Evaluation of the revised trauma and injury severity scores in elderly trauma patients.

BACKGROUND: Severity-of-illness scoring systems have primarily been developed for, and validated in, younger trauma patients. AIMS: We sought to determine the accuracy of the injury severity score (ISS) and the revised trauma score (RTS) in predicting mortality and hospital length of stay (LOS) in trauma patients over the age of 65 treated in our emergency department (ED).
MATERIALS AND METHODS: Using the Illinois Trauma Registry, we identified all patients 65 years and older treated in our level I trauma facility from January 2004 to November 2007. The primary outcome was death; the secondary outcome was overall hospital length of stay (LOS). We measured associations between scores and outcomes with binary logistic and linear regression.
RESULTS: A total of 347 patients, 65 years of age and older were treated in our hospital during the study period. Median age was 76 years (IQR 69-82), with median ISS 13 (IQR 8-17), and median RTS 7.8 (IQR 7.1-7.8). Overall mortality was 24%. A higher value for ISS showed a positive correlation with likelihood of death, which although statistically significant, was numerically small (OR=1.10, 95% CI 1.06 to 1.13, P<0.001). An elevated RTS had an inverse correlation to likelihood of death that was also statistically significant (OR=0.48, 95% CI 0.39 to 0.58, P<0.001). Total hospital LOS increased with increasing ISS, with statistical significance decreasing at the highest levels of ISS, but an increase in RTS not confirming the predicted decrease in total hospital LOS consistently across all ranges of RTS.
CONCLUSIONS: The ISS and the RTS were better predictors of mortality than hypothesized, but had limited correlation with hospital LOS in elderly trauma patients. Although there may be some utility in these scores when applied to the elderly population, caution is warranted if attempting to predict the prognosis of patients.

INTRODUCTION

Trauma has long been a leading cause of morbidity and mortality in our population. Many scoring systems have been derived in order to allocate appropriate resources to patients based on the severity of their injuries. The injury severity score (ISS) and revised trauma score (RTS) are instruments used to analyze the severity of a patient's illness and stratify the patient accordingly. These scoring systems have been used to predict mortality among patients. The ISS, first described in 1974, is a scoring system for patients with multiple injuries that has shown a linear correlation with mortality, morbidity and hospital length of stay (LOS).[1] The RTS, described in 1989, is a physiologic scoring system based on a patient's initial set of vital signs and glasgow coma score (GCS) that has been shown to have a strong correlation with the probability of survival.[23]

The elderly patient population increased eleven-fold between the years 1900 and 1994, while the remainder of the population only increased three-fold during that same time period.[4] The elderly population has continued to grow since that time and will continue to represent an increasing proportion of patients in our trauma bays. Additionally, elderly trauma patients have an increased morbidity and mortality for a given severity of injury, a decreased 5-year survival compared to their uninjured counterparts, as well as an increase in intensive care unit (ICU) stays and longer hospital LOS.[5-8] Nonetheless, aggressive treatment for elderly trauma patients is paramount, as patients who survive such events often return to independent living.[9] The RTS and ISS were developed on young trauma patients in the 1970s and 1980s, respectively.[1-3] As such, analysis on the use of these scoring systems to stratify the elderly patient population has revealed inconsistent results.[10-12] Determining the utility and applicability of these scoring systems to elderly trauma patients is important to understand the potential limitations in relying on these scores for predicting the prognosis of patients and treatment decisions, thereof.

We sought to determine the accuracy of the ISS and the RTS as predictors of mortality and hospital LOS in elderly trauma patients, using data from the Illinois Trauma Registry. We were also interested in mortality in our elderly population compared with mortality as predicted using the RTS and ISS. Our hypothesis was that neither RTS nor ISS would successfully predict LOS or mortality for elderly patients.

