OBJECTIVES: To characterize the dispensing patterns of Antidepression(AD) and Antianxiety (AA) medications among women with psychiatric disorders before and after hysterectomy for benign indications.

METHODS: This is a retrospective cohort study of patients who underwent hysterectomy for benign indications from January 2011 to December 2016 using the IBM MarkertScan^® Research Database. Inclusion criteria included reproductive age women (≤ 50 years); enrollment for ≥1 year prior to and 1 year following hysterectomy; at least 1 depressive or anxiety disorder diagnosis; and at least 1 dispensing of AD/AA medications. We measured monthly adherence (proportion of days covered (PDC)>=0.8) and persistence to AD/AA medications (days until a 45-day gap in therapy) over 12 months after hysterectomy. Group-based trajectory modeling was used to identify dispensing trajectory groups by clustering similar patterns of monthly AD/AA medications dispensing over 6 months before and 12 months after hysterectomy. Multinomial logistic regression was used to identify factors associated with individual dispensing trajectory patterns.

RESULTS: Among 11,607 women, six dispensing trajectory groups were identified: 1) Continuously high (27.0%); 2) Continuously moderate (21.9%); 3) Continuously low (17.9%); 4) Low high (10.0%); 5) Moderate low (9.8%); 6) Low moderate (13.4%). Compared with the continuously high dispensing group, younger age and no history of mood disorder were clinical predictors of low dispensing. The discontinuation rate at 3 months was higher at 88.6% in the continuously low dispensing group and at 66.5% in the continuously low-moderate dispensing group compared to 3.7%, 13.5%, 18.5%, and 42.5% in the continuously-high, low-high, continuously-moderate, and moderate-low dispensing groups, respectively.

CONCLUSIONS: The results provided 6 distinct patterns of dispensing of AD/AA medications over 18 months. Age and mood disorder were significant predictors of low dispensing pattern among the cohorts. Patients in the low-moderate and continuously low dispensing groups had higher risk of discontinuation and reinitiating the treatment compared with the other groups.