Predicting mortality while on veno-venous extracorporeal membrane oxygenation

Abstract

the authors constructed a very nice model predicting the mortality risk while on veno-venous extra-corporeal membrane oxygenation (vvECMO) in patients with acute respiratory distress syndrome (ARDS) due to H1N1 pneumonia. These patients were analyzed retrospectively from a large Italian vvECMO network [2]. Until publication of this paper, inclusion and exclusion criteria for vvECMO treatment were based on expert opinion [3]. The authors constructed the model from the retrospective data of 60 patients, of which 82 % had H1N1 pneumonia with a survival rate of 71 %. The remaining ARDS patients had a survival rate of 52 %. In another validation set containing 74 patients, the authors tested their model: the area under the curve (AUC) of the receiver operating characteristic (ROC) was 0.69 to predict mortality while on ECMO with a ECMOnet score [4.5 points, with a sensitivity of 51 % and a specificity of 76 %.In our center, we have treated 44 patients with vvECMO: 50 % for bacterial pneumonia, 20 % for viral pneumonia and 27 % for autoimmune pneumonitis disease. The overall survival rate while on vvECMO was 73 %. When applying the ECMOnet score to our patients, the optimal cut-off point (calculated with the Youden index) was an ECMOnet score of 6.5. The AUC with an ECMOnet score [6.5 was exactly 0.69 for predicting mortality while on ECMO with a sensitivity of 50 % and a specificity of 80 % (see Fig. 1). AUC of the oxygenation index and APACHEII was 0.52 and 0.54, respectively. Therefore, it seems that the ECMOnet score is a promising scoring system with a fair specificity in predicting mortality while on vvECMO, although with a higher cut-off point. At least, the ECMOnet score is much better than traditional scoring systems such as oxygenation index or APACHEII. Intensive Care Med

Prediction of mortality in adult patients with severe acute lung failure receiving veno-venous extracorporeal membrane oxygenation: a prospective observational study

“…Comparison of original and recalibrated estimates revealed considerable differences between the study populations, but continued poor discrimination after recalibration indicated that the combination of predictors was not useful in our study population. The ECMOnet score has been externally validated twice in patient populations with a similar profile as the original study [6,14]. Recent results indicate that patients receiving ECMO due to influenza-A virus infection have a lower mortality than patients presenting with other causes of ALF [13].…”