Long‐Term Utilization Patterns of Antidepressants in Two Italian Regions: A Naturalistic Observation of Over 2 Million Individuals Across 5 Years

Background: Antidepressant drugs (AD) are commonly used and prescribed. However, factors associated with their prescription and use have not yet been fully characterized. Objective: To better characterize long-term utilization patterns of antidepressants, leverage data collected by naturalistic observation (i.e., routinely collected administrative prescription data, without experimental manipulation) across 5 years in two Italian regions. Methods: Sociodemographic factors associated with prescription in a population of more than 2,000,000 individuals were explored (multivariate logistic regression). Multivariate survival analysis was employed to investigate factors associated with the time to drug discontinuation. Diverging patterns of use were described and empirically tested by logistic regression, performed on time series of drug utilization. Results: A total of 219,351 incident users were observed. Older age and female sex were associated with a more likely prescription (age 45–64: incidence rate ratio [IRR] 1.63; age 65–84: IRR 3.266; age 85+: IRR 10.244; all p < 0.001: females vs. males IRR 1.740, Z-score 50.08, p < 0.001). Older age and the use of tricyclic or other antidepressants (compared to selective serotonin reuptake inhibitors [SSRIs]) were associated with lower discontinuation and thus longer treatment persistence (minimum hazard ratio [HR], 0.680, Z-score −36.08, p < 0.001). Predictive models could identify diverging patterns of use (area under the curve [AUC] ≥0.697). Conclusions: The long-term pharmacoutilization of antidepressants was shown to be under the influence of structural and sociodemographic factors. Considering the recent directive by the European Commission on the European Health Data Space, implications for public health policies and future research may be considered.