Improving Provider Adoption With Adaptive Clinical Decision Support Surveillance: An Observational Study

• DOI: 10.2196/10245
• Publication Date: 2018-11-25T09:26:25Z
• Authors: Sundas Khan; Safiya Richardson; Andrew Liu; Vinodh Mechery; Lauren McCullagh; Andy Schachter; Salvatore Pardo; Thomas McGinn
• Keywords: Dashboard

Successful clinical decision support (CDS) tools can help use evidence-based medicine to effectively improve patient outcomes. However, the impact of these has been limited by low provider adoption due overtriggering, leading alert fatigue. We developed a tracking mechanism for monitoring trigger (percent total visits which tool triggers) and completed tools) rates complex CDS based on Wells criteria pulmonary embolism (PE).We aimed monitor evaluate assess whether ongoing modifications would rates.As part larger trial, was using calculate pretest probability PE at 2 tertiary centers' emergency departments (EDs). The had multiple triggers: any order D-dimer, computed tomography (CT) chest with intravenous contrast, CT angiography (CTPA), ventilation-perfusion scan, or lower extremity Doppler ultrasound. A dashboard Tableau real-time rates. Based initial tool, we conducted small focus groups key ED providers elicit barriers use. identified overtriggering non-PE-related evaluations inability testing intermediate-risk patients. Thus, modified allow group not contrast orders. dialogue box, "Are you considering this patient?" added before triggered account CTPAs ordered aortic dissection evaluation.In center 1, 95,295 patients visited during academic year. an average 509 per month (average rate 2036/30,234, 6.73%) modifications, reducing 423 1629/31,361, 5.22%). In 2, 88,956 year, triggering about 473 1892/29,706, 6.37%) 400 1534/30,006, 5.12%) afterward. resulted in significant 4.5- 3-fold increase centers 1 respectively. increased monthly from 23.20/360 (6.5%) 81.60/280.20 (29.3%) 46.60/318.80 (14.7%) 111.20/263.40 (42.6%) respectively.Close postimplementation may adoption. Adaptive user feedback targeted rates, fatigue increasing Iterative improvements significantly