As a consequence, real severity of obese patients at admission could be a confounding factor on their real outcome. An extreme example of confounding is the Simpson paradox,4 in which a factor reverses the effect first observed. For example, in a study5 comparing mortality between closed and opened ICU systems, taking into account overall patients for analysis, mortality was increased in closed ICU, but, paradoxically, mortality was decreased for intubated patients and, likewise, for nonintubated patients. If the authors had not taken into account the confounding factor “intubation status,” the conclusion would have been wrong. In this study,1 in order to better control the severity despite the lack of a severity score validated on obese patients, would it be possible to match a posteriori the cause of admission in addition to the multivariate analysis?