Findings question the applicability of widely used trauma triage scores to assess older patients

TAMPA, Fla. — A retrospective registry review of trauma patients older than 55 years highlighted the Injury Severity Score as a poor predictor of mortality and the Trauma Score-Injury Severity Score as lacking specificity in its predictability.

“We feel that the [Trauma Score-Injury Severity Score] TRISS and [Injury Severity Score] ISS are both poor predictors of mortality in this at-risk, intermediate, geriatric trauma population. Furthermore, older age, lower [Glasgow Coma Scale] GCS and a lower mechanism of injury are associated with a higher mortality rate in this at-risk patient cohort,” Matthew Wilson, MD, said at the Orthopaedic Trauma Association Annual Meeting.

To assess whether existing triage scores were predictive of prognosis in a geriatric trauma population and to identify variables that predict mortality, the researchers identified 237 patients aged 55 years or older who were admitted to a level 1 trauma center and had a predicted probability of survival of 10% to 75% using the TRISS. Patients were then categorized into survivor and non-survivor groups for study. Patients had an average age of 70.7 years. More than half of the patients had low-energy mechanisms of injury and 40% of all patients were men.

“Our observed mortality was 46%, while the predicted mortality based on the TRISS was 59.2%,” Wilson said. Although the investigators found no difference between survivors and non-survivors regarding gender, non-survivors had a significantly higher mean age compared to survivors. In addition, non-survivors were more likely to sustain injuries from a low- vs. high-energy mechanism.

“We found that the TRISS was predictive of survival, but the ISS demonstrated a trend toward higher scores or worse injuries in the survivors as compared to the non-survivors, and the patients in our cohort who survived had a higher probability of survival based on the TRISS, but the difference between groups was quite small,” he said.

A receiver operator characteristic curve using a TRISS cut-off of 0.27 to maximize specificity and sensitivity showed the area under the curve was 0.65, Wilson said, suggesting the TRISS has a poor predictive capacity in this cohort of patients.

“We found mortality rates were higher in those patients with a lower GCS, lower respiratory rate, lower temperature and lower hematocrit rate. Traditionally accepted physiologic indices and measures of resuscitation, such as pulse rate, systolic blood pressure, shock index and base deficient, were not significantly different between the two groups,” he said.

Wilson added that existing trauma triage measures may be missing key variables, such as mechanism of injury, that impact prognosis of traumatic injuries in the geriatric population and noted it is imperative that a triage tool specific to the geriatric trauma population reflecting appropriately weighted physiologic variables and other injury-specific indices should be developed, refined and validated. — by Gina Brockenbrough, MA

Reference:

Wilson M. Paper #53. Presented at: Orthopaedic Trauma Association Annual Meeting; Oct. 16-18, 2014; Tampa, Fla.

Disclosure: Wilson has no relevant financial disclosures. The study was funded in part by an internal grant though the Carolinas Healthcare System Center of Excellence.