Can the remuneration scheme of general practitioners affect their antibiotic prescription behaviour?

Antibiotic resistance poses a significant global health threat, exacerbated by over-prescription of antibiotics, which often happens in primary care for respiratory tract infections (RTIs). Studies indicate that up to half of these prescriptions may be unnecessary. Little is known about how general practitioners' (GPs) remuneration schemes influence prescribing. GPs compensated via fee-for-service (FFS) and capitation (CAP) may face stronger incentives to prescribe antibiotics compared to salaried GPs, as prescriptions can signal quality, reduce consultation time, and aid patient retention - critical where reimbursements depend on consultations and list size. This study examines how GP remuneration influences antibiotic prescribing for RTIs using Norwegian register data.

We utilized linked registry data (2015-2019) from the Control and Payment of Health Reimbursements Database (KUHR), National GP Registry, Norwegian Prescribed Drug Registry (NorPD), and Statistics Norway. We matched antibiotic prescriptions to patient-GP contacts for RTIs. The dataset covers regular GPs (mixed FFS/CAP or fixed salary) and locum GPs (FFS or salary).

(1) probability of antibiotic prescription during RTI contacts; (2) probability of selecting non-phenoxymethylpenicillin (non-PcV). We used linear probability, logit, and probit models, controlling for GP, patient, contact, and practice attributes. To mitigate selection bias, we exploited within-GP variation among those who switched remuneration type during the study period.

Regular FFS/CAP GPs had a 12-15% higher relative probability of prescribing antibiotics for RTIs than salaried GPs, especially at initial contacts. When prescribing, they were 9-11% more likely to choose non-PcV. Switchers analyses showed that FFS/CAP increased prescription rates by 14% and non-PcV choice by 8%. Among locums, we did not find any significant difference in overall prescription rates between FFS and salary, but FFS locums favoured non-PcV by 7%.

Remuneration schemes may influence antibiotic prescribing behaviour. FFS/CAP is linked to higher prescription rates and broader-spectrum antibiotic use among regular GPs, likely due to patient retention and time-efficiency incentives. Policy interventions, such as monitoring of prescriptions depending on the remuneration type or adjustments to the remuneration scheme (e.g., antibiotic-related pay-for-performance), could promote prudent prescribing. Further research is needed on prescription appropriateness and quality impacts.