The ECMOnet score: a useful tool not to be taken absolutely
- PDF / 139,921 Bytes
- 2 Pages / 595.276 x 790.866 pts Page_size
- 47 Downloads / 208 Views
Thomas Mu¨ller Stephan Schroll Alois Philipp Christian Karagiannidis Matthias Amann Dirk Lunz Julia Langgartner Thomas Bein Marcus Fischer Matthias Lubnow
The ECMOnet score: a useful tool not to be taken absolutely Accepted: 29 April 2013 Published online: 22 May 2013 Ó Springer-Verlag Berlin Heidelberg and ESICM 2013
CO RRESPONDENCE
influenza-A and the need for venovenous ECMO since 2009 (Table 1), in addition to common respiratory parameters we tested static compliance (CRS) and corrected expired volume per minute (VEcorr), because a recent post hoc analysis showed that either a CRS of 20 ml/cm H2O or less or a VEcorr of at least 13 l/min in combination with a PaO2/FiO2 of 100 mm Hg or less identifies a higher risk subgroup [2, 3]. However, this may not be true for patients on ECMO, as we could not see a difference in our small sample. Failure of extrapulmonary organs may be a more important predictor, which is mirrored in two parameters of the new ECMOnet score, bilirubin
and creatinine. In addition, hospital length of stay before ECMO institution was found to be a predictor of death, which parallels information from the ELSO database, that length of ventilation before ECMO is associated with a worse outcome [4]. This parameter might be falsified, if H1N1 is acquired in hospital, as we have seen in several referred cases. Parameter 4 (hematocrit) and parameter 5 (mean arterial pressure) may be iatrogenically corrected by blood transfusions and vasopressors, and we wonder whether this was taken into account when calculating the score. In our cohort, further major contributors to a poor outcome were chronic
Table 1 Patient characteristics before ECMO
Dear Editor, We read with great interest the article by Pappalardo and colleagues [1], and very much appreciate their effort in identifying predictors of mortality in patients on ECMO with lung failure induced by H1N1-influenza-A using the newly developed ECMOnet score. We congratulate the authors and all participating centers on the achievement of a nationwide database for these patients, which opens up new approaches to collectively improving the outcome in these patients. Predicting the risk of mortality is important and desirable in guiding the use of restricted treatment resources such as ECMO. It is not an easy undertaking, as very high accuracy is needed to avoid the grave consequences of an incorrect decision. We agree with the authors that ventilatory and blood gas parameters before ECMO cannot sufficiently predict the final outcome in a patient on ECMO, which may be explained by the fact that the extracorporeal device will support and partly substitute for respiratory function. In our own cohort of 38 patients with H1N1-
Age (years) Male/female (n) Weight (kg) BMI (kg/m2) Hospital stay (days) Mechanical ventilation (days) SOFA score Lung injury score Bilirubin (mg/dl) Creatinine (mg/dl) Hematocrit (%) Hemoglobin (g/dl) Lactate (mg/dl) PaCO2 (mmHg) PaO2/FiO2 (mmHg) pH PEEP (cmH2O) PIP (cmH2O) VT (ml) MV (l/min) VEcorr (l/min) CRS (mL/cmH2O) Mean arterial
Data Loading...
