Clinical limitations of in vitro testing of microorganism susceptibility.

General concepts in evaluating the clinical importance of discrepancies between in vitro susceptibility tests of microorganisms and in vivo results are reviewed, and four problematic antibacterial-bacterial combinations are discussed. The three most common in vitro testing systems--agar disk diffusion, agar dilution, and broth dilution--are designed to detect the minimum inhibitory concentration (MIC) of an antimicrobial agent. However, agar and broth systems cannot include all of the biologic variables found within the human body. Factors affecting the reliability of in vitro testing systems include the limitations in interpreting MIC data, because in vitro test conditions cannot duplicate the host environment; the variability of testing media (e.g., acidic versus alkaline, differences in cation content); and the limiting effect of protein binding on an antimicrobial agent. In vitro testing systems do not consider the pharmacokinetics of the antimicrobial agent or the postantibiotic effect, whereby microbial growth is suppressed even when the antibiotic concentration falls below the MIC. The variability in drug distribution to infection sites within the body is also not considered. The following antimicrobial agent-bacterial combinations present specific problems to the clinician: trimethoprim-sulfamethoxazole and Group D enterococcus, cephalosporins and methicillin-resistant Staphylococcus aureus, aminoglycosides and Staph. aureus, and aminoglycosides and Pseudomonas aeruginosa. Despite the limitations of currently available systems, in vitro testing of the susceptibility of microorganisms can be an invaluable tool in selecting antimicrobial therapy when it is used in conjunction with data regarding the clinical course of the infected patient.