BACKGROUND
Traumatic brain injury (TBI) is a leading cause of pediatric morbidity and mortality. Evidence-based guidelines for management of pediatric TBI were last updated in 2012. TBI outcomes may depend not just on injury severity but also on the timeliness and quality of care provided and on prevention of second insults. A recent retrospective multi-center study found a 1% increase in guideline adherence to be associated with a 6% decrease in inpatient mortality. We previously developed a novel, real-time algorithmic clinical decision support (CDS) system to guide pediatric TBI anesthesia care. In this study, we evaluate the impact of CDS on anesthesia provider adherence with best practice process measures and on key performance indicator (KPI) outcomes during neurosurgeries at a level I trauma center.
METHODS
CDS was activated in patients who were under 19 years with diagnosis of acute TBI undergoing urgent or emergent craniotomy. Eligible cases were automatically detected from a surgery scheduling system. The anesthesia information management system (AIMS) database was sampled every minute for KPIs of interest. Messages were generated for unwanted KPIs and displayed on the AIMS computer screen. Providers were also prompted to follow best practice TBI anesthesia care processes. Data was collected prospectively for cases with CDS activated and also in those that CDS failed to activate. Patients meeting our inclusion criteria who underwent craniotomy prior to CDS implementation were sampled retrospectively.
RESULTS
Data from 39 patient cases was included for analysis: 20 with CDS and 19 without CDS. Median ages were 8.9±5.7 years (CDS) and 9.9±5.6 years (no CDS). Both groups were approximately 75% male and epidural hematoma was the most common diagnosis. Median arrival GCS was 8 (IQR 3-14; CDS) and 8 (IQR 4-15; no CDS). CDS was associated with improvement in 5/11 TBI process measures: arterial blood gas sampled, neuromuscular blocker use, coagulopathy treated, hyperglycemia treated, and ICP transduced. Median KPI event duration was lower in the CDS group for 4/7 KPIs: Duration of hypocarbia, by 44%; hypotension, 29%; hypothermia,12%; and hyperthermia, 15%. The CDS was highly sensitive, detecting 89% of monitor KPIs and 100% of lab KPIs, and 100% specificity was achieved.
CONCLUSIONS
CDS implementation for pediatric TBI anesthesia care is feasible and reliable, with high rates of case capture and event detection. CDS was associated with improved adherence to some TBI process measures, and a reduction in duration in the majority of unwanted KPI events. This is the first demonstration of improvement in guideline adherence attributed to CDS, and these preliminary results suggest CDS as a potential strategy to reduce second insults and improve TBI guideline adherence during anesthesia care.
