Prevalence and associations of general practitioners' ordering of "non-symptomatic" prostate-specific antigen tests: A cross-sectional analysis

Aims: Testing for asymptomatic prostate cancer with prostate specific antigen (PSA) is of uncertain benefit. Most relevant authorities recommend against screening, and for informed patient choice. We aimed to establish the prevalence and associations of non-symptomatic PSA-testing of men aged 40 or older by early-career general practitioners (GP registrars). Methods: A cross-sectional analysis from the ReCEnT cohort study of registrars' consultations, 2010-2014 (analysed in 2016). Registrars record 60 consecutive consultations each 6-month training term. The outcome factor was ordering an asymptomatic PSA test (a PSA ordered for an indication that was not prostate-related symptoms or prostatic disease monitoring). Independent variables were patient, registrar, practice, consultation and educational factors. Results: A total of 856 registrars contributed details of 21,372 individual consultations and 35,696 problems/diagnoses of males 40 or older. Asymptomatic PSAs were ordered for 1.8% (95%CI: 1.7-2.0%) of consultations and for 1.1% (95%CI: 1.0-1.2%) of problems/diagnoses. Multivariable associations of asymptomatic PSA testing (compared with problems/diagnoses for which a PSA was not ordered) included patient age (OR 2.32 [95%CI: 1.53-3.53] for 60-69years compared with 40-49), patient ethnicity (OR 0.40 [95%CI: 0.19-0.86] for non-English speaking background), the patient being new to both the registrar and practice (ORs 1.46 [95%CI: 1.08-1.99] and 1.79 [95%CI: 1.03-3.10]), the number of problems/diagnoses addressed (OR 1.44 [95%CI: 1.25-1.66] for each extra problem) and more pathology tests being ordered (OR 1.88 [95%CI: 1.79-1.97] for each extra test). Conclusion: GP registrars frequently order asymptomatic PSA tests. Our findings suggest that non-compliance with current guidelines for PSA screening may be relatively common and that targeted education is warranted.