Improving Diagnostic Testing Decisions For Pediatric Minor Head Trauma In The Emergency Department: A Two-Year Prospective Implementation Study

Research Objective The risk of serious intracranial injury in pediatric patients with minor head trauma (MHT) is less than 5%; most computerized tomography (CT) scans in MHT are normal or contribute little to management, yet expose children to unnecessary radiation. Despite evidence‐based risk classification criteria from the Pediatric Emergency Care Applied Research Network (PECARN) for assessing appropriate CT use during emergency department (ED) visits, barriers persist to replacing unnecessary scans with structured observation. Field readiness assessments at Intermountain Healthcare suggest that physicians often believe they know the risk factors for traumatic brain injury (TBI) but sometimes misremember elements. Information retrieval when delivering ED care can be cumbersome. Many physicians also perceive ordering CT scans is the safest course of action despite a lack of significant symptoms. We theorized that targeting evidence‐based education at the individual scan decision point, coupled with timely performance feedback, would increase cognitive support for assessing risk of clinically‐important TBI (ciTBI), reducing potentially unnecessary scans. Study Design We conducted a prospective pre‐post comparison implementation study. The primary implementation strategies were two‐fold. First, we embedded an alert containing an easy‐to‐understand, information‐rich graphic providing current PECARN risk stratification criteria and supporting evidence for classifying ciTBI, along with a risk assessment prompt linked to a CT order. Second, we provided timely feedback on performance and local prompting to educate physicians. Uptake and effectiveness measures included % adherence change in PECARN guidelines and the CT scan rate. Safety was evaluated by counting 48‐hour readmissions with clinical evidence of ciTBI confirmed via chart review. Acceptability, fidelity and feasibility were assessed using qualitative analysis. Statistical analysis was conducted using tests of proportions. Population Studied Approximately 14,000 pediatric patients presenting with MHT at 22 EDs from January 2019–December 2020 within a single, integrated delivery system including urban, rural and frontier locations and a children's hospital. Principal Findings Year 1 adherence to PECARN guidelines was 98.7% with a 14% reduction in the CT scan rate for pediatric MHT patients across geographies with no readmissions for ciTBI (Table). Results were sustained in Year 2 despite increased patient acuity in 2020 due to the novel coronavirus pandemic. Subsequent field discussions found good acceptance by physicians noting the alert was relevant, timely and easy to understand. Implementation fidelity was high given routinization of the alert into clinical workflow. Conclusions Combining local performance feedback with use of an information‐rich text alert was associated with significant improvements in adherence to PECARN guidelines and a reduction in the CT scan order rate for diagnosis of clinically‐important TBI in MHT patients without impacting safety. The routinized nature of the alert was associated with good practice sustainment over multiple years across geographies. Implications for Policy or Practice Simple, information‐rich text alerts may prove useful as an implementation strategy for updating physicians on changes in evidence‐based triage and risk classification criteria often associated with de‐implementation of legacy clinical practices. Measure Baseline‐Dec2018 Rolling‐6mo Uptake‐Dec2019 Rolling‐12mo Sustainment‐Dec2020 Rolling‐12mo z‐score/p‐value % adherence to PECARN guidelines 95.66% 98.74% 99.32% z = −10.3; p < 0.001 CT scan rate 33% 29% 29% z = 6.0; p < 0.001 48‐hour readmission for ciTBI 0 0 0