Journal of Medical Economics ISSN: 1369-6998 (Print) 1941-837X (Online) Journal homepage: https://www.tandfonline.com/loi/ijme20 Cervical cancer prevention in Morocco: a model- based cost-effectiveness analysis Wadie Messoudi, Toufik Elmahi, Chakib Nejjari, Nabil Tachfouti, Ahmed Zidouh, Ghali Saadani, David Moriña & Mireia Diaz To cite this article: Wadie Messoudi, Toufik Elmahi, Chakib Nejjari, Nabil Tachfouti, Ahmed Zidouh, Ghali Saadani, David Moriña & Mireia Diaz (2019) Cervical cancer prevention in Morocco: a model-based cost-effectiveness analysis, Journal of Medical Economics, 22:11, 1153-1159, DOI: 10.1080/13696998.2019.1624556 To link to this article: https://doi.org/10.1080/13696998.2019.1624556 © 2019 The Author(s). Published by Informa UK Limited, trading as Taylor & Francis Group View supplementary material Accepted author version posted online: 28 May 2019. Published online: 17 Jun 2019. Submit your article to this journal Article views: 905 View related articles View Crossmark data Full Terms & Conditions of access and use can be found at https://www.tandfonline.com/action/journalInformation?journalCode=ijme20 JOURNAL OF MEDICAL ECONOMICS 2019, VOL. 22, NO. 11, 1153–1159 https://doi.org/10.1080/13696998.2019.1624556 Article 0051-RT.R1/1624556 ORIGINAL RESEARCH Cervical cancer prevention in Morocco: a model-based cost-effectiveness analysis Wadie Messoudia, Toufik Elmahia, Chakib Nejjarib, Nabil Tachfoutib, Ahmed Zidouhc, Ghali Saadania, ~ad and Mireia Diaze,f David Morin a Laboratory of Coordination of Studies and Research in Analysis and Economic Forecast, Faculty of Law, Economics and Social Sciences, Universite Sidi Mohamed Ben Abdellah, Fez, Morocco; bLaboratory of Epidemiology, Clinical Research and Community Health, Faculty of Medicine, Universite Sidi Mohamed Ben Abdellah, Fez, Morocco; cLalla Salma Foundation for Cancer Prevention and Treatment, Rabat, Morocco; dBarcelona Graduate School of Mathematics (BGSMath), Departament de Matematiques, Universitat Autonoma de Barcelona (UAB), Cerdanyola del Valles (Barcelona), Spain; eUnit of Infections and Cancer (UNIC-I&I), Cancer Epidemiology Research Program, Institut Catala d’Oncologia (ICO) – IDIBELL, L’Hospitalet de Llobregat, Barcelona, Spain; fCentro de Investigacion Biomedica en Red (CIBERONC), Madrid, Spain ABSTRACT ARTICLE HISTORY Objective: Cervical cancer is a huge public health issue in Morocco which represents the second most Received 19 March 2019 frequent and fatal cancer among women. Countries that have not yet introduced the HPV vaccine Revised 20 May 2019 could benefit greatly, but before implementation it is necessary to perform country-specific economic Accepted 24 May 2019 assessments that include current screening practices. KEYWORDS Methods: A Markov model was developed to simulate the natural history of HPV and cervical cancer Cervical cancer; HPV so as to calculate the long-term health benefits and costs of HPV vaccination and current screening vaccination; screening; by visual inspection with acetic acid (VIA). Starting from a previous transition probability matrix used visual inspection with for a model from Spain, the present model was calibrated to cervical cancer incidence from Morocco. acetic acid (VIA); Cost survey data was used to estimate the cost of screening and clinical procedures from the public cost-effectiveness; healthcare perspective. Incremental cost-effectiveness ratios were calculated as 2018US$ per additional low-resource settings year of life saved (YLS) and both costs and health outcomes were discounted at 3%. Results: The expected reduction in lifetime risk of cervical cancer for current screening would be 14% JEL CLASSIFICATION CODES at a cost of US$551/YLS compared with no intervention, assuming VIA every 3 years in women aged I18; I19 30–49 at 10% coverage. HPV vaccination of pre-adolescent girls at 70% coverage would reduce the lifetime risk of cervical cancer by 62% at a cost of US$1,150/YLS, compared with no intervention. When implementing HPV vaccination in combination with current screening, vaccination would be dominated, and the combined strategy would provide a 69% reduction at a cost of US$2,843/YLS, compared with screening alone. Current screening would be dominated by vaccination when screen- ing coverage is higher than 15%, whereas the combined strategy rapidly exceeds US$4,000/YLS. Conclusions: HPV vaccination could be highly effective and cost-effective in Morocco. Current screen- ing would be good value for money compared with no intervention, but scaling-up screening cover- age would make it inefficient compared with vaccination. Introduction Cervical cancer is the third most common cancer among Cost-effectiveness analyses are increasingly required in the women worldwide, and this health issue mostly occurs healthcare decision-making process for the purpose of (85%) in low- and middle-income countries, accounting for informing about which new products to include in