Stroke risk among new users of typical vs. atypical antipsychotics among patients aged ≥65 years


These research results are from a retrospective, real-world, observational study. This web-based application provides an interactive platform to explore all analysis results generated as part of this study, as a supplement to a full manuscript currently in development for to a peer-reviewed journal. During manuscript development and the subsequent review period, these results are considered under embargo and should not be disclosed without explicit permission and consent from the authors.

Below is the abstract of the manuscript that summarizes the findings:

Backgroud: We estimated stroke risk associated with new exposure to haloperidol, or any typical antipsychotic, vs. atypical antipsychotic among patients aged ≥65 years regardless of dementia status.

Methods: IBM MarketScan Medicare Supplemental Database (MDCR) data (01 January 2001 to 31 December 2017) were used. Stroke risk for new users of typical antipsychotics (T1 cohort) or haloperidol (T2 cohort) was compared with new users of atypical antipsychotics (C1 cohort) aged ≥65 years. Crude incidence rate (IR) and incidence proportion of stroke were estimated within each cohort and gender subgroup. Three propensity score (PS) matching strategies were employed: unadjusted (crude), Sentinel PS replication, and a large-scale regularized regression model (adapted PS).

Results: Overall, 36,734 (T1), 24,074 (T2), and 226,990 (C1) patients were included. Crude IRs for stroke per 1000 person-years were 17.67 (T1), 23.74 (T2), and 14.17 (C1). In pre-planned analyses, PS-matched calibrated hazard ratio (cHR) for stroke T1 vs. C1 cohort was 1.08 (95% calibrated confidence interval [cCI]=0.75, 1.55) with Sentinel PS strategy and 1.31 (95% cCI=1.07, 1.60) with adapted PS strategy. The cHR for stroke in patients of T2 vs. C1 was 1.69 (95% cCI= 1.08, 2.75) with Sentinel PS strategy and 1.45 (95% cCI= 1.17, 1.80) with adapted PS strategy.

Conclusion: Stroke risk in elderly new users of haloperidol was elevated compared to new users of atypical antipsychotics and was elevated for typical antipsychotics using the adapted PS strategy.

Below are links for study-related artifacts that have been made available as part of this study:

Figure 2. Preference score distribution. The preference score is a transformation of the propensity score that adjusts for differences in the sizes of the two treatment groups. A higher overlap indicates subjects in the two groups were more similar in terms of their predicted probability of receiving one treatment over the other.
Download plot
Figure 4. Systematic error. Effect size estimates for the negative controls (true hazard ratio = 1) and positive controls (true hazard ratio > 1), before and after calibration. Estimates below the diagonal dashed lines are statistically significant (alpha = 0.05) different from the true effect size. A well-calibrated estimator should have the true effect size within the 95 percent confidence interval 95 percent of times.
Download plot