Corticosteroids for severe CAP: the pros COMMENTARYSevere community-acquired pneumonia (CAP) occurs in approximately 10% of hospitalized patients with CAP, and it still carries a high morbidity and mortality.In a multicenter study on severe pneumococcal CAP, the mortality of this population was 29%, with high rates of patients requiring mechanical ventilation and a shock. (1)Patients with severe CAP might die despite early and adequate antibiotic treatment, which is probably partially due to an imbalanced and disproportionate local and systemic inflammatory response that contributes to the impairment of alveolar gas-exchange, sepsis and end-organ dysfunction. (2)here is no doubt that systemic adjunctive corticosteroid therapy attenuates the local and systemic inflammatory response (3) and may potentially decrease acute respiratory distress syndrome, sepsis and mortality.In a model of Pseudomonas aeruginosa in mechanically ventilated piglets, we observed a lower lung bacterial burden and less severe histological pneumonia in piglets that were treated with corticosteroids plus antibiotics. (4)In humans, several randomized controlled trials (RCTs) have been performed, with the participants largely being hospitalized, non-severe CAP patients.The results of these trials have been negative (5) or have demonstrated a reduction in the length of stay (6) or in the period required to reach clinical stability. (7)10)(11) A meta-analysis that included some of these studies (12) demonstrated that the pooled effect of steroids in severe CAP is a reduction in mortality.However, in most of the RCTs, the following pitfalls are present:1.The inclusion of patients with a low severity of CAP, i.e., PORT I to III classes, who are.These patients have a low mortality and, consequently, it would be very difficult to perform a RCT in this population with a primary end-point of mortality.2. The inclusion of patients independent of the initial level of inflammation.According to the rationale of using steroids in CAP, a high inflammatory response is imperative.Until now, this variable has not been taken into account.In addition, patients with a high inflammatory response (such as a high C-reactive protein -CRP) have a higher rates of treatment failure (13) and mortality. (14)