Evaluation of a prioritization system for mild traumatic brain injury case ascertainment from emergency department records

Objective To evaluate modifications to a tiered priority chart review system designed to efficiently identify patients with mild traumatic brain injury (mTBI), via medical chart review, that presented to an emergency department or urgent and primary care center (ED/UPCC) for research purposes. Methods We initially created a tiered priority chart review system and applied it to screening N=17,072 electronic ED/UPCC medical charts in study 1 ( Clinicaltrials.gov ID# NCT04704037 ). Chief complaints with high positive predictive value (PPV) for correctly identifying possible/probable mTBI cases were moved to a higher tier and those with low positive predictive value were downgraded to create a tiered priority chart review system. This revised system was then used in a second research study ( Clinicaltrials.gov ID# NCT05365776 ), and PPV values were calculated for the new sample (N=4,434). The original and revised tiered priority system were compared with respect to overall efficiency. Results PPV for specific chief complaint key terms varied markedly from 0% to 61% and resulted in an empirically-driven resorting of the priority tiers. After excluding clearly ineligible participants, 49% of charts reviewed in the first study and 60% of charts reviewed in the second study were identified as a possible or probable mTBI. This indicates an improvement in overall efficiency (12%; χ 2 (1)=114.7, p<.001) compared to the original system. Conclusion The revised tiered priority chart review system was more efficient at identifying patients with mTBI for the purposes of mTBI study recruitment.