ECMO for infection

To the Editor.— Dr Hocker and his colleagues1 are to be commended for taking on the difficult task of evaluating the success of ECMO in early-onset group B streptococcal infection when a controlled trial is ethically and practically not feasible. In their study they used historical controls (a commonly criticized method, but all that is realistically available) to identify risk factors for early-onset group B streptococcal sepsis and to evaluate the effectiveness of ECMO intervention.