Harms and benefits of cervical cancer screening among non-attenders in Switzerland: The transition towards HPV-based screening.

ABSTRACT

Human papillomavirus (HPV) testing is replacing cytological screening for cervical cancer. Our aim was to assess the expected benefits and harms of different cervical screening strategies. This study is sub-analysis of a previous cost-effectiveness study with a target population of unscreened women without cervical cancer aged ≥ 25 years. A recursive decision-tree with one-year cycles was used to model the life-long natural HPV history. Markov cohort simulations were used to assess the expected outcomes from the model. The outcomes of three strategies were compared with the absence of screening HPV-testing on self-collected vaginal samples (Self-HPV) followed by colposcopy (Self-HPV/colpo), Self-HPV and triage with cytology (Self-HPV/PAP), cytology and triage with HPV (PAP/HPV). All screening strategies resulted in reductions in cancer cases and deaths. Self-HPV strategies were associated with a lower cancer incidence and mortality life-long, not only when performed every 3 years but also when Self-HPV was performed every 5 years vs cytology every 3 years. The gain in life expectancy obtained was 82 days with Self-HPV/colpo, 81 days with Self-HPV/PAP and 75 days with PAP/HPV compared to no screening. The number of lifetime total visits was greater with PAP/HPV compared with the Self-HPV strategies (13.13 vs < 3). The number of conizations remained relatively stable with the change of screening frequency and strategy. Self-HPV may represent a reasonable balance of harms and benefits when performed every 5 years compared to cytology every 3 years. Self-HPV/PAP yielded the most efficient harm to benefit ratio when using colposcopy as a proxy for harms.