A meta-analysis of the published literature on the effectiveness of antimicrobial soaps.

Abstract

The goal of this research was to conduct a systematic quantitative analysis of the existing data in the literature in order to determine if there is a difference between antimicrobial and nonantimicrobial soaps and to identify the methodological factors that might affect this difference. Data on hand washing efficacy and experimental conditions (sample size, wash duration, soap quantity, challenge organism, inoculum size, and neutralization method) from published studies were compiled and transferred to a relational database. A total of 25 publications, containing 374 observations, met the study selection criteria. The majority of the studies included fewer than 15 observations with each treatment and included a direct comparison between nonantimicrobial soap and antimicrobial soap. Although differences in efficacy between antimicrobial and nonantimicrobial soap were small (∼0.5-log CFU reduction difference), antimicrobial soap produced consistently statistically significantly greater reductions. This difference was true for any of the antimicrobial compounds investigated where n was >20 (chlorhexidine gluconate, iodophor, triclosan, or povidone). Average log reductions were statistically significantly greater (∼2 log CFU) when either gram-positive or gram-negative transient organisms were deliberately added to hands compared with experiments done with resident hand flora (∼0.5 log CFU). Our findings support the importance of using a high initial inoculum on the hands, well above the detection limit. The inherent variability in hand washing seen in the published literature underscores the importance of using a sufficiently large sample size to detect differences when they occur.