MATERIALS AND METHODS

Study Design, Setting and Population

We conducted a retrospective review of the Illinois Trauma Registry to identify patients, aged 65 years and greater, who presented at our emergency department (ED) between January 2004 and November 2007. Any patient, 65 years and older, whose injury required activation of the trauma system and admission to the hospital was included in this study. Elderly was defined as any patient over the age of 65, which was chosen as our definition based on similar cutoffs in the literature.[13-17]

The Illinois Department of Public Health requires that all level 1 or level 2 trauma centers submit data on patients who sustain traumatic injuries requiring treatment at a trauma center and are then admitted to a trauma center; are transferred to a trauma center; or are dead on arrival (DOA) or die in the ED. The Illinois Trauma Registry does not include patients who are discharged to home; patients who are dead on arrival or are not transported to the hospital; and patients who only meet criteria for trauma team activation.

The study was approved by our hospital's Institutional Review Board.

Input Measures

The following data were abstracted from the trauma registry: admission date, patient age, ISS, RTS, hospital LOS, ICU LOS and discharge information (home, nursing home, morgue). ISS and RTS are calculated during the course of the patient's care.

Outcome Measures

The primary outcome of this study was in-hospital death, defined as death at any point during hospitalization. The secondary outcome of this study was overall hospital LOS, which was measured in days.

Data Analysis

We measured the association between ISS and mortality and between RTS and mortality using binary logistic regression. Models were created with and without gender as a variable to evaluate its influence. We measured the association between ISS and overall hospital LOS and between RTS and overall hospital LOS with generalized linear models, which extend the general linear model by removing assumptions of normality, linearity, and constant variance, allowing a more robust analysis of these associations. Specifically, we incorporated ISS or RTS scores as equally spaced binned variables and created generalized Poisson log linear models, using hybrid parameter estimations, and Pearson chi-square scale parameters to obtain more conservative variance estimates and significance levels, and the sequential Sidak method to adjust for multiple comparisons. The accuracy of each score in predicting mortality was determined by calculating the area under receiver operating characteristic (ROC) curves, with either ISS or RTS as test variables, and outcome (death) as the state variable, using non-parametric distribution assumptions without adjusting for additional confounding factors. Statistical analyses were performed using IBM SPSS Statistics version 19.0 (SPSS Inc., an IBM Company, Chicago, IL), with P values less than 0.05 considered significant.

RESULTS

A total of 347 patients over age 65 were treated and admitted from January 2004 until November 2007. The median age of our patient population was 76 (IQR 69 to 82) [Table 1]. There were 153 females (44%) and 194 males (56%). Of the 347 patients in our study, 149 required a stay in the ICU. The overall in-hospital mortality for the entire patient population of this study was 24%. The median ISS score was 13 (IQR 8 to 17) [Table 1]. The median RTS score was 7.8 (IQR 7.1 to 7.8) [Table 1]. Results of logistic regression modeling are shown in Table 2. A higher value for ISS showed a positive correlation with likelihood of death, which although statistically significant, was numerically small (OR=1.10, 95% CI 1.06 to 1.13, P<0.001). An elevated RTS had an inverse correlation to likelihood of death that was also statistically significant (OR=0.48, 95% CI 0.39 to 0.58, P<0.001). Inclusion of gender in the model did not show a statistically significant dependence in outcome for the ISS (OR 0.62, 95% CI 0.35 to 1.1, P=0.10) or for the RTS (OR 0.59, 95% CI 0.31 to 1.1, P=0.11) [Table 2]. Table 3 shows the results of generalized modeling of associations of ISS and RTS with LOS. Total hospital LOS increased with increasing ISS, with statistical significance decreasing at the highest levels of ISS, but an increase in RTS not confirming the predicted decrease in total hospital LOS consistently across all ranges of RTS. Both the ISS and RTS had fair-to-moderate discriminatory power for mortality in our study population, with an area under the ROC curve for prediction of mortality with the ISS of 0.78 (95% CI 0.71 to 0.84, Std. Error=0.03) [Figure 1] and with the RTS of 0.81 (95% CI 0.75 to 0.87, Std. Error=0.03) [Figure 2].

