Assessing the Effectiveness of Interventions Implemented by Nurses to Reduce Medication Administration Errors in Hospitalised Acute Adult Patient Settings: Systematic Review and Meta-Analysis.

The pooled analysis combined randomised and non-randomised studies, and all included studies had at least one risk of bias element. Most studies used direct observation without blinding to outcomes, and many were before-and-after or quasi-experimental designs. Heterogeneity was very high (I2 = 94%), so the average effect may not apply to any one ward. The funnel plot was asymmetric, indicating possible publication bias. Studies used different definitions and reporting standards for medication administration errors, making comparison difficult. Only English-language studies were included. The paper reports inconsistent participant counts for the meta-analysis (2034 in one place and 35,046 in another). Several interventions were multicomponent, so the effect of a single part cannot be isolated.

The authors declare no conflicts of interest. The first author conducted this review through a NIHR/NHSE East Midlands Clinical Academic Internship.

Do not conclude that any nurse-led intervention reliably reduces medication administration errors in adult acute wards. The overall pooled reduction was based on highly heterogeneous studies, had possible publication bias, and only the workflow smart technologies subgroup reached statistical significance.