Patterns of medication errors across therapeutic drug classes: a retrospective analysis of incident reports in a Saudi tertiary hospital (2023–2025)

Background Medication errors vary in type, cause, and severity across different therapeutic drug classes, yet drug-class–specific error profiles remain poorly characterized. Understanding these patterns is essential for developing targeted prevention strategies. This study aimed to evaluate how medication error types, stages, and causes vary across therapeutic categories among voluntarily reported errors in a tertiary hospital. Methods A retrospective cross-sectional study was conducted at Al-Noor Specialist Hospital in Makkah, Saudi Arabia, analyzing medication error reports from January 2023 to August 2025. Errors were classified by therapeutic drug class, error type (dosing, selection, information, duplication, other), stage of the medication process, and cause (system-related, environmental, staff experience). Error types and causes were compared across drug classes using chi-square tests with false-discovery-rate correction. Multivariable logistic regression models examined associations between drug classes and error characteristics, adjusting for patient demographics, year, and error stage, with Holm adjustment for multiple comparisons. Results Among 6,546 medication-error reports, antibiotics/antimicrobials (32.8%) and cardiovascular drugs (22.5%) accounted for the largest shares of reports. Dosing errors predominated across most drug classes, with highest proportions observed in electrolytes/supplements (78.9%) and respiratory drugs (75.3%). Cardiovascular drugs showed higher proportions of duplication errors (25.6% vs. 10.9%) and selection errors (4.1% vs. 1.3%) than other classes (both q < 0.001), whereas only three errors (0.05%) resulted in harm, precluding comparison across drug classes. Antibiotics were significantly associated with increased staff experience-related errors (OR = 1.80, 95% CI: 1.51–2.14), while electrolytes were associated with environmental causes (OR = 1.43, 95% CI: 1.20–1.71). Drug class remained a significant independent predictor of error type and cause after adjustment for confounders and multiple comparisons. Conclusion Patterns of reported medication errors vary across therapeutic drug classes, with cardiovascular drugs showing higher proportions of selection and duplication errors. These patterns describe reported errors, not error rates, but may help to prioritize targeted quality improvement.