Prostate-specific antigen testing: uncovering primary care influences

OBJECTIVES To examine influences on the behaviour of General Practitioner (GP) in relation to prostate-specific antigen (PSA) testing. SUBJECTS AND METHODS In Northern Ireland in 2003-2004, all GPs (1067) were invited to complete a self-administered postal questionnaire survey that was then matched with a regional PSA-testing database. The main outcome measures were individual GP responses for demographic, practice and training characteristics, PSA testing behaviour and perceived influences, matched against GP-initiated first PSA tests performed in 2003 and 2004 (22 207 tests). RESULTS In all, 704 GPs (66%) responded and 49% of these reported awareness of the national guidelines, which was highest among those attending postgraduate meetings. PSA tests were more likely to be ordered by full-time male GPs who had attended a local postgraduate urology meeting; ran a 'well-man' clinic; tested men with unrelated complaints; and were not in a training practice. Testing levels were highest among GPs who had been practising for 21-30 years and those in rural practices. Awareness of national guidelines or having had a postgraduate post in urology did not affect testing behaviour. After adjusting for gender, working hours, duration in practice and urban/rural setting, independent influences increasing testing behaviour were: testing men with a positive family history or unrelated complaints; testing any man who requests it; and previous experience of prostate cancer being detected in an asymptomatic patient by PSA testing. Working in an accredited training practice was associated with lower testing levels. CONCLUSION There are complex influences on the PSA testing behaviour of GPs; addressing these influences could contribute to the rationalization of testing. A low awareness of national guidelines indicates a need for new strategies to disseminate and implement guidelines. The influence of local educational meetings on PSA testing is an unharnessed force.