Cost-effectiveness of cervical cancer screening: cytology versus human papillomavirus DNA testing

Objective To determine the most cost-effective screening programme for cervical cancer. Design Cost-effectiveness analysis from a societal perspective. Setting The Netherlands. Population Dutch women who have not been invited for human papillomavirus (HPV) vaccination. Methods We calibrated the microsimulation screening analysis (MISCAN) model to Dutch epidemiological data. We used this model to consider nine screening strategies that use: (i) cytological testing with cytology triage for borderline/ mildly abnormal smears; (ii) HPV testing with cytology triage for HPV-positive smears; or (iii) cytological testing with HPV triage for borderline/ mildly abnormal smears. For each strategy, we varied the number of screening rounds, the time interval, the age of the first screening, and the type of cytological testing (conventional or liquid-based cytology). Main outcome measures Quality-adjusted life years (QALYs) gained and costs from a societal perspective. Results Under the base-case assumptions, primary HPV testing with cytology triage is the most cost-effective strategy. Using costeffectiveness thresholds of (sic)20 000 and (sic)50 000 per QALY gained yields optimal screening programmes with three and seven screening rounds, respectively. The results are sensitive to several uncertain model inputs, most importantly the costs of the HPV test. For women aged 32 years or younger, primary cytology screening is more cost-effective than primary HPV testing. Conclusions Increasing the interval between screening rounds and changing the primary test from cytology to HPV testing can improve the effectiveness and decrease the costs of cervical cancer screening in the Netherlands.