national nearly 12% of all female-related cancers2. In Morocco, cer- public healthcare systems or optimizing those that already vical cancer ranks second among women with an age-stand- exist1. Mathematical models are commonly used in these ardized incidence rate of 17.2 per 100,000 women and an analyses to simulate diseases and the long-term health out- age-standardized mortality rate of 12.6 in 2018. There are comes and economic impact of different interventions that currently more than 200 virus strains of human papillomavi- cannot be explored through experimental studies. The ultim- rus (HPV) classified according to the degree of carcinogenesis ate goal is to determine a set of strategies that result in risk, of which at least 13 are cancer-causing3. HPV types 16 good value for money. and 18 are the most virulent and are responsible for 70% of CONTACT Mireia Diaz mireia@iconcologia.net Unit of Infections and Cancer, Cancer Epidemiology Research Programme (CERP), Institut Catala d’Oncologia-IDIBELL, Av. Granvia 199-203, 08908 L’Hospitalet de Llobregat, Barcelona, Spain Supplemental data for this article is available online at https://doi.org/10.1080/13696998.2019.1624556. ß 2019 The Author(s). Published by Informa UK Limited, trading as Taylor & Francis Group This is an Open Access article distributed under the terms of the Creative Commons Attribution-NonCommercial-NoDerivatives License (http://creativecommons.org/licenses/by-nc-nd/4.0/), which permits non-commercial re-use, distribution, and reproduction in any medium, provided the original work is properly cited, and is not altered, transformed, or built upon in any way. www.tandfonline.com/ijme 1154 W. MESSOUDI ET AL. cervical cancer cases globally, as well as a varying proportion through the four FIGO stages. Women may die from cervical of other cancer and non-cancer diseases4,5. cancer in the cancer stages, or may die from other non- Screening activities have showed that early detection can cervical cancer causes in every health state and every cycle. reduce the morbidity and mortality associated with cervical The model was developed with Microsoft Excel 2010. cancer6. Several screening tests with different performances in terms of sensitivity and specificity are currently available, such as the Pap smear, visual inspection with acetic acid (VIA) and Calibrating the model HPV DNA. In Morocco, opportunistic cervical cancer screening For our initial calibration step, a transition probability matrix based on VIA every 3 years in women aged 30–49 was from a previous validated model for Spain was used to launched in 20107,8. However, the coverage rate at the reproduce cervical cancer incidence from Morocco14,15. To national level is only 6–10% of the targeted women, with ensure that the model fits the epidemiological data from 80–90% of all cervical cancer cases being diagnosed at an Morocco, a manual calibration process was performed to advanced stage7,9,10. This low coverage rate might be reduce uncertainty in transition probabilities selected from explained by the lack of mass communication and public various studies. The model was calibrated to reproduce age- awareness campaigns about the cervical cancer screen- specific cervical cancer incidence prior to launching the ing program7. screening program in 2010 (Supplementary Figure S2). Three prophylactic vaccines against HPV infections are currently available and have proved to be highly safe and effective for the prevention of HPV vaccine-type infection Strategies considered and associated diseases11. The bivalent vaccine protects Three cervical cancer preventive strategies were included in against HPV types 16 and 18; the quadrivalent vaccine pro- the base-case analysis: current screening in Morocco of tects against HPV types 6, 11, 16, and 18; and the nonavalent women aged between 30 and 49 years with a VIA test every vaccine targets the same HPV types as the quadrivalent vac- 3 years at 10% coverage rate; HPV 16/18 vaccination of pre- cine as well as types 31, 33, 45, 52, and 58. The World Health adolescent girls at 14 years old with two doses at 70% cover- Organization (WHO) points out that developing countries age rate assuming 90% efficacy and lifelong immunity; and with difficulties in implementing screening programs might combined vaccination and screening. reap an important benefit from HPV vaccination as a primary prevention strategy for cervical cancer disease12,13. On this basis, health policy decision-makers in countries Demographic and epidemiological data like Morocco need to know the cost-effectiveness of HPV The distribution of the Moroccan female population by age vaccination to decide the best way to allocate their scarce group was extracted from the United Nations’ population healthcare resources. Hence, the aim of this project is to prospects for 201516. Annual transition probabilities depict- assess the cost-effectiveness of implementing an HPV vaccin- ing the natural