Facebook Twitter YouTube


Development of Computerized Clinical Decision Support (CDS) to Implement the Pecarn Clinical Prediction Rules for Children with Minor Blunt Head Trauma: A Multicenter Clinical Trial

Sunday, October 21, 2012
Room 270 (Morial Convention Center)
Marguerite Swietlik, MSN, CPNP1, Peter Dayan, MD, MSc1, Nathan Kuppermann, MD1, Howard Goldberg, MD2, Marilyn Paterno, MBI2, Deepika Pabbathi2, Molly Schaeffer2, Beatriz Rocha, MD, PhD2, Jeffrey Hoffman, MD1, Robert Grundmeier, MD1, Dustin Ballard, MD, MBE3, Evaline Alessandrini, MD1, Lalit Bajaj, MD, MPH1, Leah Tzimenatos, MD1, Robert Norris, MD3, Steve Offerman1, Dustin Mark1, David R. Vinson1, Mamata Kene1, Grant Jones1, Mike Dean1, Charlie Casper1, Eric Tham, MD, MS1 and Pediatric Emergency Care Applied Research Network1, (1)Pediatric Emergency Care Applied Research Network (PECARN), New York, NY, (2)Partners HealthCare System, Wellesley Hills, MA, (3)Clinical Research in Emergency Services & Treatments (CREST) Network

Purpose:To develop standardized CDS to implement the PECARN prediction rules for children with minor blunt head trauma in a multicenter clinical trial. Across 8 US emergency departments the CDS will provide real-time delivery of predictive statements regarding risk of clinically important traumatic brain injury (ciTBI) and recommendations concerning the need for head CTs based on patient-specific data entered into a single-platform electronic health record (EHR) by clinicians.  The CDS has the potential to reduce clinically-unnecessary head CTs.

Case Report: Through computerized CDS, we will implement the PECARN head trauma prediction rules to identify patients at very low risk for ciTBI. All sites receiving CDS use Epic® (Madison, WI) and are collecting standardized clinical assessment data for the study in the EHR. Six sites will use Epic®-based CDS in which rules are built, analyzed and risk statements delivered using Epic® CDS functionality. Two sites will use remote, web-service-based CDS. The Epic-based CDS is being built at one site, pilot tested, and exported to all others. For sites accessing the web service CDS, assessment data will be transmitted to Partners HealthCare’s Enterprise Clinical Rules Service (ECRS). Partner’s ECRS will return risk statements consistent with those developed in Epic’s® internal CDS rules engine. The risk statements will be presented to the provider in real-time (as soon as data entry is complete) and stored in the Epic® EHR. The context, language and format of the risk statements are being developed iteratively among the research, clinical, and informatics personnel across all sites. Knowledge management experts at PHS are playing a key role in developing the many risk statements to ensure both CDS strategies provide the same information for similar patients. To accept the build, each site will reserve a set range of EHR identification numbers and perform site-based testing. Sites accessing the web services CDS are deploying middleware to allow Epic® to call an external web service and store the returned risk statements in the EHR. CDS build for both Epic® and web services includes more than 250 criteria combinations and more than 30 risk statements. Contingency planning includes the development of statements for patients not meeting the PECARN very low risk criteria for ciTBI, those having incomplete data and those for whom a precise estimate of ciTBI is unavailable. 

Discussion: Development and dissemination of CDS across a multicenter study has enormous potential for improving consistency in delivery of care. Our central development strategy for the CDS build with distribution via export-import decreases risk of variability in the build across sites, decreases risk of build error and allows for uniform collection and comparison of outcomes data.