Evaluation of automation on the Gold Coast University Hospital Pharmacy department

Loading...
Thumbnail Image
File version
Author(s)
Primary Supervisor

Grant, Gary D

Other Supervisors

Kelly, Fiona S

Editor(s)
Date
2021-11-12
Size
File type(s)
Location
License
Abstract

Background: Pharmacy robotics have been implemented globally to create medication management efficiencies. However, translation to the Australian public hospital environment has not been evaluated. It is also not known what factors influence pharmacy staff to use new technologies that could guide future automation implementation strategies. Aim: The primary aim of this research is to evaluate the introduction of a Pharmacy Robotic Dispensing System (PRDS) in an Australian public hospital pharmacy and the impact it had on pharmacy staff medication supply activities in both the dispensary and imprest management. Secondly, this research aims to identify factors that influence hospital pharmacy staff acceptance of robotics during implementation and over time. Method: This study was conducted in a tertiary public hospital on the Gold Coast, Australia and took place in two phases, each phase consisting of a staff survey and a time in motion study. Phase one: May 2016; an initial assessment prior to the PRDS becoming operational for baseline data. Data for the staff survey and time and motion study were collected and analysed in this phase. Phase two: August 2017; PRDS had been operational for 15 months and the study was repeated. Data for the staff survey and time in motion study were remeasured in this phase for comparative analysis Staff survey: A single centred, prospective, longitudinal cohort quantitative study. Pharmacy staff were surveyed using the Extended Technology Acceptance Model (ETAM). Fischers exact test and correlation analysis of paired responses was used to identify significant factors influencing usage of the PRDS between the two time points. Time in motion study: A single centred, prospective, longitudinal time in motion study. Time stamped video footage was collected tracking medication orders in numbered containers throughout the pharmacy dispensary. Activities observed and recorded comprised of dispensing, labelling, assembly of items, number of items and time waiting to be actioned. Data was analysed and compared overall turnaround times; time waiting to be actioned; dispensing, assembly and labelling times; plus dispensing rates (items/person/hours). Fitbit Zip® used to anonymously track pharmacy assistant steps per day in the dispensary. Dispensing software (iPharmacy ®) supplied the location of stocked medication and electronic tracking databases provided data for imprest medication supply turnaround time, picking and waiting for action times. Median turnaround times and Mann Whitney U test used to examine relationships between phases, plus mean steps per day and two tailed t-test used to analyse changes in movement. Results: Sixty four respondents completed surveys during phase one (70%: 64/91) and 34 paired surveys were collected fifteen months later in phase two (53%: 34/64). Respondents were predominantly young, female with a tertiary qualification. Initial perceptions of the pharmacy robot did not change over time, with the exception of ‘reliability’. Participants found the robot less reliable than expected after working with it for fifteen months. Departmental leaders had greatest influence on technology acceptance during implementation and over time. Other key factors correlating with pharmacy staff acceptance included: how useful the robot was perceived to be; ease of use; and how relevant the robot was for an individual role. Higher levels of education had a negative association with usage during implementation and age was not a factor. Medication stored in the robot was limited to 46% (n=20,771 full packs) of total pharmacy holdings (n= 45,437 full packs). At baseline, 774 orders were received in the dispensary over five days increasing by 13% to 887 in phase two (p<0.01). Dispensary workload increased, staff levels remained constant and pharmacy assistant movement significantly reduced (p<0.05). However, there were no significant changes to dispensing rates and turnaround times. Conclusion: This study identified critical insights influencing staff acceptance of pharmacy robots that will help inform future implementation. The influence of pharmacy leaders emerged as key influence on technology acceptance. Leveraging on this influence a communication strategy prior to implementation should include information on useful functions and known benefits of the system customised for individual roles. In addition, pharmacy robotics has the potential to absorb increased workload and reduce staff movement in the dispensary when staffing levels remained constant. However, turnaround times alone are too simplistic as a sole measure of benefits for robotics in Australian public hospital pharmacy.

Journal Title
Conference Title
Book Title
Edition
Volume
Issue
Thesis Type

Thesis (Masters)

Degree Program

Master of Medical Research (MMedRes)

School

School of Pharmacy & Med Sci

Publisher link
Patent number
Funder(s)
Grant identifier(s)
Rights Statement
Rights Statement

The author owns the copyright in this thesis, unless stated otherwise.

Item Access Status
Note
Access the data
Related item(s)
Subject

Pharmacy Robotic Dispensing System

Australian public hospital pharmacy

dispensary and imprest management

hospital pharmacy staff

Gold Coast

Australia

Persistent link to this record
Citation