history of cervical cancer and HPV infection ation program in Morocco and the current screening with were also extracted from published epidemiologic articles VIA at different frequencies, both strategies separately, and under the assumption that the mechanism of cervical car- in combination. cinogenesis from initial infection is universal and that it does not differ greatly between countries17–20. Epidemiological Methods data on cervical cancer incidence and mortality by age were derived from the cancer registries of the Grand Casablanca The model Region and from the ICO-HPV Information Center21,22. The A Markov model was developed to simulate the natural his- sensitivity and specificity of VIA were derived from a system- tory of HPV infection and cervical cancer in order to calculate atic review conducted to compare the test accuracy of the the long-term health benefits and costs of different cervical HPV DNA test, cytology, and VIA23. The baseline assumptions cancer preventive strategies. As shown in Supplementary on screening and vaccination are shown in Table 1. Figure S1, the static model includes 12 boxes that represent mutually exclusive health states (healthy, HPV infection, cer- Economic data vical intraepithelial lesions (CIN) grade 1–3, International Federation of Gynecology and Obstetrics (FIGO) cervical can- Since it is important to use country-specific information but cer stages, survival, cervical cancer mortality, and mortality there is a lack of data on costs related to cervical cancer from other causes). Women move from one health state to screening procedures in Morocco, a survey was carried out another in 1-year intervals according to some transition to estimate the direct medical costs of screening, diagnosis, probabilities. All women start the model simulations as and treatment of pre-cancerous lesions using loop electrosur- healthy and can move to the HPV-infected state by acquiring gical excision procedure (LEEP) and treatment of cervical can- the infection with some probability. Women who have cer from the public healthcare perspective in 2015 Moroccan acquired HPV infection could develop CIN, and those with dirhams. The data collection and cost calculation methods CIN 1 or with CIN 2/3 could progress, regress, or stay in the are both included in the Supplementary Appendix. All costs same state. Once in the cancer state, a woman may not were converted to 2018 US dollars (US$) using GDP deflators regress to other health states, and instead progresses and average annual exchange rates24 (Table 1). JOURNAL OF MEDICAL ECONOMICS 1155 Table 1. Baseline assumptions and direct medical costs indexed at year 2018. Variable Value Range of values used in the sensitivity analysis Baseline assumptions VIA screening coverage 10% 5–70% Sensititvity of VIA test 69% 30–80% Sensititvity of colposcopy 95% Vaccination coverage 70% 30–100% Vaccine protection against HPV 16/18 90% Duration of HPV 16/18 protection Lifelong Direct medical costs (US$2018) Screening test (VIA) 36.7 18.35–73.4 Diagnostic Coloposcopy 36.6 Biopsy 27.1 CIN treatment Procedure of CIN1 63.7 Procedure of CIN2/3 136.7 Cancer treatment Stage I & II 1,026.9 Stage III 2,083.9 Stage IV 1,508.5 HPV vaccine cost per dose Vaccine 10.0 5.0–35.0 Administration 5.0 The lower price reported for HPV vaccine was offered by in the results, one-way and two-way sensitivity analyses were Merck to the Global Alliance for Vaccines and Immunizations carried out for the vaccination’s coverage and cost, as well (GAVI) for the poorest countries (US$4.50 per dose), but as for the coverage, sensitivity, and cost of the VIA test. Morocco is not eligible for GAVI support25. The negotiated price by the Revolving Fund of the Pan American Health Organization (PAHO) for low- and middle-income countries Results in Latin America is US$10–US$15 per dose26. Based on this, The expected reduction in lifetime risk of cervical cancer for we assumed an initial vaccine cost of US$10 per dose and a current screening is 14% compared with no intervention, US$5 administration cost. assuming VIA every 3 years in women aged 30–49 at 10% For the purpose of identifying those strategies that coverage (Figure 1). Pre-adolescent vaccination alone has an appear to be relatively good value for money, usually a cost- important impact on lifetime risk of cervical cancer, reducing effectiveness threshold is defined. However, there is no uni- it by 63% compared with no intervention and by 48% com- versal criterion and there is no consensus on how it should pared with VIA every 3 years. An additional reduction of 7% be derived27. For illustrative purposes, the approach of the is also obtained when combining vaccination with the exist- Commission on Macroeconomics and Health was used, since ing screening program compared with vaccination alone. there is no specific local threshold defined for Morocco28,29. A cost-effectiveness analysis was performed, comparing This heuristic suggests that