The impact of new technologies in cervical cancer screening: Results of the recruitment phase of a large randomised controlled trial from a public health perspective Paolo Giorgi-Rossi 1 * , Nereo Segnan 2 , Marco Zappa 3 , Carlo Naldoni 4 , Manuel Zorzi 5 , Massimo Confortini 3 , Monica Merito 1,6 , Jack Cuzick 7 , Guglielmo Ronco 2 and the NTCC Working Group 1 Agency for Public Health, Lazio Region, Rome, Italy 2 Unit of Cancer Epidemiology, Centro per la Prevenzione Oncologica (CPO), Turin, Italy 3 Centro per lo Studio e la Prevenzione Oncologica, Florence, Italy 4 Centro di riferimento screening, Assessorato alla Sanita`, Regione Emilia-Romagna 5 Istituto Oncologico Veneto, Padova, Italy 6 Laboratory of Economics and Management, Sant’Anna School of Advanced Studies, Pisa, Italy 7 Queen Mary’s School of Medicine and Dentistry and Cancer Research UK, London, United Kingdom The decision to introduce liquid-based cytology (LBC) and HPV as screening tests involves criteria based on resource consumption. We used cross-sectional data at recruitment from the NTCC trial [ISRCTN81678807] on 28,000 women aged 35–60, randomised to receive a conventional Pap test or LBC plus HPV. We computed the resources employed to detect a CIN21 with different screen- ing strategies. In order to result in the same overall cost per CIN21 detected as screening by conventional cytology, the unit cost of LBC used alone should be less than that of a conventional Pap while its unit cost may be up to 20% higher if HPV-triage for Atypical Squamous Cells of Undetermined Significance is applied together. With the same criterion the unit cost of HPV used alone may be about 20% higher than that of a Pap-test using a 1 pg/ml cut-off and over 40% higher using a 10 pg/ml cut-off. If HPV test- ing is applied with cytology-triage, a single HPV test may cost 20– 30% more than a conventional Pap to result in the same overall cost per CIN21 detected. ' 2007 Wiley-Liss, Inc. Key words: human papilloma virus; cervical cancer; cancer screening; health services workload; costs In recent years new technologies potentially applicable for cer- vical screening, particularly liquid-based cytology (LBC) and mo- lecular testing for Human Papilloma Virus (HPV), have become available. HPV infection is causally related to cervical cancer and is present in virtually all invasive cancers 1 ; data suggest that the infection precedes cytologically detectable intraepithelial lesions by a median of 12 years, 2,3 therefore a negative HPV test may pro- vide substantial protection for a very long period. 4 Although the accuracy of LBC compared to conventional cytology has been questioned, 5 LBC is now widely used in many countries. On the other hand, conventional cytology has worked very effectively for 50 years, with an 80% potential reduction in cervical cancer mor- tality, in the best hands. 1 It is clear that any decision to change a screening test that works that well will involve an analysis of the human and technological resources available and cost considera- tions, 6–13 since there are only residual health benefits to be gained. To date, LBC has been accepted for screening by the English and the Scottish Health Authorities on the basis of cost-effective- ness and organisational considerations. 14–16 HPV testing has only been recommended for triaging ASCUS cytology. 1,17–19 Many authors have proposed HPV as a primary screening test, 20–33 but to date no governmental agency or organised screening pro- gramme recommends its use as such. We conducted a randomised controlled trial, the NTCC (New Technologies for Cervical Cancer) study, designed to provide lon- gitudinal data on the detection of high-grade Cervical Intraepithe- lial Neoplasia (CIN) by conventional cytology, which was applied to the conventional arm of the study, and by new technologies, which were applied to the experimental arm. The latter involved 2 phases, a first with HPV testing and LBC combined and a second phase with HPV testing only. Final cost-effectiveness analyses will be based on these longitudinal data. Data on the cross-sec- tional accuracy of HPV and LBC vs. conventional cytology derived from Phase 1 have already been published. 34,35 In this article, also based on Phase 1, we compare the referral rates and work load of 4 different strategies for a single round of screening using the new technologies available in cervical cancer prevention in women over 35 years of age. We also estimate the unit cost per new test at which the new techniques results in an overall cost per high-grade lesion detected equal to that by con- ventional Pap test screening. Since costs can vary widely in differ- ent settings this approach allows locally relevant values to be con- sidered. 36 In addition, it allows the local evaluation of the thresh- old at which the new strategy becomes competitive with conventional cytology. We considered only women aged 35 years or more because the younger group followed a different protocol. Methods The data presented here are part of the findings of a large rando- mised controlled trial with 2 arms. The project was conducted with 9 organised cervical screening programs in 7 Italian regions. We excluded from the present anal- ysis 2 centres that, unlike the others, did not directly refer women to colposcopy after Atypical Squamous Cells of Undetermined Significance (ASCUS) conventional cytology. All programmes routinely invite women aged 25–64 years old every 3 years. In this analysis we consider women aged 35–60 years from 7 centres. The methods have been described in detail elsewhere, 34 here we briefly summarise the trial design and setting. General trial methods Women aged 25–60 who attended the cervical cancer screening programme for a new round were asked to enter the trial. Women who were pregnant, had had a hysterectomy, who had never had sexual intercourse or were recently treated for CIN (in the last 5 years) were not eligible. After giving informed consent, women were individually rando- mised to the 2 groups, with a 1:1 ratio. In the conventional arm women with ASCUS or more severe cytology were referred for colposcopy, as in routine screening. *Correspondence to: Agenzia di Sanita ` Pubblica, Regione Lazio, via di S. Costanza 53, Roma 00198. Fax 1390683060463. E-mail: giorgirossi@asplazio.it Received 10 October 2006; Accepted after revision 11 June 2007 DOI 10.1002/ijc.23055 Published online 27 August 2007 in Wiley InterScience (www.interscience. wiley.com). Int. J. Cancer: 121, 2729–2734 (2007) ' 2007 Wiley-Liss, Inc. Publication of the International Union Against Cancer