Purpose of review
The review examines potential confounders hampering measurement of the impact of multidrug resistance in nosocomial infections. Methodological techniques dealing with the problem of confounding are discussed and current findings in how multidrug resistance affects outcome in patients with nosocomial infection are highlighted.
Outcome comparisons between patients infected with multidrug-resistant pathogens and patients infected with susceptible microorganisms are hampered by confounders such as differences in disease severity, prolonged hospitalization prior to onset of infection (exposure time), the causative pathogen, the type of infection, and the rate of appropriate empirical antimicrobial therapy. The confounding effect can be countered by means of either multivariable regression techniques or matched cohort studies, or a combination of both. Recent literature on the impact of multidrug resistance (methicillin-resistant Staphylococcus aureus, vancomycin-resistant enterococci, extended-spectrum β-lactamase producing Enterobacteriaceae, etc) is conflicting and highly dependable on the way disturbing variables are accounted for.
Recent data underscore that the impact of multidrug resistance on the outcome of nosocomial infection might differ depending on the study population, type of infection, type of pathogen and appropriateness of therapy, and hence, that any conclusion drawn prior to accurate accounting for imbalanced confounders is premature and potentially false.