an intervention should be the current screening in Morocco using VIA every 3 years considered highly cost-effective if the ICER is less than the with different frequencies of screening. It was found that, as country’s per capita gross domestic product (GDP) and cost- the frequency increases, both the effectiveness and the costs effective if the ICER is less than three times the per capita rise, and screening with VIA at any frequency would be GDP. The estimated per capita GDP in Morocco for the year dominated by VIA once-in-a-lifetime that would be cost-sav- 2018 is US$2,86024. ing compared with no intervention (Table 2A). Vaccination alone at a 70% coverage rate would cost US$1,150 per YLS compared with no intervention (Table 2B), well below the Outcomes, measurements, and cost- per capita GDP threshold. The current screening program effectiveness analysis with VIA in Morocco (10% coverage to women aged 30–49 For each strategy, the model predicts the life expectancy (LE) years) would cost US$551 per YLS compared with no inter- and the lifetime cost per woman. The incremental cost- vention (Table 2C). When implementing HPV vaccination in effectiveness ratio (ICER) was the measurement used to combination with existing screening, vaccination alone perform the cost-effectiveness analysis, defined as the differ- would be dominated, and the combined strategy has an ence in cost between two interventions, divided by the dif- incremental cost-effectiveness ratio of US$2,843 per YLS ference in health (LE in our case). Therefore, the ICER compared with screening alone, which would be considered represents the incremental cost associated with one add- cost-effective based on the GDP per capita threshold itional year of life saved (YLS). Both cost and health out- (Table 2C). When the screening coverage rate is higher than comes were discounted at an annual rate of 3%. To identify 15%, vaccination is no longer dominated, instead screening which parameters and assumptions are the most influential alone is dominated and the ICER for the combined 1156 W. MESSOUDI ET AL. No intervenon No intervenon 60 VIA screening 3y-VIA screening 50 Vaccinaon at 14 yo Vaccinaon 14% Cervical cancer incidence Vaccinaon + + Vaccinaon VIA screening 3y-VIA screening (per 100,000 woman) 40 30 48% 20 7% 10 0 0-14 15-39 40-44 45-49 50-54 55-59 60-64 65-69 70-74 >=75 Age-group Figure 1. Estimated impact on cervical cancer incidence for different strategies. Table 2. Discounted life expectancy (LE), discounted total lifetime cost per woman, and incremental cost-effectiveness ratio for evaluated strategies. Strategy Discounted cost per woman (US$) Discounted LE ICER (US$/YLS) A. VIA screening alone by frequency at 10% coverage No intervention 3.49569 54.95990 20y-VIA screening (once-in-a-lifetime) 2.48417 54.97585 CS 10y-VIA screening (twice-in-a-lifetime) 3.76766 54.97547 Dominated 5y-VIA screening (four times-in-a-lifetime) 6.97354 54.97526 Dominated 3y-VIA screening (seven times-in-a-lifetime) 11.70498 54.97480 Dominated B. Vaccination alone at 70% coverage No intervention 3.49569 54.95990 Vaccination (girls 14 years old) 22.36102 54.97630 1,150 C. VIA screening at 10% coverage and vaccination at 70% coverage No intervention 3.49569 54.95990 3y-VIA screening (seven times-in-a lifetime) 11.70498 54.97480 551 Vaccination (girls 14 years old) 22.36102 54.97630 Dominated Vaccination þ 3y-VIA screening (seven times-in-a-lifetime) 32.06297 54.98196 2,843 D. VIA screening and vaccination, both at 70% coverage No intervention 3.49569 54.95990 3y-VIA screening (seven times-in-a-lifetime) 73.61780 54.97157 Dominated Vaccination (girls 14 years old) 22.36102 54.97630 1,150 Vaccination þ 3y-VIA screening (seven times-in-a-lifetime) 95.05145 54.98143 14,170 strategy increases rapidly, exceeding US $4,000/YLS above VIA test is due to the low coverage rate of 10% assumed for 20% (Table 2D and Supplementary Table S4). the current existing screening and the small population cov- The one-way sensitivity analysis shows that the cost- ered (from 30 to 49), which excludes most of the female effectiveness of adding vaccination to the existing screening population at risk for cervical cancer from screening. program remains relatively stable on varying the cost per dose up to current tender prices in high-resource countries (US$35 per dose) and vaccination coverage (up to 100%); Discussion in neither case does the ICER exceed the per capita GDP. Our results show that scaling up coverage of screening with The combined strategy is also stable on varying the sensitiv- VIA more than 15% in Morocco is not cost-effective, whereas ity of the VIA test and the cost of VIA, whereas it is sensitive HPV 16/18 vaccination for pre-adolescent girls at age 14 with to the increase in coverage of the VIA test (Figure 2)30. a coverage rate of 70% is highly cost-effective and could Indeed, when the coverage of VIA goes from 5% to 100%, reduce cervical cancer by 63%. Provided the