Trauma Severity Scores Inaccurate Predictors of Mortality in Older Patients
October 9th, 2016
CHARLOTTE, NC – Tools used to predict mortality in trauma patients do not work very well in an older cohort being treated at an urban Level 1 trauma center, according to a new study.
The research, led by investigators from the Carolinas Trauma Network in Charlotte, instead suggested that traditional triage factors such as age, vital signs and anatomic injuries might require different weighting in the geriatric trauma population to account for age-related differences in injury characteristics and physiology. The study was presented recently at the Orthopedic Trauma Association annual meeting in Tampa.
For the study, researchers identified patients 55 or older who had presented to a Level 1 trauma center from 2008-2011, focusing on those with a predicted probability of survival of 10%-75% based on the Trauma Score-Injury Severity Score (TRISS).
After accounting for 10 patients who had died in the ED, a cohort of 237 patients was divided into eventual survivors and non-survivors for analysis. Noting that age, mechanism of injury, laboratory values, and vital signs upon arrival at the trauma center are triage variables reported to have a role in survival prediction, the researchers calculated the injury severity score (ISS) and TRISS were for both survivor and non-survivors.
During the index hospitalization, 109 (46%) of the patients and 28 (54%) survived, with no gender difference noted. The mean age for non-survivors, 74, was significantly higher than for survivors, who average 67, however, and a much higher percentage of the non-survivors, 68%, were injured as result of a low energy mechanism, i.e. a fall from a standing height, compared to the survivors, 43%.
The Glasgow Coma Scale was notably lower for non-survivors vs. survivors as were temperature, respiratory rate and hematocrit, according to the study. Pulse rate, blood pressure, shock index and base deficit on arrival were not significantly different, however.
The review found that the TRISS was predictive of survival (TRISS 0.35 vs. 0.46; P < 0.001) while the ISS was significantly lower for non-survivors than survivors (ISS 23 vs. 26; P <0 .001).
“In spite of its widespread adoption and use, the ISS is a poor predictor of mortality in an intermediate-risk geriatric trauma population as evidenced by lower triage scores for non-survivors when compared with survivors,” the authors conclude. “Those patients in our cohort who survived had a higher probability of survival based on the TRISS, but the difference between groups was quite small, suggesting that the TRISS lacks the requisite specificity to be used as an accurate prediction model in the geriatric patient.”
The researchers note that older age, lower GCS, and a low-energy mechanism of injury were associated with a higher mortality rate at the urban Level I trauma center.
“Given the inability of existing measures to adequately predict mortality in older adults, existing measures may be missing key variables that impact survival of traumatic injuries,” they add. “This information sets the stage for the development of a triage tool specific to the geriatric trauma population with appropriately weighted risk factors.