Objective: Within Veterans Affairs (VA) nursing homes (NHs), quality issues have a tremendous impact on the population with serious mental illness (SMI), who are more likely than their non-SMI Veteran counterparts to use NH services. We examined recent trends in quality indicators (QIs) measuring poor performance of VA NHs and whether the facility-level QIs vary with SMI concentration within the facility.
Methods: From VA administrative records including Minimum Data Set assessments, we identified all residents in the 135 VA NHs between fiscal years 2005 (FY05) through FY07. We used a zero-inflated Poisson regression to assess trends in and facility-level predictors of 3 process-related QIs: depression without antidepressant therapy; bladder/bowel incontinence without a toileting plan; and physical restraint use. Facility-level predictors included collocated special care units, rurality, staffing, physical plant characteristics, SMI prevalence, and SMI admission volume.
Results: During FY05–FY07, restraint use declined from 1.2% to 1.1% and incontinence without a toileting plan from 25.8% to 22.1%, but untreated depression increased from 5.1% to 5.5%. Despite overall gains in quality, higher SMI prevalence was associated with higher odds of physical restraint use and lack of toileting plan. Higher SMI prevalence was also associated with higher frequency of untreated depression. Other characteristics such as complex building structure were predictive of variation in quality, but the relationships were not consistent across QI types.
Conclusion: VA NHs had significant improvements in these examined QIs during the study period. Nonetheless, overall poorer quality was observed at sites with higher SMI concentrations.
*Center for Statistical Consultation and Research, University of Michigan
†Department of Veterans Affairs, Ann Arbor Center of Excellence (COE), Serious Mental Illness Treatment Resource and Evaluation Center (SMITREC)
‡Department of Health Management and Policy, School of Public Health
§Department of Psychiatry, University of Michigan Medical School, Ann Arbor, MI
Supported by the Department of Veterans Affairs, Health Services Research and Development Grant IAD 06-055.
The authors declare no conflict of interest.
Reprints: Hyungjin M. Kim, ScD, Center for Consultation and Research, University of Michigan, 3555 Rackham, Ann Arbor, MI 48109-1070. E-mail: email@example.com.