cost per vacci- the ICER jumps from US$997 per YLS to US$5,702 per YLS nated girl is US$10 per dose, we have found that vaccination compared with no intervention, exceeding the per capita alone costs US$1,150 per YLS, well below the 2018 per capita GDP threshold at a 50% coverage rate unless the cost of the GDP (US$2,860). Even when the vaccine cost per dose is VIA test drops by one third (Supplementary Figure S3). At increased up to current tender prices in high-resource coun- coverage of 75%, the combined strategy exceeds the per tries, pre-adolescent HPV vaccination would be considered capita GDP threshold unless the cost of the VIA test drops good value for money. The existing screening program with by one half. At the current price, the ICER for the combined VIA every 3 years for women aged between 30 and 49 years strategy is below three times the per capita GDP, even at at 10% coverage costs US$551 compared with no interven- 100% coverage. The slight variation on the sensitivity of the tion. However, this would be considered dominant compared JOURNAL OF MEDICAL ECONOMICS 1157 “Vaccinaon followed by 3y-VIA screening” compared to “No intervenon” ICER = US$ 1,295 / YLS Cost of VIA (/2 - x2) 1,058 1,768 VIA sensivity (80% - 30%) 1,263 1,355 VIA coverage (5% - 70%) 997 4,252 Vaccine cost per dose ($5 - $35) 978 2,334 Vaccinaon coverage (30% - 100%) 479 1,703 0 1,000 2,000 3,000 4,000 Incremental cost-effecveness rao (US$/YLS) Figure 2. One-way sensitivity analysis for vaccination followed by screening with VIA performed every 3 years. with vaccination at US$10 per dose and 70% coverage rate The WHO recommends introducing the HPV vaccines when coverage of VIA is higher than 15%. The high cost of into national immunization programs and using VIA in screening and subsequent follow-up and treatment together resource-constrained settings11,37. However, if HPV testing is with an inefficient screening with VIA and low coverage in a feasible, the WHO suggests a strategy of screening with an small fraction of women at risk of cervical cancer makes HPV test over a strategy of screening with VIA. This is screening alone unattractive in Morocco compared to HPV increasingly achievable, since new methods of HPV testing vaccination at reasonably low prices. Vaccination combined for low-resource settings that require low staff qualifications with existing screening practices in Morocco costs US$2,843 are being developed, alongside alternative sample collection per YLS, which is less than Morocco’s per capita GDP. The methods such as cervical specimen self-sampling38. higher the screening coverage, the greater the ICER, exceed- Furthermore, based on estimates of the manufacturing costs ing the per capita GDP threshold at a 50% coverage rate of HPV vaccines, the price for the GAVI alliance for introduc- unless the cost of the VIA test drops by one third. ing the HPV vaccine in the poorest countries could be To our knowledge, our study is the first cost-effectiveness around $0.50 per dose; much lower than the current price of analysis calibrated to the Moroccan context that takes into $4.5039. The same study states that prices for lower- and account the existing screening program. Only a simple Excel- middle-income countries could also be lowered. based model was used to project the population-level health Although there have been more studies on cervical cancer and economic impact of preadolescent HPV vaccination in 20 prevention in Africa in recent years, there are still several countries, including Morocco31. However, this model was not countries with little to no research ever conducted in this calibrated to Morocco and did not include screening strat- area40. Given the substantial morbidity and mortality caused egies. The authors conclude that HPV vaccination would be by this cancer in this setting, more research is needed to cost-saving at a cost of $5 per dose, cost-effective at a cost of inform about which feasible, sustainable strategies can maxi- $26.75 per dose, and would exceed the cost-effectiveness mize women’s health. Our analysis is the first attempt to threshold at a cost of $54.25 per dose. The calibration proced- assess the epidemiologic and economic impact of HPV vac- ure is a crucial aspect for the model’s performance, especially cination and existing screening practices in Morocco. The when it involves policy decisions15. Several cost-effectiveness ultimate aim is to provide information in the long-term on analyses in low- and middle-income countries have examined value for money for cervical cancer prevention strategies in the value of HPV vaccination in combination with VIA32–36. All Morocco to help decision-makers who have to work on pri- of these studies are in line with our results, considering HPV oritization of public health investment. vaccination to be cost-effective in resource-poor settings if the There are several limitations to our approach, including vaccine cost decreases significantly in relation to the current the model’s structure, input data, assumptions, and uncer- out-of-pocket price. In such circumstances, the best cervical tainty of the parameters. Some of these are inherent in deci- preventive option would be a combined strategy of VIA sion-analytic models and were addressed following scientific screening with HPV vaccination. criteria of good practices or using standard approaches41,42. 