Controversia sugli steroidi nella sepsi e shock settico: una meta-analisi

CRD COMMENTARY:

This is a clearly-written review. A more extensive literature search may have revealed more relevant articles though contact with experts in the field should have revealed most of them. The studies were assessed for quality but no details of the factors used were given. The authors highlight problems in the methodology of the primary studies; these include the lack of consistent definition of septic shock, insufficient detail on the patient characteristics, mortality rates defined at different points in time, variable type of steroid, variable duration of steroid therapy, trials over a considerable time period (from 1963 to 1988), lack of information on the time of onset of treatment, and differences in study populations, as reflected in the variability of mortality in the control groups. There was some investigation of heterogeneity but further investigation of the effect of quality assessment and the year of study on the result would have been welcome, as would heterogeneity testing of the studies used to assess the effectiveness of steroids in Gram-negative infections. In view of the lack of statistical significance of the effect of corticosteroids on patients with Gram-negative septicaemia, and the lack of information on the primary studies, it is not possible to support the authors' conclusion that there is evidence of a positive effect of steroids for this subgroup of patients.