A systematic review of the methods used to assess the association between appropriate antibiotic therapy and mortality in bacteremic patients.

Studies of the association between inappropriate antibiotic therapy and mortality among bacteremic patients have generated conflicting findings. We systematically reviewed these studies to identify methodological heterogeneity that may explain the lack of agreement. We identified 51 articles that met the inclusion criteria, and we extracted the following data: study design, definition and measurement of variables, and statistical methods. Only 8 studies (16%) defined inappropriate antibiotic therapy as that which was inactive in vitro against the isolated organism(s) and not consistent with current clinical practice recommendations and distinguished between empiric and definitive treatment. Thirty-four studies (67%) measured the severity of illness, but only 6 (12%) specified the time at which it was measured. The methodological recommendations suggested in this article are intended to improve the validity and generalizability of future research. In brief, future studies should define "inappropriate" therapy on the basis of in vitro susceptibility data, should separately evaluate empiric and definitive therapy, and should control for the baseline severity of illness.