Title Design, Implementation and Evaluation of rational antibiotic prescribing decision support system to neonatal intensive care unit

Abstract

Introduction: In order to prevent further complications caused by infection, physicians start experimental antibiotic treatment for neonates only based on clinical symptoms. Therefore, in many cases, this causes the prescription of unnecessary or inappropriate antibiotics and, as a result, creates antibiotic resistance. Purpose: The purpose of this research is to design and evaluate a decision support system for the rational prescription of antibiotics, based on the risk of infection and the pattern of antibiotic resistance. Methods and Materials: The current research was conducted in three phases: descriptive, developmental, and evaluation. In the descriptive phase, first by conducting a systematic review and meta-analysis, the pattern of antibiotic resistance in Iran was determined. Then, during a cross-sectional study, the category of problems related to antibiotic treatment in the neonatal intensive care unit was determined. Further, the clinical risk factors of developing sepsis in newborns were identified using a fuzzy Delphi study, and the decision rules for sepsis diagnosis were finalized by holding focused group meetings. In the development phase, the system was designed based on a user-centered method and a three-layer architecture in the ASP.net environment. In the evaluation phase, the accuracy, sensitivity, and specificity of the system were determined. The impact of the system on the prescribing behavior of physicians based on the scenario was evaluated and finally, the applicability of the system was evaluated. Findings: The knowledge base was designed based on 26 clinical risk factors and 24 decision rules regarding neonatal sepsis. Then, the decision support system was designed based on the category of antibiotic treatment problems and the type of antibiotic resistance pattern in each geographical region, web-based and in the form of three data entry modules, decision support, and reports. The accuracy, sensitivity, and specificity of the system are 85.7%, 83.3%, and 90%, respectively. The system had a significant effect on reducing the number of suspected sepsis cases and changing sepsis risk groups (P<0.001). In 76.7% of cases, after using the system, prescriptions were modified. In evaluating the usability of the system, the usefulness and satisfaction of the system had the highest score. Conclusion: The decision support system for rational antibiotic prescription helps the neonatologist to accurately diagnose sepsis. As a result, the use of unimportant antibiotics can be reduced, and wrong prescriptions can be avoided to a large extent. Keywords: Neonatal infection, Antibiotic, Antibiotic resistance, Drug-related problems, Experimental treatment, Sepsis risk factor, infant, Decision support system, Medication error, Rational drug prescription