Prognostic value of automated pupillometry in traumatic brain injury: a systematic review and meta-analysis

Clinical Snapshot

Key Findings

• P Value: Mortality: p-value not explicitly stated but CI excludes unity marginally; Functional outcome: statistically significant (CI excludes unity); ICP: not statistically significant (CI crosses zero)

• Effect Size: Mortality: OR 1.84 (4 studies, n = 354, I² = 84%); Functional outcome: OR 3.64 (5 studies, n = 384, I² = 39%); ICP correlation: MD −1.05 NPi units in patients with intracranial hypertension (3 studies, n = 190, not statistically significant)

• Primary Outcome: Abnormal quantitative pupillary metrics (most commonly NPi < 3 or absent reflexes) were associated with increased odds of in-hospital mortality and poor functional outcome at 3–6 months in adult TBI ICU patients

• Nnt Or Sensitivity: No NNT, sensitivity, or specificity data reported in the abstract. No diagnostic accuracy metrics (AUC, ROC) provided. Prognostic OR for poor functional outcome (3.64) is the most clinically informative estimate, though absolute risk data are not available to calculate NNT.

• Confidence Interval: Mortality: 95% CI 1.01–3.35; Functional outcome: 95% CI 1.63–8.10; ICP: 95% CI −2.65 to 0.54

Clinical Application

The NeurOptics NPi-200/300 device is commercially available and provides bedside, non-invasive, operator-independent pupillary assessment within minutes. It can be integrated into standard neurological monitoring workflows in ICUs equipped with the device. Training requirements are minimal compared to invasive ICP monitoring. However, device cost and availability may limit uptake in resource-constrained settings. Quantitative pupillometry devices (NeurOptics NPi-200/300) are available in Australia but are not universally deployed across trauma ICUs. The TGA has not issued specific guidance on QP as a prognostic tool for TBI. The device is not currently listed on the PBS or MBS as a reimbursable item. RACGP and ANZICS guidelines for TBI management do not yet formally incorporate QP as a standard prognostic metric. Australian neurosurgical and neurocritical care units managing TBI (particularly Level 1 trauma centres) may consider QP as an adjunct to existing monitoring, consistent with the Brain Trauma Foundation guidelines that emphasise multimodal monitoring. The findings of this meta-analysis, while preliminary, are relevant to Australian ICU practice given the significant burden of TBI in the Australian population (approximately 22,000 hospitalisations annually). Formal incorporation into Australian clinical guidelines would require higher-certainty evidence from prospective controlled studies. Adult patients with moderate-to-severe TBI admitted to the ICU, assessed within 72 hours of injury. Most applicable to patients requiring neurocritical care monitoring where ICP management decisions are being made. Not directly applicable to mild TBI or paediatric populations based on current evidence.

Abstract

Traumatic brain injury (TBI) remains a major global cause of mortality and long-term disability. Early prognostication is critical for guiding management decisions. Quantitative pupillometry (QP), using automated devices such as the NeurOptics NPi-200/300, enables bedside measurement of pupillary reactivity through indices like the Neurological Pupil Index (NPi), constriction velocity (CV), and dilation velocity (DV). We conducted a PRISMA-compliant systematic review and meta-analysis to evaluate the prognostic value of QP in adults with TBI. Observational studies of quantitative pupillometry within 72 h of traumatic brain injury in adult ICU patients were included. Studies limited to mild TBI were excluded, although mixed-severity cohorts were eligible. Primary outcomes were mortality and poor functional outcome at 3-6 months (GOS ≤ 3 or mRS ≥ 3). Secondary analyses assessed differences in pupillometry metrics according to intracranial hypertension defined by direct ICP measurements (> 20-25 mmHg) or proxy markers reflecting ICP-directed interventions. Odds ratios were pooled for dichotomous outcomes and mean differences for continuous outcomes using random-effects models, with heterogeneity quantified by I². Four studies (n = 354) reported mortality data. Abnormal quantitative pupillary metrics (most commonly NPi < 3 or absent reflexes) were associated with increased odds of in-hospital death (OR 1.84, 95% CI 1.01-3.35; I² = 84%). Five studies (n = 384) assessed functional outcomes, showing significantly greater odds of poor outcome among patients with abnormal pupillary metrics (OR 3.64, 95% CI 1.63-8.10; I² = 39%). Three studies (n = 190) evaluated intracranial pressure, demonstrating lower NPi values in patients with intracranial hypertension compared with those without elevated ICP, though this difference did not reach statistical significance (MD - 1.05, 95% CI - 2.65 to 0.54). Early quantitative pupillometry is associated with increased mortality and worse functional outcomes in TBI in observational studies. Lower quantitative pupillary metrics were also observed in the presence of intracranial hypertension. Despite substantial heterogeneity, these findings support its potential as a prognostic tool. Further standardization, validation and controlled studies are needed to refine its clinical application.

References

1. 1.Vieira, R. C., Duarte Dos Santos, C. S., Stracieri Janchevis Preiss, M., Rabello, P. H. M., Carlotti, C. G., & Figueiredo, E. G. (2026). Prognostic value of automated pupillometry in traumatic brain injury: a systematic review and meta-analysis. Neurosurgical Review. https://doi.org/10.1007/s00134-024-07756-2