Background: Glasgow Coma Scale (GCS) and the pupillary light reflex (PLR) are important prognostic tools for traumatic brain injury (TBI). This study compared the predictability of GCS, GCS plus manual PLR (GCS-P), GCS plus Neurological Pupil index (GCS-NPi), and average NPi (avgNPi) in predicting discharge outcome in patients diagnosed with TBI. Methods: Data were obtained from a multicenter prospective registry that included 175 subjects with TBI. A nonlinear mixed model (NLMIXED) approach was used to determine which of the following independent variables (GCS, GCS-P, GCS-NPi, and avgNPi) is a better predictor of modified Rankin Scale (mRS) at discharge by fitting four predictive models for comparison. Results: The NLMIXED model for longitudinal data determined that GCS, GCS-P, GCS-NPi, and avgNPi were all significant predictors of mRS at discharge (P<0.001). Age was a significant predictor of the discharge mRS (P<0.001). There was a strong significant correlation between the four predicting variables (P<0.05). The maximum likelihood estimation (MLE) of GCS was -0.17 (P<0.001), MLE of GCS-P was -0.17 (P<0.001), MLE of GCS-NPi was -0.17 (P<0.001), and the MLE of avgNPi was -0.39 (P<0.001).
Conclusion:Our findings suggest that any of the four variables (GCS, GCS-P, GCS-NPi, and avgNPi) could be used as a potential predictor of discharge mRS in a patient with TBI. This warrants future investigations to explore the combination of pupillary reactivity scores and NPi with GCS for prognostication in patients with TBI.