Suicidal Ideation in US Nursing Homes: Association With Individual and Facility Factors

Objective: To assess prevalence of suicidal ideation (SI) among new postacute and long-stay nursing home (NH) admissions and examine the associations with individual and NH-level factors.

Setting/participants: A total of 1,864,102 postacute and 304,106 long-stay admissions to just over 15,000 NHs between 7/1/2014 and 6/30/2015.

Measurement: Using 100% of the national Minimum Data Set 3.0, we identified SI and key covariates. SI was based on responses to one item on the PHQ-9 scale. For postacute residents, SI was measured at admission and discharge. For long-stay residents, SI was assessed at admission and assessments closest to 90, 180, and 365 days thereafter. Patient sociodemographics, functional and cognitive status, comorbid conditions, and other covariates were included as independent variables, as were several NH-level factors. Logistic regression models were fit to estimate SI risk at admission and at subsequent time intervals.

Results: Observed 2-week prevalence rates of SI were highest at admission (1.24% for postacute and 1.84% for long stays) and declined thereafter at each subsequent time interval. The odds of SI were significantly increased for residents with severe depression at admission and all subsequent intervals. Residents in for-profits had significantly lower rates of SI, compared with those in not-for-profits.

Conclusions: Our findings demonstrate that SI risk in NHs is highest at admission and subsequently declines. We found several potentially modifiable individual-level risk factors for SI. The identification of SI may be seriously underreported in for-profit-facilities. Future research may be needed to explore how the PHQ-9 item on SI is understood by residents and recorded by staff.