J. Baker1, D. A. Millar1, V. Heh1, M. Goodman1, T. Pritts1, C. Janowak1 1University Of Cincinnati,Division Of Trauma And Critical Care Surgery,Cincinnati, OH, USA
Introduction:
There has been significant interest into the consequences of chest wall injury on the outcomes injured patients. It is assumed that worse severity of injury correlates with worse outcomes. The two most widespread chest wall injury severity scores, the organ injury scale (OIS) and abbreviated injury scale (AIS), are based on expert opinion and it is unknown if they correlate with clinical endpoints. We set out to determine if the OIS and AIS are associated with clinical outcomes.
Methods:
A single institution trauma registry was queried from January 2014 and December 2017 for all patients with bony chest wall injury. All patient with blunt chest wall injuries were included and assessed for both OIS and AIS scores on their respective scales. Outcomes assessed included mortality, complications including respiratory failure, tracheostomy, and readmissions. Descriptive statistics were performed. The receiver operating characteristics (ROC) areas under the curve (AUC) were determined to measure discriminatory accuracy of scoring systems.
Results:
Over four years, 19% of all trauma patients (n=3033) presented with a blunt chest wall injury, of which 2821 survived to discharge (7.0% mortality). The distribution of chest wall OIS scores was: 1 = 562 (18.5%), 2 = 1649 (54.4%), 3 = 621 (20.5%), 4 = 190 (6.3%), 5 = 11 (0.4%). In comparison, chest wall AIS score for the same patient was greater than OIS in 48.7%, the same in 46.7%, and lower in 4.6%. The ROC AUC for mortality, respiratory failure, and tracheostomy were 0.68, 0.64, and 0.67 respectively. The ROC AUC for overall readmission, readmission for injuries, and readmission specific for chest injuries were 0.52, 0.54, and 0.59 respectively. For readmission outcomes, the AUCs for the AIS score were slightly better than the OIS score (Table 1).
Conclusion:
There is a weak association between OIS score and mortality, respiratory failure, and tracheostomy and a poor association with hospital readmissions. The AIS score performed minimally better than OIS for discriminating readmission outcomes. It is important that outcomes-based research use scoring systems and instruments that are linked to relevant clinical outcomes. There may be an opportunity to revise chest wall injury scoring systems to facilitate future outcomes based research and practice improvement processes.
