Background: Medicines have an important role in the treatment of many health conditions. Interventions to improve medicine use aim to maximize potential benefits and minimise the potential for harm. Pharmaceutical product promotion frequently reflects the tension between commercial and health priorities, as promoted medicines do not always represent advances in patient care and may introduce risk of harm. The extent and types of interactions between pharmaceutical companies and prescribers has expanded, with regulatory standards not fully reflecting all current promotional activities.
Objectives: To assess the effects of pharmaceutical industry promotional interactions with prescribers on the quality, quantity, cost of prescribed medicines, and on formulary requests.
Methods: Multiple study designs were eligible for this Cochrane Effective Practice and Organisation of Care (EPOC) review, including the following: randomised trials, controlled before-after studies, interrupted time series studies, cohort studies, and cross-sectional studies. Included studies had to have a measure of pharmaceutical industry interaction with clinicians, either individually or as part of a group, and of clinicians’ prescribing (appropriateness, quantity and/or cost). Types of interactions were grouped as either direct receipt of information or gifts and payments, including free samples, from pharmaceutical companies.
Results: Of 11,758 identified records, 69 studies met our inclusion criteria, including the following: 3 RCTs, 4 interrupted time series, 12 controlled cohort studies, 7 controlled before-after studies, 1 case-control study, and 42 cross-sectional studies, 30 of which use United States Open Payments data. A total of 19 of these studies examined effects on prescribing appropriateness, 37 on quantity, 25 on costs, and 1 on formulary requests. Types of interactions mainly included gifts/payments, free samples, and information from sales representatives. Several studies showed a dose-response relationship between payment amounts, including numbers of meals provided, and prescribing outcomes. In addition, eight studies examined the effects of conflict-of-interest policies or other restrictions on clinicians’ interactions with industry.
Conclusions: The evidence base includes a range of study designs with differing risks of bias. This variety of designs creates challenges in interpreting the evidence. However, these challenges are common to reviews of public health interventions in which interventions cannot be easily assessed through randomised trials. Final results will be presented. Patient, public, and/or healthcare consumer involvement: None.