Abstract 200: Discordance Between Monitor-measured and EMS Documented Respiratory Rates in Major Traumatic Brain Injury: Implications for Injury Scoring Systems

Abstract

Background: Respiratory rate (RR) is a key component in commonly-used trauma scoring systems [e.g., Revised Trauma Score (RTS), TRISS]. Imprecise documentation of RR introduces misclassification when these tools are used in trauma research. By identifying each waveform, nasal cannula end tidal CO2 (NCCO2) accurately measures RR in non-intubated patients. Objective: Evaluate the relationship between EMS-documented RR measurements in patient care records (PCRs) vs. true RR recorded by non-invasive NCCO2 monitoring in major TBI patients who were never actively ventilated. Methods: Among spontaneously-breathing, major TBI cases (moderate/severe/critical), continuous NCCO2 data (Philips MRx™) were evaluated from the EPIC Prehospital TBI Study (NIH 1R01NS071049). RR classifications for RTS/TRISS were then established for each case using both PCR-documentation and monitor data. Routine monitor data (including RR) were available to EMS providers on the display at all times during care. Results: Included: 158 cases from 7 Arizona EMS agencies [(7/13-7/17; median age 55 (range 18-94); 65% male]. The Table shows RTS/TRISS case classification by PCR and monitor RR. PCR-documented RR frequently failed to correctly classify cases: RR <6 (0/10; 0%); 6-9 (3/21; 14.3%; >29: (11/34, 32.4%), normal (67/93, 72.0%; Table). In total, PCR documentation misclassified 48.7% of cases (77/158). Conclusion: These findings identify a major contributor to inaccurate trauma scoring. Since RTS and TRISS are used widely in research, this has important implications for study enrollment, case ascertainment, confounding, and risk-adjustment in injury studies. Whenever possible, QI and research studies should utilize monitor data to identify and evaluate RR and other vitals rather than relying on PCR documentation. Future development of monitor-based, real-time feedback technology might improve trauma scoring precision and provider identification of RR abnormalities.