Understanding physician prescription behaviors: a systematic review and meta-analysis of macro, meso, and micro-level influences.

Abstract

Background: Prescription is a complex act that reflects the physician's expertise and authority. While some factors affecting prescription decisions have been studied, empirical findings often conflict, leaving our understanding of prescription behaviors limited and fragmented.

Objective: To assess the factors influencing physicians' drug prescribing habits by applying Strong Structuration Theory. Factors are categorized at: physician, practice, patient, industry, and system level.

Methods: Pubmed, Scopus, and ISI Web of Science were searched from inception to June 2025. Peer-reviewed studies were included if they were published in English, empirical, and assessed at least one factor influencing physicians' prescribing behaviors. Studies reporting the effect of covariates on prescriptions using Odds Ratios were included in the meta-analysis.

Results: 146 studies were selected for the review. At the macro-level, physicians were more likely to prescribe after being exposed to marketing activities by pharmaceutical industries, and for privately insured patients. Meso-level factors, such as practice ownership and setting, showed conflicting results, with no significant effect observed in the meta-analysis. Micro-level influences were the most prevalent in literature. Patient requests had a significant positive effect on prescriptions. Physician-level influences were inconsistent across most variables, except gender, where male physicians were more likely to prescribe. This effect was not confirmed by the meta-analysis, which showed heterogeneity across studies.

Conclusion: This study highlights the complexity of prescribing behaviors and the challenges in designing effective micro-level policies. Policymakers should therefore consider the multiple influences on prescribing to design targeted interventions that promote high-quality prescribing practices.