital costs per NP. CONCLUSION The use of active primary care provided by NPs may have prevented the occurrence of some hospitalizable events, but its major effect was allowing cases to be managed more cost-effectively.