OP0325 Effect of three health it interventions on ra disease activity score documentation in an academic rheumatology clinic

Background Routine assessments of disease activity are a key part of high quality rheumatoid arthritis (RA) care. Recommendations published in 2012 detailed the preferred disease activity measures, but little has been published about how to best implement these measures to guide routine clinical practice. Customization of the electronic health record (EHRs) provides an opportunity to systematically collect disease activity scores in order to adequately implement treat-to-target approaches to RA treatment. Objectives To evaluate the overall impact of three health information technology (IT) initiatives at the point of care on the proportion of RA patient visits with disease activity scores recorded in an academic outpatient clinic. Methods We studied three initiatives designed to facilitate disease activity score documentation: 1. An EHR flowsheet, 2. Public reporting of physicians’ performance, 3. An EHR SmartForm to facilitate the calculation and documentation of a validated RA measure, the Clinical disease activity index (CDAI). The initiatives were implemented at different time points over the study period. The study cohort included all adult RA patients with at least 2 face-to-face encounters in the outpatient rheumatology clinic at University of California, San Francisco between June 2012 – June 2017. Clinical data and covariates were retrieved from the EHR data warehouse. To examine trends in CDAI documentation over time, we created a quality control chart (p-chart) (figure 1, where vertical lines indicate onset of each of the initiatives). For each initiative, we analysed our data using a two-pronged pre-post approach. First, we developed multiple logistic regression models in which the outcome was documentation of CDAI, controlling for patient age, sex, race/ethnicity, language and insurance category as well as physician sex and years in practice. Second, logistic mixed effect models were used to account for repeated visits by patients to the clinic and clustering by physicians. Results We included data from 7406 encounters from 978 unique patients. Mean (SD) age was 58.9 (16) years, 82% were female, 44% were racial/ethnic minorities, and 59% had public insurance. Over a 60 month period, overall documentation of CDAI scores per month in the clinic increased from 0% to 64% (figure 1). Results from mixed effect logistic modelling showed that Initiative 1 significantly increased CDAI recordings (OR=40.14 p Conclusions Introduction of a flowsheet and public reporting of physician performance within the practice significantly improved performance, but institution of the SmartForm did not further improve on these gains. However, gains were maintained through the end of the five-year study period. Future work should focus on whether improved CDAI documentation is associated with improved patient outcomes, such as lower disease activity and improved physical function. Disclosure of Interest None declared