OBJECTIVES: Improving patient safety by reducing medication errors during prescribing at care transitions (for example, hospital admission or discharge) is a global policy priority. Standard medicines reconciliation is typically used to identify medication errors at care transitions. These procedures are imperfect and can miss potentially harmful medication errors. Decision-makers are now seeking evidence demonstrating the added-value of intervention strategies to reduce undetected medication errors at care transitions. To achieve this, analysts must first quantify the current prevalence of medication errors missed by standard medicines reconciliation. This quantity is challenging to estimate because undetected medication errors are unobservable. This study aims to demonstrate a framework enabling analysts to estimate the prevalence of undetected medication errors.

METHODS: The framework defines total medication errors as the sum of medication errors detected and undetected by standard medicines reconciliation. Two data sources are required: (1) observational data on the performance of conventional medicines reconciliation to estimate the number of errors detected; (2) the relative risk reduction in medication errors detected by medicines reconciliation versus no medicines reconciliation. First, estimate the total medication errors: medication errors detected by medicines reconciliation divided by one minus the relative risk reduction. Then, estimate the total undetected medication errors: subtract the total detected medication errors (observed) from the total medication errors (estimated). A probabilistic UK-based case study illustrates the framework.

RESULTS: Observational data report that 5,910/44,496 (13.3%) medication orders at admission had a medication error detected during medicines reconciliation (Ashcroft et al., 2015). A Cochrane meta-analysis estimated a relative risk reduction in medication errors of 0.13 following medicines reconciliation. Therefore, the estimated total medication errors is 15.3% of prescribed items. The total undetected medication errors is 1.98% of prescribed items.

CONCLUSIONS: A robust understanding of undetected medication errors will help demonstrate the added-value of intervention strategies to improve patient safety at care transitions.