Effect of Selective Serotonin Reuptake Inhibitors on Healthcare Utilization in Patients with Post-Traumatic Stress Disorder and Alcohol Use Disorder

Aims: The objective of this study is to address equivocation in estimates of selective serotonin reuptake inhibitor initiation (SSRI) effect on all-cause and alcohol-related ER visits, and medical or psychiatric admissions within 2 years of initial Post-Traumatic Stress Disorder (PTSD) diagnosis in patients with PTSD and Alcohol Use Disorder (AUD). Methods: This study is a quasi-experimental, new-user-design cohort study of 3235 patients seen at the VA North Texas Healthcare System between January 1, 2000 and December 31, 2016. High dimensional propensity score (HDPS) techniques were used to estimate likelihood of SSRI initiation within 30 days of first PTSD diagnosis. Propensity scores were used to calculate weights for likelihood of SSRI initiation which were used to control for baseline covariates in estimations of SSRI medication effect on odds of each outcome occurring. Results: Compared to those who did not receive SSRIs, patients prescribed an SSRI within 30 days showed significantly lower odds of alcohol-related ER visits (OR=0.668, 95%CI = 0.476 to 0.938, P = 0.02) and alcohol-related medical admissions (OR=0.583, 95%CI = 0.399 to 0.851, P = 0.005). Limitations: Inconsistent assessment of PTSD severity necessitated the use of HDPS models to control for baseline confounding. Our study design mimicked intent-to-treat trial design and therefore could not control for SSRI initiations after the 30-day grace period following initial PTSD diagnosis. Conclusions: SSRI initiation in patients with AUD and PTSD is associated with significantly reduced odds of alcohol-related medical hospitalization and alcohol-related ER visits within 2 years of first PTSD diagnosis. Additional studies are needed to verify these results.