New HPV Model for Cervical Screening – Center for Health Decision Science

CHDS faculty Nicole Gastineau Campos and Jane Kim, with colleagues at the National Cancer Institute, developed a new simulation model of how human papillomavirus (HPV) infection can progress toward cervical cancer, intended for evaluating emerging screening approaches. The article reporting the model was published in the Journal of the National Cancer Institute.

Instead of tuning model inputs until outputs fit observed data (i.e., model calibration), the team directly estimated the transition rates from infection to precancer from cohort, trial, and cross-sectional studies, so that the uncertainty in those estimates could be measured. Precancer was defined strictly as an expert-confirmed high-grade lesion (CIN3 or worse) with a carcinogenic HPV type, rather than the multiple lesion grades CIN1 and CIN2 that are no longer considered to be on the causal pathway to cervical cancer used in many other models. The model was built for Southeastern Brazil, and the authors reported that its projections aligned well with age-specific HPV prevalence in São Paulo across the 13 carcinogenic types it included.

The authors noted that transitions from precancer to cancer could not be directly observed and remained highly uncertain. They added that the model is being applied to inform Brazil’s pending decisions on national adoption of HPV-based screening. Future work will evaluate the cost-effectiveness of strategies for resource-limited settings, including a screen-triage-treat approach with HPV testing of self-collected samples and AI-assisted visual evaluation.

Learn more: Read the full publication, A New Microsimulation Model of HPV Natural History for Comparison of Novel Cervical Screening Strategies