Knowing your patient better: a nationwide cohort study of antibiotic prescribing for respiratory tract infections in Norwegian primary care

Background Antibiotic overuse drives antimicrobial resistance. In high-income countries, most antibiotics are prescribed in primary care, and respiratory tract infections (RTIs) are the leading reason, although most outpatient RTIs are viral. Diagnostic uncertainty is a major physician-level driver of precautionary prescribing. Repeated contacts between a patient and the same general practitioner (GP) build relationship-specific knowledge and trust that might reduce this uncertainty, but evidence is scarce because most studies measure relational continuity at the patient level rather than at the specific doctor−patient pair. Methods In this nationwide register-based retrospective cohort study, we linked four Norwegian nationwide registries covering all primary care contacts during 2010−2019 (about 176 million contacts). We measured relational continuity directly as the running count of prior face-to-face contacts within each patient-GP pair and estimated linear probability models of antibiotic prescribing for RTI contacts with GP fixed effects, adjusting for patient, GP, and contact characteristics, separately for patients listed and not listed with the consulting GP. Reporting follows the STROBE guideline. Findings Antibiotic prescribing was substantially higher when the current RTI contact was the first patient-GP encounter compared with RTI contacts with prior interactions (any reason) within patient-GP pair. Prescribing probability decreased with the number of prior contacts and stabilised within roughly four to eight encounters. Relative to the first contact, prescribing fell by up to 5.2−7.6 percentage points for patients not listed with the GP, and by up to 1.0−1.8 percentage points for listed patients. Independent of encounter number, listed patients had a 4.3−4.5 percentage point lower probability of receiving a prescription. Interpretation Both formal registration and accumulated personal contact were associated with lower antibiotic prescription probability. Stable, contact-rich doctor−patient relationships are a plausible lever for antimicrobial stewardship.