PS-120 Challenges in evaluation of electronic switch modules for home medication

Abstract

Background Most clinical decision support systems (CDSS) offer the opportunity to automatically switch patients’ home drug treatment to drugs that are listed in the hospital formulary. Only limited data are available regarding the quality of those automatic switches in clinical practice. Purpose To evaluate a methodology for comparing electronic switch modules incorporated in hospital CDSS. Material and methods A classification model with 13 categories and a six-item scale was developed to determine the quality of switches from home to hospital medicines. This model was applied to 250 drugs and three different CDSS, implemented in three university hospitals. Electronic switches were compared to manual switches by two experienced clinical pharmacists for each hospital. The functionalities of the systems were assessed by a questionnaire. Results The type of switch differed significantly within the three hospitals: same drug: 27–36%, generic substitution: 48–62%, therapeutic substitution (TS): 4–12%, special order (SO): 3–13%. Although the majority of switches was conducted correctly in our setting, the quality ranking of the switch differed significantly, especially in case of TS and SO, although the numbers of those switches were low. Incorrect and/or incomplete switches were caused either by missing alternatives in the hospital formulary and/or limited functionality of the software in terms of customisation. In theory, all types of switches were possible with each system; however, the algorithms and the covariates considered with which to suggest an appropriate drug were different. Each system had different functionalities with no system scored highly in all aspects. Conclusion The assessment of switch quality seems to depend on the diversity of the hospital formulary and on the ability to implement hospital-specific policies. In future, if the selection of drugs used to test the feasibility of electronic systems were narrowed to TS switches, potential limitations of such systems could be assessed more thoroughly. References and/or acknowledgements No conflict of interest.