Meta-analysis of Polymyxin Use in Patients

In this chapter, we systematically reviewed studies that assessed polymyxin's effectiveness and summarized results through meta-analysis. The outcomes addressed were all-cause mortality, assuming that for patients with severe multidrug-resistant infections survival is the most important outcome, and resistance development, important for future patients. Most clinical data on polymyxins in the literature are from retrospective, observational studies at high risk of bias. The majority of clinical studies were unpowered to examine mortality controlling for other risk factors. The studies had no control of dosage regimens and treatment modifications. We identified several areas of missing data, in particular randomized controlled trials (RCTs) examining treatment options for carbapenem-resistant Gram-negative bacteria, different dosage regimens, polymyxins versus alternative antibiotics (e.g. aminoglycosides, tigecycline), and monotherapy versus specific combination therapies. Ideally, mortality and development of resistance should be examined in RCTs, with further longitudinal studies required for the latter.