Physician Documentation of Resident to Resident Mistreatment in Residential Care Facilities

While mistreatment of long-term care residents by staff is unconscionable, this proposal asserts that the greatest threat of elder abuse to American nursing home residents comes not from staff, but from other cohabitants in the form of resident-to-resident elder mistreatment (R-REM).
Despite pilot data suggesting R-REM is prevalent, it is both under-studied and under-reported.

Accordingly, the specific aims of this project are to:
1. Enhance institutional recognition of R-REM by deriving R-REM information from five different sources, including two added for this project: forensic medical record review, and accident/incident reports. Additionally, a gold standard consensus conference classification is proposed for a random sample of residents.
2) Examine the convergence of R-REM reports across the five different methodologies.
3) Identify the most accurate mechanism for detecting and reporting R-REM.
4) Develop profiles to describe the types of people reported by each different source.
5) Investigate the existing policies and procedures for reporting R-REM in each facility.
6) Develop institutional guidelines for the reporting of R-REM episodes.

This project will afford the opportunity (as a piggy-back to ongoing state and federal parent grants) to augment the number of reporting mechanisms examined including two additional methods. The New York State Department of Health (NYSDOH) parent grant is being conducted in five urban facilities, and is a quasi-experimental evaluation of a staff training program in recognition of R-REM. The National Institute on Aging (NIA) grant is a prevalence study which is being conducted in an additional sample of five suburban nursing homes. This NIJ-funded project will collect additional reporting source data from five urban nursing homes enrolled in the NYSDOH study during phase 1 (years 1 and 2), and will replicate these findings through cross-validation in three suburban facilities as part of the NIA study in phase 2 (year 3).

The target population will be that of the parent NYSDOH grant, i.e., long-stay cognitively intact, mild, moderate and 'low'-severe cognitively impaired nursing home residents from larger downstate (urban) facilities. During phase two, the target population will be three additional, smaller suburban facilities. The anticipated sample size for the proposed project is at least 600 urban residents and 300 suburban residents. Analyses will be cross-sectional and longitudinal. The analytic approaches relate to examination of agreement, and to prediction of multiple, correlated binary outcomes from covariate risk factors. The method used will be a generalization of the logistic regression model to the multivariate context.

ca/ncf