Category: Professional Posters
Purpose: Patients with AF are five times more likely to experience a stroke. Current guidelines recommend the use of stroke risk scores such as the CHA2DS2-VASc to identify high-risk patients for primary stroke prevention through prophylactic anticoagulation. Recent studies found that approximately 39% of high-risk non-valvular AF patients remain untreated. This study was designed to assess baseline rates of risk score documentation and anticoagulation to determine local guideline compliance and areas of opportunity for electronic clinical decision support tools.
Methods: This baseline assessment was conducted through retrospective analysis of electronic health record (EHR) data. An EHR report was built to query for patients diagnosed with AF during a face-to-face, emergency department, or admission encounter from January 1, 2018 to December 31, 2018. The report was designed to first identify the total number of unique AF patients, and then determine the rates of risk score documentation and anticoagulation in this population. Documentation rates for the currently recommended CHA2DS2-VASc as well as the older CHADS2 score were collected. Due to limitations with the data source, scores were only captured if they were entered in structured fields (i.e. flowsheets). Data were also collected regarding encounter types associated with score documentation, with a random sample of encounters manually reviewed. With recent guidelines recommending the use of direct oral anticoagulants (DOAC) over warfarin in eligible patients, the report was built to capture separate utilization rates for warfarin, apixaban, betrixaban, dabigatran, edoxaban, and rivaroxaban.
Results: The EHR report identified 16,230 unique AF patients. The report found that the CHADS2 score was not utilized during the specified timeframe but had historically (prior to 2018) been documented for 17% (n= 2,753) of patients. During 2018 a CHA2DS2-VASc score was documented for 12% (n= 1,887) of patients. The majority of documented scores were attributable to an anticoagulation clinic encounter (84%) or telephone encounter (15%). A supplemental EHR query searching for CHA2DS2-VASc scores documented in unstructured fields (i.e. freeform provider notes) yielded a variety of formats and documentation methods that made utilization of these scores for any standardized systemic intervention unfeasible. Of the unique AF patients 60% (n= 9,771) had an oral anticoagulant ordered during the study period. Apixaban was ordered most frequently (31%), followed by warfarin (20%) and rivaroxaban (13%), with negligible use of dabigatran, edoxaban, and betrixaban.
Conclusion: This baseline assessment revealed that only 12% of AF patients in our system had stroke risk scores documented in structured data fields. Risk scores documented in unstructured fields could not be reliably abstracted. To be able to employ electronic clinical decision support tools to promote guideline compliance, standardized risk score documentation must be improved upon first.