Abstract
OBJECTIVE: The aim of this research is to implement and reliably evaluate primary human papillomavirus (HPV) screening in an established and routinely running organised, large-scale population-based screening programme. PARTICIPANTS: Resident women in the Stockholm/Gotland region of Sweden, aged 56-60 years were randomised to either (1) screening with cervical cytology, with HPV test in triage of low-grade cytological abnormalities (old policy) or (2) screening with HPV testing, with cytology in triage of HPV positives (new policy). OUTCOME: The primary evaluation was the detection rate of cervical intraepithelial neoplasia grade 2 or worse (CIN2+). RESULTS: During January 2012-May 2014, the organised screening programme sent 42 752 blinded invitations with a prebooked appointment time to the women in the target age group. 7325 women attended in the HPV policy arm and 7438 women attended in the cytology arm. In the new policy, the population HPV prevalence was 5.5%, using an accredited HPV test (Cobas 4800). HPV16 prevalence was 1.0% (73/7325) and HPV18 prevalence was 0.3% (22/7325). In the HPV policy arm, 78/405 (19%) HPV-positive women were also cytology positive. There were 19 cases of CIN2+ in histopathology, all among women who were both HPV positive and cytology positive. The positive predictive value for CIN2+ in this group was 33.3% (19/57). In the cytology policy, 153 women were cytology positive and there were 18 cases of CIN2+ in histopathology. Both the total number of cervical biopsies and the number of cervical biopsies with benign histopathology were much lower in thepositive predictive value policy (49 benign, 87 total vs 105 benign, 132 total). CONCLUSION: Primary HPV screening had a similar detection rate for CIN2+ as cytology-based screening, already before follow-up of HPV-positive, cytology-negative women with new HPV test and referral of women with persistence. TRIAL REGISTRATION NUMBER: NCT01511328.