Evaluating the appropriateness of antimicrobial treatment in hospitalized patients: a comparison of three methods

Background: Approximately 20–50% of antimicrobial therapy in hospitalized patients is considered inappropriate, which may be associated with increased morbidity and mortality. The best method for evaluation of appropriateness is not well defined. Aim: To evaluate the rate of appropriate antimicrobial therapy in a secondary hospital using three different methods, and determine the rate of agreement between the different methods. Methods: A point prevalence study included all adult hospitalized patients receiving systemic antimicrobial therapy during 2016, screened on a single day. Clinical, laboratory and therapeutic data were collected from patient files, and appropriateness was rated with a qualitative evaluation by expert opinion. In addition, a quantitative evaluation was performed according to 11 quality indicators (QIs) rated for each patient. A strict definition of appropriateness was fulfilled if six essential QIs were met, and a lenient definition was fulfilled if at least five QIs were met. Agreement between methods was analysed using kappa statistic. Findings: Among 106 patients included, rates of appropriateness of antimicrobial therapy ranged from 20% to 75%, depending on the method of evaluation. Very low agreement was found between the strict definition and expert opinion (kappa=0.068), and medium agreement was found between the lenient definition and expert opinion (kappa=0.45). Conclusions: Rates of appropriateness of antimicrobial therapy varied between evaluation methods, with low to moderate agreement between the different methods.