What the study found
Robotic process automation, or RPA, reduced the time needed to abstract cancer registry data in a production electronic health record, or EHR, environment. The size of the time savings differed by registry, with larger gains for gastric cancer than for breast cancer.
Why the authors say this matters
The authors conclude that RPA can improve clinical data workflows when it is matched with organizational readiness and strong monitoring. They also suggest it may help reduce the manual burden of repetitive data abstraction tasks.
What the researchers tested
The researchers implemented RPA for gastric and breast cancer registries within a live EHR system at a tertiary hospital. They compared per-patient extraction time before and after implementation, manually verified all RPA outputs against source records, and interviewed 14 stakeholders using semi-structured interviews analyzed with the PARiHS framework.
What worked and what didn't
RPA was applied to 70 gastric cancer variables and 83 breast cancer variables. Mean abstraction time per patient fell from 19.5 to 5.1 minutes for gastric cancer, a 74% decrease, and from 25.4 to 17.8 minutes for breast cancer, a 30% decrease. Based on 2024 surgical volumes, the study estimated more than 260 hours of manual labor could be saved per year.
What to keep in mind
The abstract says the quantitative findings are based on time savings, but it also notes that formal quantitative assessments of data accuracy were not performed. The study was done in one tertiary hospital and focused on gastric and breast cancer registries, so the scope was limited to those settings and workflows.
- RPA lowered cancer registry abstraction time in a live EHR setting.
- Time savings were larger for gastric cancer than for breast cancer.
- All RPA-extracted outputs were manually checked against source records.
- Interviewed participants said RPA was well suited to repetitive tasks.
- Successful implementation depended on clinician cooperation and continuous output monitoring.