1158 W. MESSOUDI ET AL. For example, our model inevitably requires assumptions our results and to help guide the next steps on cervical can- about the costs of interventions and procedures, the effect- cer prevention in Morocco. Finally, a country like Morocco, iveness of VIA screening and vaccination, and about cervical where there has been problems with screening coverage, cancer incidence and HPV prevalence data. As the quality of should also consider evaluating the cost-effectiveness of HPV the results is only as good as the quality of the assumptions self-collection, which has demonstrated to be a powerful and synthesized data, the most reliable information from tool to increase coverage and has the potential to overcome Morocco was included in the model43. Subsequently, to many of the barriers shown by other screening techniques. ensure the credibility of the results, model predictions on cervical cancer incidence were calibrated to real empirical Transparency data from Morocco. In addition, sensitivity analyses were per- formed which can help appreciate the impact of uncertain- Declaration of funding ties in the key parameters. However, some other limitations This work was partially supported by grants from the Instituto de Salud are more difficult to address. For example, our model does Carlos III through the projects RD12/0036/0056, PI11/02090, PIE16/00049, not capture herd immunity effects, nor does it include the and PI16/01254 (Co-funded by European Regional Development Fund. potential benefits of vaccination on other non-cervical HPV- ERDF, a way to build Europe). With the support of the Secretariat for related diseases such as other anogenital cancers or genital Universities and Research of the Department of Business and Knowledge warts, and it does not reflect the impact of cross-protection of the Government of Catalonia. Grants to support the activities of research groups (SGR 2017–2019). Grant number 2014SGR756, against other non-vaccine HPV types. These limitations may 2014SGR1077 and 2017SGR1718. The research leading to these results has lead the health benefits of HPV vaccination to be under-esti- received funding from the European Union’s Seventh Framework mated. Further, as our model is a simple representation of Programme (FP7/2007–2013) under grant agreement n Health-F3-2013- the natural history of HPV infection and cervical cancer and 603019 (COHEAHR) and MNSIRSES [FP7-People-2013-IRSES612216]. The not a detailed model with screening complex strategies that research leading to these results has received funding from RecerCaixa incorporate different visits, all women diagnosed with a CIN [2015ACUP00129]. We thank CERCA Programme/Generalitat de Catalunya lesion are treated and there is no loss-to-follow-up. In order for institutional support and Lalla Salma Foundation for the Prevention and Treatment of Cancers for financial and technical support. None of to deal with all these issues, it would be necessary to use these entities played a role in data collection or analysis, or in the inter- more complex approaches such as dynamic microsimulation pretation of the results. models. Another limitation is that our analysis is carried out under the favorable assumption of high efficacy against HPV Declaration of financial/other relationships 16/18 with two doses; therefore, any evidence regarding the reduction of the efficacy in two doses, the inclusion of wan- WM, TE, CN, NT, AZ, GA, DM and MD state that they have no competing ing immunity or the need for a booster to obtain long-term personal or financial interests in relation to this study. MD-Institutional protection would lead to less attractive results for HPV vac- support: The HPV vaccine trials and epidemiological studies were spon- sored by GlaxoSmithKline (GSK), Merck, Roche, and Sanofi Pasteur MSD. cination. Finally, although the threshold based in the GDP is A peer reviewer on this manuscript has disclosed being a principal often used in cost-effectiveness analysis in low- and middle- investigator of several HPV screening projects sponsored in part by BD income countries, recent papers suggest that thresholds rep- Diagnostics, receiving honoraria from BD Diagnostics, being a principal resenting likely health opportunity costs tend to be below investigator of EU HORIZON2020 SME contract 666-800, principal investi- one GDP per capita44,45. In our analysis, the ICER for vaccin- gator of study sponsored in part by Agena Biotech, Genomica SAU, and ation represents an eleventh of the per capita GDP in Life River