Evaluating Automated Cancer Registration Accuracy
Author Information
Author(s): Giovanna Tagliabue, Anna Maghini, Sabrina Fabiano, Andrea Tittarelli, Emanuela Frassoldi, Enrica Costa, Silvia Nobile, Tiziana Codazzi, Paolo Crosignani, Roberto Tessandori, Paolo Contiero
Primary Institution: Istituto Nazionale per lo Studio e la Cura dei Tumori
Hypothesis
How accurate are automated cancer registration codes compared to manual registration?
Conclusion
The automated procedure is efficient for routine case generation, effectively reducing the workload for manual checking.
Supporting Evidence
- 59% of automatically generated cases were accepted without major discrepancies.
- Discrepancies were mostly trivial, such as minor errors in names and dates.
- The automated system effectively reduced the workload for manual checking.
Takeaway
This study looked at how well computers can register cancer cases compared to people doing it by hand. It found that computers did a pretty good job!
Methodology
The study compared automatically generated cancer cases with those produced manually over one year, using clinical records and pathology reports.
Potential Biases
Potential bias due to variations in coding practices between automated and manual methods.
Limitations
The study may not account for all types of discrepancies in coding practices.
Participant Demographics
Cancer cases from the Lombardy Cancer Registry in 1997.
Digital Object Identifier (DOI)
Want to read the original?
Access the complete publication on the publisher's website