Prescribing Appropriateness and Clinical Outcomes in Hospitalized Adults with COVID-19 in Jazan, Saudi Arabia: A Retrospective Cohort Study.

Background: The COVID-19 pandemic required repeated revision of national treatment protocols; evaluating prescribing against the guidance then in force is central to stewardship. Objective: To characterize clinical outcomes of adults hospitalized with COVID-19 in Jazan, Saudi Arabia, and to assess prescribing appropriateness using the Medication Appropriateness Index (MAI) benchmarked against contemporaneous Ministry of Health (MOH) guidance. Methods: Single-center retrospective cohort study of all adults (aged 18 years or over) admitted with COVID-19 between 1 March 2020 and 28 February 2021. Two trained clinical pharmacists independently applied the MAI to 1018 prescriptions. All associations were examined in univariable, unadjusted analyses. Results: Hypertension (44.1%), diabetes (38.7%), end-stage renal disease (13.7%), and cardiac disease (12.4%) were prevalent. In-hospital mortality was 26.8% (121/451 patients with a determinate outcome). Diabetes (OR 2.02, 95% CI 1.32-3.08), hypertension (OR 1.92, 95% CI 1.26-2.93), and cardiac disease (OR 2.04, 95% CI 1.15-3.64) were each associated with higher unadjusted odds of death; patients who died were on average 15.8 years older (p < 0.0001). Overall, 936 of 1018 prescriptions (91.9%) were appropriate; rates ranged from 99.4% (311/313) for enoxaparin to 42.9% (6/14) for hydroxychloroquine, the latter based on a small number of prescriptions. Inappropriate prescriptions were flagged mainly in the duration (82.9%) and indication (70.7%) domains, and receipt of at least one inappropriate prescription was not associated with mortality (OR 0.59, 95% CI 0.31-1.12; p = 0.10). Conclusions: Anticoagulants and corticosteroids were largely prescribed in line with contemporaneous MOH guidance; several repurposed agents were not. These findings support an MAI-based audit against the guideline version in force as a tool for quality control.