Cost-effectiveness of human papilloma virus vaccination in Iceland

Transcription

1 Acta Obstetricia et Gynecologica. 2009; 88: ORIGINAL ARTICLE Cost-effectiveness of human papilloma virus vaccination in Iceland KRISTJAN ODDSSON 1, JAKOB JOHANNSSON 2, TINNA LAUFEY ASGEIRSDOTTIR 3 & THOROLFUR GUDNASON 1 1 Directorate of Health, 170 Seltjarnarnes, 2 Department of Oncology, Landspitali University Hospital, and 3 Department of Economics, University of Iceland, 101 Reykjavik, Iceland Abstract Objective. To evaluate the likely cost-effectiveness of introducing routine HPV vaccination in Iceland. Design. Prospective costeffectiveness analysis of human papilloma virus (HPV) vaccination. Setting and sample. Population of 12-year-old girls in the Icelandic population. Methods. A model was developed, comparing a cohort of all 12-year-old girls alive in year 2006, with or without vaccination. The model was based on the epidemiology of cervical cancer in Iceland and its premalignant stages as well as the costs involved in the treatment of each stage, assuming that the vaccines only prevent infections caused by HPV 16/18 at an efficacy of 95% and participation rate of 90%, no catch-up vaccination, no vaccination of boys and no booster dose needed. All costs were calculated on the basis of the price level of mid-year 2006 with a 3% discount rate. Incremental cost-effectiveness ratio calculations were performed and sensitivity analysis was carried out on factors most relevant for cost-effectiveness. Results. Vaccination costs in excess of savings would be about e /year. Vaccination would reduce the number of women diagnosed with cervical cancer by almost 9, prevent the death of 1.7 women and result in 16.9 quality-adjusted life years gained annually. The incremental cost-effectiveness ratio was calculated to be about e18.500/quality-adjusted life year saved. Conclusion. HPV vaccination seems to be cost-effective in Iceland, but this was sensitive to various parameters in the model, mainly the discount rate, the price of the vaccines and the need for a booster dose. Key words: HPV, cervical cancer, cost-effectiveness Introduction Cervical cancer is the second most common cancer among women worldwide and it is estimated that 493,000 new cases are diagnosed annually and 274,000 of these women die from the disease (1). Screening for CIN 2 and 3 precancerous lesions (cervical intraepithelial neoplasia grade 2 and 3) of the cervix with Pap (Papanicolaou) smear has lowered the incidence and number of deaths in the developed world (2). Human papilloma viruses (HPV16 and/or 18) are the causal agents in approximately 70% of cervical cancer (3), but a recent study in Iceland has shown a ratio of 60% (4). At least 15 types of HPV are associated with cervical cancer (5). HPV also causes anal and genital warts, especially HPV 6 and 11 (6,7). In addition, HPV infection has been related to anal, oropharyngeal, vulval and vaginal cancer and juvenile laryngeal papillomatosis, which may progress to cancer (1,8). About 17 new cases of cervical cancer are diagnosed each year in Iceland, most being among women who have not participated in the screening program. More women are diagnosed with precancerous lesions and over 300 conizations are carried out annually because of CIN 2/3 changes (9,10). Two vaccines against HPV infections have been licensed and registered in Iceland: Cervarix Ò from GlaxoSmithKline, UK, a bivalent HPV16/18 vaccine and Gardasil Ò from Merck and Co., Inc, USA, a quadrivalent HPV16/18/6/11 vaccine. Several costeffectiveness studies and health technology Correspondence: Kristjan Oddsson MD, Directorate of Health, 170 Seltjarnarnes, Iceland. Tel: / Fax: oddsson@hotmail.com (Received 6 April 2009; accepted 8 September 2009) ISSN print/issn online Ó 2009 Informa UK Ltd. (Informa Healthcare, Taylor & Francis AS) DOI: /

