A single-centre before-after study: implementing and evaluating drug life cycle management in pharmacy IV admixture service

Abstract Managing drug expiry and the usable period of finished infusions is a key component of medication safety in a pharmacy intravenous admixture service (PIVAS). We developed a pharmacy service model based on drug life cycle management (DLCM) and evaluated its preliminary effects on inventory and workflow indicators in a single-centre PIVAS. This single-centre, uncontrolled before-and-after quality-improvement study was conducted in the PIVAS of a tertiary general hospital in Chongqing, China. January 2024 served as the baseline month and February–May 2024 as the post-implementation period. The DLCM model comprised three phases: drug quality management based on a three-tiered information system integrating the hospital information system (HIS), an intelligent secondary storehouse, and a central scheduling machine; rational drug-use management including pharmacist prescription review, individualised infusion-batch decisions, and drug information management; and drug therapy management including finished-infusion beyond-use date (BUD) guidance, QR-code infusion prompts, real-time administration monitoring, and one-click medicine return. Outcomes included the near-expiry inventory rate, the proportion of near-expiry drug varieties, the configuration-to-administration interval (from configuration completion to infusion start), and clinical department satisfaction; each post-implementation month was compared with the January 2024 baseline. Compared with January 2024, the near-expiry inventory rate (by value) decreased from 4.91 to 2.05% in May 2024 (relative reduction 58.2%; Spearman ρ = − 1.0, P = 0.017), and the proportion of near-expiry varieties decreased from 6.04 to 3.04% (relative reduction 49.7%; Cochran–Armitage trend test, P = 0.012). The mean configuration-to-administration interval decreased from 206.8 min in January to 149.5 min in May 2024 (Welch’s t-test, P < 0.0001; Cohen’s d = 0.36), with significant reductions in all seven infusion batches. Overall clinical department satisfaction increased from 2.40 ± 0.74 to 4.33 ± 0.62 on a 5-point scale (paired t-test, P < 0.0001). Implementation of the DLCM-based PIVAS model was associated with improvements in selected inventory and workflow indicators and increased staff satisfaction during the four-month post-implementation period. Because of the single-centre uncontrolled design, the one-month baseline, and the absence of patient-level safety outcomes, these findings are preliminary and no causal conclusions can be drawn. Multicentre studies with longer follow-up and direct medication-safety measures are warranted.