Objective: To assess whether Somatosensory Evoked Potentials (SEPs), recorded within 24h after ICU admission, are reliable predictors of brain death (BD) in comatose patients with acquired brain injury of various aetiologies.
Methods: SEPs were classified as absent (A), pathological (P), and normal (N). Considering SEP recordings from both hemispheres, 6 patterns were identified: NN, NP, PP, NA, AP, and AA. The final endpoint was BD.
Results: Of the 203 patients included in the study, 70 (34%) evolved toward BD. The survival analysis indicated that the combination of SEP patterns in a two-graded scale (grade 1: NN-NP-PP-NA, and grade 2: AP-AA), allowed for prediction of BD with the best accuracy. This aggregation predicted BD with a sensitivity of 75.7% (CI: 64-84), a specificity of 76.6% (CI: 68-83), a positive predictive value of 64.2% (CI: 53-74) and a negative predictive value of 84.3% (CI: 77-90) in overall patients, and with a sensitivity of 75.0% (CI: 63-84), a specificity of 84.9% (CI: 75-90), a positive predictive value of 77.5% (CI: 63-88) and a negative predictive value of 84.3% (CI: 74-91) when excluding cardiac arrest.
Conclusion: It is worth including SEPs, in association with other investigations and clinical signs, in prognostic scores of BD. The early identification of patients at high risk of evolving towards BD could help physicians to optimise management.
Keywords: Brain death; Don d’organes; Mort cérébrale; Organ donation; Organ transplantation; Potentiels évoqués somesthésiques; Predictive score; Score prédictif; Somatosensory evoked potentials; Transplantation d’organes.
Copyright © 2016 Elsevier Masson SAS. All rights reserved.