Methodologic aspects of clinical studies of perioperative antibiotic prophylaxis.

Many issues affect the design, conduct, and analysis of clinical evaluations of perioperative antibiotic prophylaxis. The most important of these are the selection of appropriate procedures for evaluation, the choice of meaningful outcomes, the comparison of outcomes among groups of patients who are really comparable to one another, the unbiased assessment of outcomes, the assessment of a number of patients sufficient to provide adequate statistical power for the detection of meaningful differences, and the generalizing of results. In evaluating perioperative prophylaxis, investigators often strive to identify small absolute differences among the risks of infections with different prophylactic regimens. For many specific procedures and drug regimens, such evaluations are limited by the difficulty of studying a sufficient number of individuals. Although randomized clinical trials are the standard for clinical evaluation, observational data may also prove useful under some circumstances.