LAUSR

BACKGROUND Veno-arterial extra-corporeal membrane oxygenation therapy (VA-ECMO) following refractory cardiogenic shock (r-CS) or cardiac arrest (r-CA) has significant morbidity and mortality. Early outcome prediction is crucial in this setting, but data on neuro-prognostication are limited. We examined the prognostic value of clinical neurological examination, using an automated device for the quantitative measurement of pupillary light reactivity. METHODS An observational cohort of sedated mechanically ventilated VA-ECMO patients was analyzed at the early phase following ECMO insertion (first 72 hours). Using the NPi®-200 automated infrared pupillometer, pupillary light reactivity was assessed repeatedly (every 12 hours) by calculating the Neurological Pupil index (NPi). Trends of NPi over time were correlated to 90-day mortality, and the prognostic performance of the NPi, alone and in combination with the 12-h PREDICT VA-ECMO score, was evaluated. RESULTS A total 100 consecutives patients were studied (51 r-CS/49 r-CA; 12-h PREDICT VA-ECMO 40%; observed 90-day survival 43%). Non-survivors (n=57) had significantly lower NPi than survivors at all time-points (all p<0.01). Abnormal NPi (<3, at any time from 24 to 72 hours) was 100% specific for 90-day mortality, with 0% false positives. Adding 12-h PREDICT VA-ECMO score to the NPi provided the best prognostic performance (specificity 100% [95% confidence interval 91-100%], sensitivity 60% [46-72%], area under the ROC curve 0.82). CONCLUSIONS Quantitative NPi alone had excellent ability to predict a poor outcome from day 1 after VA-ECMO insertion, with no false positives. Combining NPi and 12-h PREDICT-VA ECMO score increased sensitivity of outcome prediction, while maintaining 100% specificity.