Impact of the transition to HPV-based primary screening in Portugal's organized cervical cancer screening program: A controlled interrupted time-series analysis (2014-2023)

Objectives: To evaluate the impact of transitioning from cytology to primary human papillomavirus (HPV) testing on cervical cancer screening (CCS) performance in Portugal's Central Region. Study design: Retrospective, population-based evaluation using a controlled interrupted time-series (CITS) approach. Methods: CCS registry data (2014-2023) were analyzed in six-month intervals. Three performance indicators (participation, test positivity, and priority referrals) were modeled using negative binomial regression with appropriate offsets. A COVID-19 covariable (national lockdown, March-June 2020) adjusted for the temporary suspension of screening services. The organized breast cancer screening (BCS) program served as an external control to distinguish CCS-specific effects from system-wide temporal fluctuations. Results: The analysis included 594,074 CCS and 888,184 BCS screening episodes. Following HPV implementation, CCS evolved differently from the control group across all outcomes. Priority referrals showed the strongest effect, with a four-fold immediate increase in CCS not observed in BCS (IRR: 4.31; 95% CI: 3.89-4.79). Participation and test positivity also diverged between programs, although with smaller magnitude changes. Post-intervention trends differed across all outcomes, although the COVID-19 pandemic, occurring shortly after implementation, complicates the interpretation of temporal patterns. Conclusions: Transition to HPV-based screening was associated with changes in screening processes, including increased identification of high-risk cases requiring priority referral while maintaining participation. By incorporating an external control, the CITS approach strengthens attribution of observed effects to HPV implementation rather than background system dynamics. These findings support HPV-based screening within organized programs and highlight its role in improving risk stratification and program performance.