Abstract
Background
General practitioners (GPs) are often the first point of contact for adults presenting with symptoms of depression or anxiety. Practices may vary due to patient and physician tendencies, influencing treatment pathways. This study aimed to quantify variation in GPs’ diagnostic tendencies for depression and/or anxiety and their association with subsequent healthcare utilisation and sickness absence.
Methods
Using linked national registry data from 2008 to 2021, we studied adults assigned a new GP due to list closures, list reductions, or immigration. We constructed two measures of each GP’s diagnostic tendency: one based on their diagnostic rate among patients seen before the index patient was assigned (prior measure), and one based on patients seen after (current measure). Using ordinary least squares (OLS) regressions with fixed effects, we first examined the association between each tendency measure and the likelihood of receiving a depression and/or anxiety diagnosis, to verify that the measures capture meaningful variation in diagnostic tendencies. We then estimated associations between GP tendency and psychiatric and somatic healthcare utilisation and sickness absence over three-year follow-up.
Results
Diagnostic tendency exhibited moderate variation across GPs. A 1%-point increase in diagnostic tendency was associated with a 0.47 (95% CI: 0.44–0.49) percentage-point increase in the probability of receiving a diagnosis using the current measure, and 0.32 (95% CI: 0.29–0.34) using the prior measure, corresponding to approximately 27 and 13 additional diagnoses per 1,000 newly assigned patients, respectively, when comparing GPs one standard deviation above versus below the mean. A 5%-point increase in diagnostic tendency was associated with a 0.4 (95% CI: 0.2–0.5) percentage-point increase in non-acute psychiatric outpatient care and a 0.3 (95% CI: 0.1–0.4) percentage-point increase in repeated sickness absence, consistent across time and instruments, corresponding to approximately 2 to 6 and 3 to 4 additional cases per 1,000 patients per year, respectively. Associations with other outcomes were smaller and imprecise.
Conclusion
GPs vary in the extent to which they diagnose depression and/or anxiety, affecting patients’ likelihood of psychiatric care use and sick leave. Whether this is beneficial or harmful remains unclear and requires further investigation.