Biotech. The peer reviewers on this manuscript have no other Morocco, which may indicate that vaccination would be relevant financial relationships or otherwise to disclose. cost-effective with a much lower threshold; whereas the ICER for combined vaccination and existing screening with VIA Ethics approval and consent to participate and low coverage is between one and two times the per This manuscript has been revised for its publication by the Clinical capita GDP, which may indicate that it could be displacing Research Ethics Committee of Bellvitge University Hospital. Verbal more health than it generates. informed consent was obtained from all participants in this study. Data of participants were anonymized for the purposes of this analysis. The confidential information of the patients was protected according to Conclusions national normative data. Based on our results, healthcare policymakers in Morocco should explore the possibility of supporting the implementa- Acknowledgements tion of pre-adolescent HPV vaccination for girls. However, None reported. they should also consider that introducing HPV vaccine into a national immunization program depends not only on the cost-effectiveness, but also on other important factors such ORCID as affordability, sustainable financing, avertable burden, feasi- Mireia Diaz http://orcid.org/0000-0001-9360-4548 bility, and equity. Further analyses with more complex approaches that include other non-cervical HPV-related dis- eases, vaccines that provide protection against more high- References risk HPV types, and other screening tests such as HPV testing [1] Rawlins MD. Crossing the fourth hurdle. Br J Clin Pharmacol. as suggested by the WHO are needed in order to reconfirm 2012;73:855–860. JOURNAL OF MEDICAL ECONOMICS 1159 [2] Ferlay J, Colombet M, Soerjomataram I, et al. Global Cancer [24] The World Bank [Internet]. World Development Indicators, WDI online Observatory: Cancer Today. Lyon, France: International Agency version. [cited 2019 May] Available from: https://databank.worldbank. for Research on Cancer; 2018 [cited 2018 Oct 15]. Available from: org/data/reports.aspx?source=world-development-indicators. https://gco.iarc.fr/today. [25] GAVI alliance. Millions of girls in developing countries to be pro- [3] Papillomavirus Episteme (PaVE). National Institutes of Health (NIH), tected against cervical cancer thanks to new HPV vaccine deals. National Institute of Allergy and Infectious Diseases (NIAID). [cited [cited 2017, December 15]. Available from: http://www.gavi.org/ 2019 May] Available from: https://pave.niaid.nih.gov/#home. library/news/press-releases/2013/hpv-price-announcement/. [4] Clifford GM, Smith JS, Plummer M, et al. Human papillomavirus [26] Levin CE, Van Minh H, Odaga J, et al. Delivery cost of human types in invasive cervical cancer worldwide: a meta-analysis. Br J papillomavirus vaccination of young adolescent girls in Peru, Cancer. 2003;88:63–73. Uganda and Viet Nam. Bull WHO. 2013;91:585–592. [5] Mun ~oz N, Bosch FX, de Sanjos e S, et al. Epidemiologic classifica- [27] Kvizhinadze G, Wilson N, Nair N, et al. How much might a society tion of human papillomavirus types associated with cervical can- spend on life-saving interventions at different ages while remain- cer. N Engl J Med. 2003;348:518–527. ing cost-effective? A case study in a country with detailed data. [6] Vaccarella S, Franceschi S, Engholm G, et al. 50 years of screening Popul. Health Metr. 2015;13:15. [28] Bertram MY, Lauer JA, De Joncheere K, et al. Cost-effectiveness in the Nordic countries: quantifying the effects on cervical cancer thresholds: pros and cons. Bull WHO. 2016;94:925–930. incidence. Br J Cancer. 2014;111:965–969. [29] Marseille E, Larson B, Kazi DS, et al. Thresholds for the cost-effective- [7] Selmouni F, Sauvaget C, Belakhel L, et al. Organization and evalu- ness of interventions: alternative approaches. Bull WHO. 2015;93: ation of a pilot cervical cancer screening program in Morocco. Int J 118–124. Gynaecol Obstet Off Organ Int Fed Gynaecol Obstet. 2016;132:25–28. ~a D, Rodrıguez-Sal [30] Diaz M, Morin es V, et al. Moving towards an [8] Selmouni F, Belakhel L, Sauvaget C, et al. Evaluation of the organized cervical cancer screening: costs and impact. Eur J national cervical cancer screening program in Morocco: achieve- Public Health. 2018;28:1132–1138. ments and challenges. J Med Screen. [cited 2019 Jan 16]. [31] Kim J, Sharma M, O’Shea M, et al. Model-based impact and cost- DOI:10.1177/0969141318824627. effectiveness of cervical cancer prevention in the Extended [9] Berraho M, Obtel M, Bendahhou K, et al. Sociodemographic fac- Middle East and North Africa (EMENA). Vaccine. 2013;31:G65–G77. tors and delay in the diagnosis of cervical cancer in Morocco. [32] Diaz M, Kim JJ, Albero G, et al. Health and economic impact of Pan Afr Med J. 2012;12:14. HPV 16 and 18 vaccination and cervical cancer screening in India. [10] Elmajjaoui S, Ismaili N, El Kacemi H, et al. Epidemiology and out- Br J Cancer. 