Invasive Cervical Cancer Detection After Transition to Primary Human Papillomavirus Screening: Population-Based Linkage Study in Portugal.
Invasive cervical cancer (ICC) is a key outcome for evaluating cervical cancer screening (CCS) programs. Monitoring ICC trends and distinguishing cancers detected within and outside the organized screening program are essential for assessing program impact, particularly during transitions such as the introduction of human papillomavirus (HPV) based primary screening.
We conducted a retrospective population-based study in Portugal's Central Region (2014 - 2023). Deterministic linkage was performed between the organized CCS database and hospital morbidity records. Invasive cervical cancer cases were classified as screen-detected or non-screen-detected based on prior screening history. Incidence rates were calculated using population denominators. Poisson regression models with log link and log-transformed population offsets were fitted to estimate incidence rate ratios (IRR) comparing cytology-based (2014 - 2019) and HPV-based (2020 - 2023) periods, adjusting for temporal trends.
A total of 654 ICC cases were included. Screening participation remained broadly stable over time. The HPV-based period was associated with a significantly higher rate of screen-detected ICC (IRR = 1.94; 95% CI 1.11 - 3.40) compared with the cytology period. In contrast, ICC incidence outside the organized screening program did not differ significantly between periods (IRR = 0.94; 95% CI 0.65 - 1.35).
The transition to HPV-based primary screening was associated with a shift in ICC detection towards the organized screening program, without evidence of a corresponding increase in incidence among women diagnosed outside the organized screening program. These findings should, however, be interpreted with caution in light of the early phase of implementation and potential external factors influencing access to diagnosis. Overall, the results support a shift in detection within the screening pathway rather than changes in underlying disease occurrence. Strengthening integration between screening, hospital, and cancer registry data remains essential for robust evaluation of cervical cancer screening policies.
We conducted a retrospective population-based study in Portugal's Central Region (2014 - 2023). Deterministic linkage was performed between the organized CCS database and hospital morbidity records. Invasive cervical cancer cases were classified as screen-detected or non-screen-detected based on prior screening history. Incidence rates were calculated using population denominators. Poisson regression models with log link and log-transformed population offsets were fitted to estimate incidence rate ratios (IRR) comparing cytology-based (2014 - 2019) and HPV-based (2020 - 2023) periods, adjusting for temporal trends.
A total of 654 ICC cases were included. Screening participation remained broadly stable over time. The HPV-based period was associated with a significantly higher rate of screen-detected ICC (IRR = 1.94; 95% CI 1.11 - 3.40) compared with the cytology period. In contrast, ICC incidence outside the organized screening program did not differ significantly between periods (IRR = 0.94; 95% CI 0.65 - 1.35).
The transition to HPV-based primary screening was associated with a shift in ICC detection towards the organized screening program, without evidence of a corresponding increase in incidence among women diagnosed outside the organized screening program. These findings should, however, be interpreted with caution in light of the early phase of implementation and potential external factors influencing access to diagnosis. Overall, the results support a shift in detection within the screening pathway rather than changes in underlying disease occurrence. Strengthening integration between screening, hospital, and cancer registry data remains essential for robust evaluation of cervical cancer screening policies.
Authors
Sousa Sousa, Fonseca-Moutinho Fonseca-Moutinho, Gomes Gomes, Loureiro Loureiro, Nascimento Nascimento, Solheiro Solheiro, Saavedra Saavedra, Ponte Ponte, Barros Barros, Pereira Pereira, Rebelo Rebelo, Caeiro Caeiro, Goes Goes, Soares Soares
View on Pubmed