Improving laboratory workforce efficiency using AI-assisted digital cytology within an HPV-based cervical screening programme: A model-based evaluation for the NHS Cervical Screening Programmes.

OBJECTIVES

We aimed to estimate the time required to review and report cervical cytology slides within a Human Papillomavirus (HPV)-primary cervical cancer screening programme, where cytology is used for triage following a positive HPV test, comparing Artificial Intelligence (AI)-assisted digital cytology (Genius) with manual microscopy.

DESIGN

A decision-tree model of the flow of slides through the National Health Service (NHS) England cervical screening cytology laboratory workflow, parameterised using published evidence and expert input.

SETTING

Cytology laboratories within the NHS Cervical Screening Programme in England and Scotland.

RESULTS

Screening and reporting 479,125 cytology slides annually in England was estimated to require 31,842 staff hours with AI-assisted digital cytology versus 103,151 hours with manual microscopy. The mean time per slide review and reporting was 4.0 vs 12.9 minutes, respectively, corresponding to a potential 69% increase in overall productivity.

CONCLUSIONS

AI-assisted digital cytology can support more resilient cervical screening programmes in the UK and other countries by reducing staff time required for slide review and reporting. These findings may inform policymakers considering strategies to mitigate workforce shortages.