2008;99:230–238. come of cervical cancer in National Institute of Morocco. BMC [33] Goldie SJ, Diaz M, Kim S-Y, et al. Mathematical models of cervical can- Womens Health. 2016;16:62. cer prevention in the Asia Pacific region. Vaccine. 2008;26:M17–29. [11] WHO. WHO j Human Papillomavirus (HPV) position paper. Geneva: [34] Praditsitthikorn N, Teerawattananon Y, Tantivess S, et al. Economic WHO; 2017 [cited 2017, November 4]. Available from: http://www. evaluation of policy options for prevention and control of cervical who.int/immunization/policy/position_papers/hpv/en/. cancer in Thailand. PharmacoEconomics. 2011;29:781–806. [12] Cutts FT, Franceschi S, Goldie S, et al. Human papillomavirus and [35] Guerrero AM, Genuino AJ, Santillan M, et al. A cost-utility analysis HPV vaccines: a review. Bull WHO. 2007;85:719–726. of cervical cancer screening and human papillomavirus vaccin- [13] Tsu V, Murray M, Franceschi S. Human papillomavirus vaccination ation in the Philippines. BMC Public Health. 2015;15:730. in low-resource countries: lack of evidence to support vaccinating [36] Chanthavilay P, Reinharz D, Mayxay M, et al. The economic evalu- sexually active women. Br J Cancer. 2012;107:1445–1450. ation of human papillomavirus vaccination strategies against cer- [14] Georgalis L, de Sanjos e S, Esnaola M, et al. Present and future of vical cancer in women in Lao PDR: a mathematical modelling cervical cancer prevention in Spain: a cost-effectiveness analysis. Eur approach. BMC Health Serv Res. 2016;16:418. J Cancer Prev Off J Eur Cancer Prev Organ ECP. 2016;25:430–439. [37] WHO. WHO guidelines for screening and treatment of precancer- [15] Morin ~a D, de Sanjos e S, Diaz M. Impact of model calibration on ous lesions for cervical cancer prevention. Geneva: World Health cost-effectiveness analysis of cervical cancer prevention. Sci Rep. Organization; 2013. 2017;7:17208. [38] Toliman PJ, Kaldor JM, Tabrizi SN, et al. Innovative approaches to [16] World Population Prospects-Population Division-United Nations. cervical cancer screening in low- and middle-income countries. [cited 2018, April 11]. Available from: https://esa.un.org/unpd/wpp/. Climacteric. 2018;21:235–238. [17] Myers ER, McCrory DC, Nanda K, et al. Mathematical model for [39] Clendinen C, Zhang Y, Warburton RN, et al. Manufacturing costs of the natural history of human papillomavirus infection and cer- HPV vaccines for developing countries. Vaccine. 2016;34:5984–5989. vical carcinogenesis. Am J Epidemiol. 2000;151:1158–1171. [40] Finocchario-Kessler S, Wexler C, Maloba M, et al. Cervical cancer [18] Torvinen S, Nieminen P, Lehtinen M, et al. Cost effectiveness of prevention and treatment research in Africa: a systematic review from a public health perspective. BMC Womens Health. 2016;16:29. prophylactic HPV 16/18 vaccination in Finland: results from a [41] McGhan WF, Al M, Doshi JA, et al. The ISPOR good practices for modelling exercise. J Med Econ. 2010;13:284–294. quality improvement of cost-effectiveness research task force [19] Annemans L, R emy V, Oyee J, et al. Cost-effectiveness evaluation report. Value Health J Int Soc Pharmacoeconomics Outcomes Res. of a quadrivalent human papillomavirus vaccine in Belgium. 2009;12:1086–1099. PharmacoEconomics. 2009;27:231–245. [42] Trikalinos TA, Siebert U, Lau J. Decision-analytic modeling to [20] Szucs TD, Largeron N, Dedes KJ, et al. Cost-effectiveness analysis of evaluate benefits and harms of medical tests—uses and limita- adding a quadrivalent HPV vaccine to the cervical cancer screening tions. Med Decis Making. 2009;29:E22–E29. programme in Switzerland. Curr Med Res Opin. 2008;24:1473–1483. [43] Griebsch I. Economic evaluation in health care: merging theory [21] Cancer registries, Morocco. [cited 2018, March 23]. Available with practice. New York: Oxford University Press; 2001. from: http://www.irc.ma/en/statistiques/registre-des-cancers/. [44] Ochalek J, Lomas J, Claxton K. Cost per DALY averted thresholds [22] Bruni L, Albero G, Serrano B, et al. [Internet]. ICO/IARC for low- and middle-income countries - Research Database, The Information Centre on HPV and Cancer (HPV Information Centre). University of York. 2015 [cited 2019, April 29]. Available from: Human Papillomavirus and Related Diseases in Morocco. https://pure.york.ac.uk/portal/en/publications/cost-per-daly- Summary Report 10 December 2018. [cited 2018 Apr]. Available averted-thresholds-for-low-and-middleincome-countries(12487fa5- from: https://www.hpvcentre.net. e63f-4ac3-9fa4-03b2795065eb).html. [23] Mustafa RA, Santesso N, Khatib R, et al. Systematic reviews and [45] Shillcutt SD, Walker DG, Goodman CA, et al. Cost effectiveness meta-analyses of the accuracy of HPV tests, visual inspection with in low- and middle-income countries: a review of the debates acetic acid, cytology, and colposcopy. Int J Gynaecol Obstet. 2016; surrounding decision rules. PharmacoEconomics. 2009;27: 132:259–265. 903–917.