P0746SUSTAINABILITY OF AN ELECTRONIC TECHNOLOGY BASED INTERVENTION IN GENERAL PRACTICE TO IMPROVE THE DETECTION AND MANAGEMENT OF THE INTERRELATED CHRONIC VASCULAR DISEASES

Abstract Background and Aims Chronic kidney disease (CKD), type 2 diabetes (T2DM) and cardiovascular disease are interrelated chronic vascular diseases that cause substantial morbidity and mortality. A recent 15-month randomised stepped wedge trial targeting these conditions using an electronic technology audit tool together with education, monitoring and support to general practices (Chronic Disease IMPACT) showed increased diagnostic testing for CKD and T2DM in those at risk, increased urinary albumin monitoring in people with T2DM and increased coded diagnosis of CKD, no changes for cardiovascular disease and reduced coded eye examinations in patients with T2DM. This study aimed to test if these outcomes would be sustained 12 months after the initial study period (27 months after study commencement). Method 15 months after initial trial commencement the 9 practices involved in the original trial received further training on quality improvement and electronic technology tool use as well as a folder containing educational resources. They retained access to the electronic technology tool but received no further education or support for 12 months. Only 8 practices were included (as with the original analysis) due to compromised data quality in one practice resulting from a recent merging of practices with different electronic medical record systems. Active patients (having attended their GP practice ≥3 times within 24 months) age ≥18 were included. Variables with evidence for effect at 15 months were re-assessed at 27 months to test sustainability of outcomes. As there were no control practices at 27 months, time was considered as a linear predictor instead of a random intercept in the analysis. Net effect is defined as the product of effect over the course of the entire 27 months and change between effect at 15 and 27 months. Raw proportion and net effect odds ratio (OR) with 95% credible interval (CI) generated through stepped wedge analysis are presented below. Analysis was conducted with R version 3.5.1. Results In the 8 included practices, at baseline, prior to intervention, there were 37,946 patients, at 15 months 37,385 and at 27 months 37,813. Testing for kidney disease in those at risk was completed in 19% at baseline, 25% at 15 months and 26% at 27 months with OR 1.4 (CI 1.2-1.6). CKD diagnosis was coded in 4.5% at baseline, 5.8% at 15 months and 6.8% at 27 months with OR 1.9 (CI 1.6-2.2). Testing for diabetes in those at risk was completed in 34%