Nursing Home Work Environment and Residents' Mental Health Outcome: Evidence from New York State

Percent of nursing home (NH) residents who are depressed or anxious is currently the only publically available mental health-related quality measure (QM) reported on the CMS NH Compare website. Although this QM has been reported quarterly for all Medicare/Medicaid certified facilities since late 2002, no study has examined the extent to which NH care processes may be associated with the variation in this outcome. Our study used Donabedian's structure-process-outcome framework and the Relational Coordination Theory to examine the association between facility-level staff cohesion (shared goals, values, group identity) and residents' symptoms of depression/anxiety.

The study population included 43,164 long-term residents from 162 NH in NYS during CY2006-2007. Data sources consisted of the Medicare Denominator File, Minimum Data Set, Online Survey Certification and Reporting file, and survey responses from 7,418 NH workers. We fit a logistic regression with facility random effects to model a dichotomous mental health QM. Key independent variable included a measure of staff cohesion (Likert-scale score). Other independent variables consisted of resident and facility characteristics.

Overall, 45.4% of residents had symptoms of depression or anxiety. Residents in NHs with better staff cohesion were 40% less likely to exhibit these symptoms (OR=0.597; CI: 0.365-0.975). Residents who were older (OR=1.013; CI: 1.011-1.015), female (OR=1.420; CI: 1.349-1.495), or white (OR=1.258; CI: 1.169-1.355) were more likely to have depressed or anxious symptoms. Other individual and facility-level factors were also found to be associated with these symptoms

Our study showed that residents in facilities with better staff cohesion had better mental health outcome.