Authors: (including presenting author): :
Chan CH(1), Wong Y(1), Cheung CT(1)
Affiliation: :
(1) Physiotherapy department, Alice Ho Miu Ling Nethersole Hospital
Keyword 1: :
Artificial Intelligence (AI)
Keyword 3: :
Staff Rostering System (SRS)
Keyword 4: :
Manual data entry
Introduction: :
•The Staff Rostering System (SRS) was implemented across the Allied Health Department in June 2020. Since then, daily duties for all professional staff have required manual entry into SRS. •In the initial phase of SRS data entry, duties were typed slowly - character by character using keyboards. Since SRS was not originally designed for Allied Health duties, frequent technical issues occurred (e.g., unintended character deletions), often forcing staff to re-enter data again. Inputting one week of duties could take more than an hour. •Additionally, duties arrangement is complex in our setting as covering Inpatient, Outpatient and Community Service. In our department, there are 4 sessions in a whole day duty, each staff’s duty can vary in all these 4 sessions. Moreover, each’s duty can vary every day in a week. With such variability, the defaulted functions (e.g. Assign Team Pattern) in SRS can hardly be applied in our department. Therefore, the complicated duties arrangement and SRS input tasks were frequently assigned to professional staff to minimize errors. •To streamline duty planning and facilitate efficient SRS entry, the AHNH Physiotherapy Department has adopted Google Sheets (similar to Microsoft Excel) to plan clinical duty since 2022. Google Sheets can automatically generate text formatted for SRS entry. Staff only needs to copy and paste this text into SRS. •However, assuming 60 minutes are required to input SRS weekly, staff would have spent 52 clinical hours annually on repetitive copy-paste tasks with little value, instead of dedicating time on patients. •Therefore, we would like to see if AI could help in streamlining this mundane data entry process.
Objectives: :
•To streamline and optimize SRS data input: - Reduce time spent on data entry - Minimize human errors - Improve OSH issue by reducing staffs’ eye strain and mental fatigue - Free up time for patient care
Methodology: :
•Our team utilized AI tools (DeepSeek) for self-taught programming. •With AI’s assistance, our team learned programming from scratch and developed an automation program using built-in Windows tools - PowerShell. •This program automatically transfers duties from Google Sheets into SRS. •This program has been implemented since April 2025.
Result & Outcome: :
•After implementing this program, staff only require 2-3 minutes manual preparation, the subsequent SRS data entry time will then be changed from 60 minutes fully manual task to 7-9 minutes fully program automation, with zero manual effort but purely computer processing. •90% reduction (from 64-66 minutes to 4-6 minutes) in manual data entry time. •If this program saves 60 minutes of staff time weekly, 52 staff hours can be saved annually (based on 52 operational weeks). •Time can be reallocated for patient care and clinical works •As an AI-automated program, the program ensures 100% SRS data entry accuracy compared to manual entry under standardized conditions. •Significantly reduced potential OSH issues such as complications of prolonged computer work, eye strain and mental fatigue. Remark: • Since the update of SRS interface in 11/2025, the SRS data entry method has changed. Accordingly, we have also updated and modified our program with the assist of AI. The data entry process will remain 100% automated. Therefore, the total amount of time for manual input will still be greatly reduced by 90%.
