Clinical outcomes after antipsychotic discontinuation in nursing home residents with dementia: a nationwide cohort study.

BACKGROUND

Antipsychotic medications in people with dementia should be avoided or limited to treatment courses of fewer than 3-4 months. It is unknown whether discontinuation of long-term antipsychotic medications in this population can improve clinical outcomes. This study evaluated clinical outcomes associated with discontinuation of long-term antipsychotic medication use compared with continued use among individuals with dementia residing in nursing homes.

METHODS

This cohort study used Medicare claims data (2013-2022) linked to Minimum Data Set clinical assessment data. The study population included nursing home residents aged ≥ 65 with dementia who had used antipsychotic medications for more than 90 days. Individuals who discontinued antipsychotic medications (gap ≥ 60 days) were matched to continuers using incidence density sampling based on time since first prescription of antipsychotic medications. Propensity score matching was applied to adjust for 86 covariates. Study outcomes included all-cause hospitalization, specific rehospitalization reasons, and all-cause mortality. Hazard ratios (HRs) were estimated using Cox proportional hazards model, and rate differences (RDs) per 1,000 person-years were estimated using generalized linear models.

RESULTS

Two thousand one hundred eleven propensity score-matched pairs were included and followed for a median of 298 days. Antipsychotic medication discontinuation versus continuation was associated with lower risk of all-cause hospitalization (HR [95% CI]: 0.83 [0.72, 0.96]; RD [95% CI]: -55.85 [-97.82, -13.88]) and hospitalization for pneumonia (HR: 0.51 [0.29, 0.90]; RD: -12.85 [-23.26, -2.44]) or urinary tract infections (HR: 0.47 [0.26, 0.83]; RD: -14.22 [-24.62, -3.83]); non-significantly lower risks of hospitalization for falls (HR: 0.88 [0.70, 1.11]; RD: -13.19 [-37.78, 11.41]), delirium (HR: 0.76 [0.57, 1.02]; RD: -19.26 [-39.18, 0.66]), stroke (HR: 0.95 [0.42, 2.16]; RD: -0.31 [-7.06, 6.44]), and all-cause mortality (HR: 0.89 [0.79, 1.02]; RD: -38.23 [-81.01, 4.54]). Subgroup analyses showed the greatest reduction in hospitalization risk was among individuals with moderate-to-severe dementia (HR: 0.78 [95% CI: 0.65, 0.93]; RD: -96.76 [-163.75, -29.77]).

CONCLUSIONS

Among nursing home residents with dementia, discontinuing long-term antipsychotic medication use was associated with lower risks of hospitalization, pneumonia, and urinary tract infections, suggesting clinical benefits in this population.