ABSTRACT
BACKGROUND: The process of obtaining a complete medication history for patients admitted to the hospital from the ED at hospital admission, without discrepancies, is error prone and time consuming. OBJECTIVES: The goal of this study was the development of a clinical decision rule (CDR) with a high positive predictive value in detecting ED patients admitted to hospital at risk of at least one discrepancy during regular medication history acquisition, along with favourable feasibility considering time and budget constraints. METHODS: Data were based on a previous prospective study conducted at the ED in Belgium, describing discrepancies in 3592 medication histories. Data were split into a training and a validation set. A model predicting the number of discrepancies was derived from the training set with negative binomial regression and was validated on the validation set. The performance of the model was assessed. Several CDRs were constructed and evaluated on positive predictive value and alert rate. RESULTS: The following variables were retained in the prediction model: (1) age, (2) gender, (3) medical discipline for which the patient was admitted, (4) degree of physician training, (5) season of admission, (6) type of care before admission, number of (7) drugs, (8) high-risk drugs, (9) drugs acting on alimentary tract and metabolism, (10) antithrombotics, antihaemorrhagics and antianaemic preparations, (11) cardiovascular drugs, (12) drugs acting on musculoskeletal system and (13) drugs acting on the nervous system; all recorded by the ED physician on admission. The final CDR resulted in an alert rate of 29% with a positive predictive value of 74%. CONCLUSION: The final CDR allows identification of the majority of patients with a potential discrepancy within a feasible workload for the pharmacy staff. Our CDR is a first step towards a rule that could be incorporated into electronic medical records or a scoring system.
