Authors (including presenting author) :
Chow CH (1), Fung CY (1), Lai KS (1), Lau CH (1), Chan WYM (1), Mak CK (1), MAK LKK (1)
Affiliation :
(1) Department of Pharmacy, Tseung Kwan O Hospital
Introduction :
Radio Frequency Identification (RFID) has been widely utilized in healthcare systems to enhance efficiency and safety. Recently it has been adopted into the new “Assemble-to-Light” Pharmacy Express Dispensing System in Tseung Kwan O Hospital as a pioneering project.
Objectives :
To assess the effect of incorporating the “Assemble-to-Light” system into Hong Kong public hospital settings in promoting medication assembly accuracy and enhancing workflow efficiency by evenly distributing workload during peak hours.
Methodology :
A single-centred, observational, pre-, and post-implementation study was conducted in Main Pharmacy, Tseung Kwan O Hospital. Workload statistics in August 2022 and August 2023 were drawn from the Pharmacy Management System to evaluate average dispensing time difference and variance. Workload distribution were drawn from May to August in 2022 and August to November in 2023 for comparison. Medication assembly error were recorded in 5-week period in pre-implementation phase and post-implementation phase.
Result & Outcome :
The mean dispensing time in the pre-implementation period is 300.87s and that in the post-implementation period is 311.10s, with a statistically significant mean difference of an increase of 10.228s (p=0.024, unpaired t-test) in the post-implementation period. The overall standard deviation of the mean of drug picking time is 215.372 seconds in the pre-implementation period, and 202.705 seconds in the post-implementation period. Levene’s test for Equality of Variances was conducted with an F value of 11.968 (p<0.001) suggesting a difference in variance of dispensing time between pre-implementation phase and post-implementation phase. A significant correlation between the standard deviation of picking time and mean picking time was found by simple linear regression (p=0.025). Medication assembly error rate was reduced from 0.199 errors per 100 items assembled to 0.142 errors per 100 items assembled (p=0.004, 2 proportions z-test) after implementation. The “Assemble-to-Light” system was able to distribute dispensing workload more evenly and reduce the rate of assembly error despite the increase in workload anticipated in public hospital pharmacy. This project also provided the proof-of-concept on the relation between workload distribution and picking time.
