This proposal directly addresses the national imperative for innovative strategies to improve care for Medicare beneficiaries and reduce health care costs. The overall objective of the proposed project is to improve the care of older individuals who reside in nursing homes (NHs), and at the same time reduce unnecessary Medicare expenditures. This goal will be accomplished by testing a quality improvement program designed to reduce the number of avoidable hospitalizations of NH residents in a randomized controlled trial. Hospitalizations of NH residents are frequent and associated with numerous complications and increased health care costs. Previous research suggests that as many as two-thirds of such hospitalizations may be avoidable. Anticipated changes in Medicare reimbursement will reduce financial incentives that favor hospitalization, but could result in reduced care quality if NH staff does not have the training and clinical tools to manage residents in the NH when acute changes occur. INTERACT (Interventions to Reduce Acute Care Transfers) is a quality improvement program that utilizes tools based on established clinical guidelines. The tools target three key strategies to reduce potentially avoidable hospitalizations: 1) preventing conditions from becoming severe enough to require acute hospital care; 2) managing selected acute conditions in the NH; and 3) improving advance care planning for residents among whom a palliative or comfort care plan, rather than acute hospitalization, may be appropriate. Preliminary research involving 30 NHs demonstrated a 17% reduction in hospitalizations compared to the same six-month period in the previous year. The cost of the intervention was $7,700; projected savings to Medicare of reduced hospital admissions from a 100-bed NH were $125,000/year. While these results are promising, the effectiveness of INTERACT in reducing hospitalizations remains to be tested in a controlled trial. The proposed project will therefore involve an interdisciplinary team of experienced NH researchers in conducting a randomized controlled trial to test the implementation of the INTERACT program. NHs randomized to INTERACT implementation will participate in 3- months of training followed by a 12-month implementation period during which they will be supported by an experienced nurse practitioner through regular teleconference calls and as-needed telephonic or email communication. The effects of implementing the INTERACT program on hospitalization rates will be compared to a randomly assigned group of usual care control NHs and a group that will self-monitor hospitalization rates only. The hypotheses to be tested are: 1) INTERACT implementation NHs will have a greater reduction in hospitalization rate than the control and monitoring only NHs during the 12-month implementation period compared to a 12-month baseline period; and 2) reductions in Medicare expenditures for hospitalizations in the INTERACT NHs will exceed the estimated costs of implementing the program.