Table 1

Patient age, injury severity scores, revised trauma scores and hospital lengths of stay of the study population

Table 2

Results of logistic regression modeling of the study population, with and without gender

Table 3

Results of generalized modeling showing association between ISS and RTS in relation to LOS

Figure 1

Area under the ROC curve for prediction of mortality with the injury severity score (ISS)

Figure 2

Area under the ROC curve for prediction of mortality with the revised trauma score (RTS)

DISCUSSION

As our patient population ages, our elderly trauma population will also continue to increase. Trauma scoring systems, such as the RTS and the ISS, are tools that can assist with the triage, prediction of prognosis and allocation of resources for trauma patients. Application of these scores to an elderly population has shown inconsistent findings.[1819]

In our study population, we found that both the RTS and ISS were better predictors of mortality than hypothesized, but had limited correlation with hospital LOS in elderly trauma patients. Patients with relatively low injury severity scores had relatively high in-hospital mortality of 24%. In our study the median age was 76, with a median RTS of 7.8 and median ISS of 13, predicting 98% survivability using the RTS and only 4-16% mortality using the ISS.[13] While these scores correlated reasonably with mortality, the mortality in our elderly trauma patient population was significantly higher than that suggested by the original scoring system descriptions. A recent study found that elderly patients with blunt trauma suffered a higher mortality at any ISS.[20] A study by Tornetta et al. addressed factors affecting morbidity and mortality in elderly trauma patients. The study included 326 elderly trauma patients over the age of 60 who had suffered from blunt trauma, with patients suffering slip and fall injuries excluded from the study. Elderly patients in their study had a mean age of 72, an overall mortality of 18% and an average ISS of 19.7.[21] A study by Perdue et al. found that mortality in elderly patients was twice that in younger patients with equivalent ISS.[7] Our patient population exhibited an increased risk of death with increased ISS, but this increase (OR of 1.10) was less than that seen with the RTS (OR of 0.48, showing decreased mortality with increasing score, Nevertheless, age itself, as a single predictor variable, showed only a small association with increase in mortality (OR of 1.04, 95% CI 1.01 to 1.08). Of note, the latest guidelines for field triage of trauma patients suggest that beginning at age 55, pre-existing co-morbidities can adversely affect a patient's recovery from traumatic injury.[22]

Studies that were unable to show that the ISS was predictive of mortality in elderly had included less severe injuries, such as slip and fall injuries.[1819] Our data excluded such patients. Elderly patients have a higher mortality for various reasons, including an increase in late mortality (>24 hours) due to in-hospital complications and a higher incidence of preexisting medical conditions, of which cardiac disease, diabetes mellitus, coagulopathy, neurologic disease, chronic obstructive pulmonary disease, malignancy, chronic kidney disease and liver disease were most commonly observed.[72023-27]

Limitations

Our analysis had several potential limitations. First, this study relied on the use of a trauma registry for data collection with potential errors, including data entry mistakes and missing data. Additionally, all patients with isolated orthopedic injuries were included in the IDPH trauma registry. We attempted to exclude patients with isolated orthopedic injuries by including only elderly patients who activated the trauma system. Since our hospital is one of the few level-one trauma centers in the area, this study may have represented a specific patient population and was therefore limited in the ability to generalize the findings to other institutions. The ISS and RTS ranges encountered in our institution may not reflect those seen at other centers. Finally, this study was a retrospective review and was therefore susceptible to the limitations inherent in such a review.

CONCLUSION

In conclusion, the ISS and the RTS were better predictors of mortality than hypothesized, but had limited correlation with hospital LOS in elderly trauma patients. Both scores had a fair-to-moderate ability to predict the outcome of mortality; however, the overall mortality of our elderly trauma patients was much higher than predicted by the scores when applied to a younger population. Although there may be some utility in these scores when applied to an elderly population, caution is warranted if attempting to predict the prognosis of patients.