orts of Matera and colleagues’ study establishing the utility of IL-10 as a diagnostic and prognostic marker of early sepsis ; however, we have some concerns. Limiting the study to the surgical population limits the generalizability. More importantly, inclusion of 28 (of the 52) patients after on-pump cardiac surgery is concerning. Pump surgery is widely accepted to result in an intense surge in infl ammatory markers, including IL-10 . Also, the duration of pump surgery can be variable and the infl ammatory response varies with the time spent on pump. e association of IL-10 with a worse prognosis (nonsurvivor group) may therefore not be valid. Cardiac ICU protocols such as the use of perioperative antibiotics were not discussed, which may aff ect mortality. In Table 3 of their article, the confi dence interval for the odds ratio of IL-10 for the prognosis of bacteremic systemic infl ammatory response syndrome patients includes the value 1 . Including this value limits the applicability. e values of bio markers were not checked daily, and therefore we cannot rule out a new increase in the levels of IL-10 secondary to subsequent episodes of infl ammation . e Sequential Organ Failure Assessment score and IL-10 values on day 1 and day 7 correlate with mortality. e utility of an expensive and timeconsuming biomarker is questioned when a simple, quick and bedside test could predict the outcome. Despite the positive results of the marker, it is diffi cult to imagine how such information would change management in the era of the surviving sepsis guidelines .