Comparison of statistical methods of antimicrobial resistance based on clinically isolated bacteria and infectionrelated nonrepetitive bacteria

Abstract:

ObjectiveTo compare the differences between two statistical methods for evaluating nonsensitivity of pathogenic bacteria to antimicrobial agents, and explore effect of nonconsideration of clinical background on evaluating extent of bacterial resistance.MethodsData of Staphylococcus aureus and Acinetobacter spp. in a hospital in the first half year of 2008, 2010 and 2013 were collected and conducted statistical analysis with two methods (method 1: based on all clinically isolated bacteria; method 2 : based on infectionrelated nonrepetitive bacteria), two methods for evaluating bacterial nonsensitive  rates to antimicrobial agents were compared.ResultsThe nonsensitive rates of Acinetobacter spp. to various antimicrobial agents : statistical results by using method 1 were generally higher than those using method 2， absolute difference between two statistical methods was 10.46%-33.77%; the nonsensitive rates of Staphylococcus aureus to various antimicrobial agents : except compound sulfamethoxazole in 2010 and 2013(difference were 6.17% and 10.21% respectively), penicillin G (difference was 3.86%), erythromycin (difference was 2.71%), and azithromycin in 2013 (difference was 2.43%), statistical results by using method 1 were generally higher than those using method 2， absolute difference between two statistical methods was 0-18.04%.ConclusionThere are deviation in the nonsensitive rates of bacterial strains to antimicrobial agents by using two different statistical methods, deviation is larger in Acinetobacter spp.. The resistance level might be incorrectly higher when evaluating the resistance status without considering clinical background of bacteria.