A pilot study of the Trauma Score (TS) was performed from July to September 1983. The Vital Signs Score (VSS) used by the ambulance paramedics, was compared with TS. Of 266 patients suitable for study, TS data was collected for 110. Other exclusions resulted in a detailed analysis of data from 65 patients among whom there were eight deaths. There was a significant correlation between TS and VSS, however, TS more accurately defined the population at risk of death. A score > 12 correlated with a mortality of zero for the TS, but for the VSS it correlated with a mortality of 4.4%. A score ≤ 12 correlated with a mortality of 61.5% for the TS but only 30% for the VSS. Stepwise regression analysis of the TS, VSS and combinations of their components was performed to determine their capacities to predict death. A combination of three components of the TS, corresponding to the Triage Index of Champion, was a better predictor than the total TS. Neither the VSS nor any combinations of its components had the predictive capacity of the total TS. If the TS and the VSS were used to select high risk patients for a particular rescue or resuscitation protocol, and scores were selected which gave 100% sensitivity with the highest possible specificity, the positive predictive values of the TS and VSS would be respectively 61.5% and 26.7%. The protocol would be administered unnecessarily to 73.3% of patients selected by the VSS, but only to 38.5% of patients selected by the TS. The TS is proposed as an aid to triage. Audit and modification of the VSS is recommended.
|Number of pages||7|
|Journal||Australian and New Zealand Journal of Surgery|
|Publication status||Published - 1986|
- probability of death
- trauma score