Computerized pharmaceutical intervention to reduce reconciliation errors at hospital discharge in Spain: an interrupted time‐series study

Summary What is known and objective It is well known that medication reconciliation at discharge is a key strategy to ensure proper drug prescription and the effectiveness and safety of any treatment. Different types of interventions to reduce reconciliation errors at discharge have been tested, many of which are based on the use of electronic tools as they are useful to optimize the medication reconciliation process. However, not all countries are progressing at the same speed in this task and not all tools are equally effective. So it is important to collate updated country‐specific data in order to identify possible strategies for improvement in each particular region. Our aim therefore was to analyse the effectiveness of a computerized pharmaceutical intervention to reduce reconciliation errors at discharge in Spain. Methods A quasi‐experimental interrupted time‐series study was carried out in the cardio‐pneumology unit of a general hospital from February to April 2013. The study consisted of three phases: pre‐intervention, intervention and post‐intervention, each involving 23 days of observations. At the intervention period, a pharmacist was included in the medical team and entered the patient's pre‐admission medication in a computerized tool integrated into the electronic clinical history of the patient. The effectiveness was evaluated by the differences between the mean percentages of reconciliation errors in each period using a Mann–Whitney U test accompanied by Bonferroni correction, eliminating autocorrelation of the data by first using an ARIMA analysis. In addition, the types of error identified and their potential seriousness were analysed. Results and discussion A total of 321 patients (119, 105 and 97 in each phase, respectively) were included in the study. For the 3966 medicaments recorded, 1087 reconciliation errors were identified in 77·9% of the patients. The mean percentage of reconciliation errors per patient in the first period of the study was 42·18%, falling to 19·82% during the intervention period ( P = 0·000). When the intervention was withdrawn, the mean percentage of reconciliation errors increased again to 27·72% ( P = 0·008). The difference between the percentages of pre‐ and post‐intervention periods was statistically significant ( P = 0·000). Most reconciliation errors were due to omission (46·7%) or incomplete prescription (43·8%), and 35·3% of which could have caused harm to the patient. What is new and conclusion A computerized pharmaceutical intervention is shown to reduce reconciliation errors in the context of a high incidence of such errors.