GENERAL PRACTITIONER DEMENTIA DOCUMENTATION: A RANDOMISED CONTROLLED TRIAL USING INTERVENTIONS OF EDUCATION AND AUDIT

Abstract

Background: General practitioners (GPs) may be good at diagnosing dementia but poor at documenting the diagnosis, creating a barrier for appropriate management. In regional areas the GP is often the most accessible doctor, and patients may expect them to identify and manage this condition. The objective of this study was to test education and audit as interventions to improve documentation of dementia diagnoses by GPs. Methods: The design was multi-centered with clusters of GPs at 16 locations randomized using Excel software into one of four groups, three intervention and one control. The intervention applied separately and together, consisted of education - a workshop on GP dementia guidelines, and audit - the proportion of documented dementia diagnoses in patients aged 65 years and over compared with a predicted prevalence of 5% in this age group. The setting was Regional North-Queensland, Australia, spanning 1000 km. Participants were 56 GPs (8% of GP population) from 27 practices, 55$ with over 20 years in practice, 33% female, 57% International Medical Graduates. The outcome measures were the change in audit at 0 and 6 months of the trial. Multivariate linear regression analysis measured the effect on this outcome of independent variables consisting of the interventions applied separately and together, years in practice and International medical training. Results: 61 GPs entered the trial with 56 completing. Five withdrew due to work commitments or relocation. Over 10,000 patient files provided data for audits. Analysis was performed on two data set; intention to treat - all completing GPs (n=56), and per protocol - GPs providing individual data who complied with their intervention (n=37). Audit significantly increased documentation in the smaller sample (p=0.017) but not in the larger (p=0.50). Education had no significant effect. Combining education and audit did not increase the treatment effect. Neither 'over twenty years in practice' nor 'international medical training' significantly influenced the outcome. Conclusions: The intention to treat analysis produced no significant outcomes possibly due to poor protocol compliance. However, per protocol analysis suggested that audit might increase GP documentation of dementia diagnoses while education may not.