2 1412 K. Oddsson et al. assessments have shown that vaccination with these vaccines appears to be cost-effective. Health authorities in many countries already have or are considering ingincluding them in their vaccination programs to vaccinate girls and young women (11 19). The aim of this study was to estimate the cost-effectiveness of HPV vaccination in Iceland against cervical cancer based on Icelandic data; incremental cost-effectiveness ratio (ICER) calculations were based on a model comparing a cohort of 12-year-old girls with and without vaccination. Material and methods A static model was developed based on Icelandic data on the epidemiology of HPV i infections, cervical cancer, its premalignant stages and the cost involved in the treatment of each stage (4,10,20 22). The model compared the cohort of all (2267) 12-yearold girls alive in the year 2006 with and without vaccination, commencing in the year A lifelong follow-up was assumed, setting up a scenario regarding diagnosis of cervical cancer, its precursors and cost based on Icelandic data when available. Otherwise information from Clifford et al. was used (23). The population numbered at the end of the year 2006, with females and girls in the 12-year-old cohort (24). In the analysis, the vaccines were assumed to only prevent infections caused by HPV16/18 (60%) (4). Vaccine efficacy was estimated to be 95% and the participation rate among 12-year-old girls who would be vaccinated at 0, 1 and 6 months before sexual debut as 90%. We anticipated no catch-up vaccination, vaccination of boys or booster doses. All costs were calculated on the basis of the costs in mid-year 2006 with a 3% discount rate. Besides health care costs, no other societal costs were included in the analysis. Also besides cost of vaccine, it was assumed that the program would be incorporated into the existing vaccination program, without any additional cost. The exact price of the vaccine in a general vaccination program is unclear, but a tender of the two licensed HPV vaccines was estimated to be 75% of the official registered wholesale price (e107 in 2006) and this price was used in base case calculations. The model assumes that cervical cancer progresses through the stages of low-grade to high-grade cervical lesions and cervical intraepithelial neoplasia (CIN 2/ 3) and finally to cervical cancer. On average, this process takes 33 years, from age 12 (t = 0; time at vaccination) until age 45 (t = 33), which is the average age of Icelandic women diagnosed with cervical cancer. These 33 years of progression to cervical cancer were divided into four stages after t = 0. The first stage is low-grade cervical lesions (LG), where it was assumed that the average age of women diagnosed with LG lesions is 24 years (t = 12) (10). If LG lesions are diagnosed it would lead to four additional Pap smears based on the guidelines of the Icelandic Cancer Society (21). The second stage includes high-grade cervical lesions (HG), where it was assumed that the average age of women diagnosed with HG is 28 years (t = 16). According to the guidelines the diagnosis of a HG lesion would lead to colposcopy and cervical biopsy (25). A third stage is CIN 2/3, where a biopsy taken because of a HG lesion may reveal CIN 2/3, which usually leads to conization (t = 16). Surveillance after conization would require three visits with Pap smears (25). A final stage is cervical cancer, where the average age of Icelandic women at diagnosis is 45 years (t = 33) (9). The treatment of cervical cancer stage 1A is usually hysterectomy with two visits scheduled annually for three years, a total of six visits. Treatment of cervical cancer stage higher that 1A includes total hysterectomy and/ or radiation and chemotherapy with the same annual follow-up schedule as stage 1A resulting in thirteen scheduled visits over five years. Sensitivity analysis was carried out on the following parameters: the price of the vaccine, administrative costs of vaccination, booster doses, broader anti-viral spectrum of the vaccines (cross-protection) and loss of work days. Incremental cost-effectiveness ratio was calculated as the net cost of vaccination per quality-adjusted life year (QALY) saved. The cost-effectiveness analysis was based on treatment costs for cervical cancer and precancerous lesions with and without vaccination for the cohort of all 12-year-old girls alive in the year The model assumed that the screening program for cervical cancer continued without change and, therefore, screening costs were not included. Cost was calculated using mid-year 2006 without value added tax (VAT). Quality-adjusted life year was calculated from presumed changes in mortality and incidence of cervix cancer as an effect of vaccination. About nine women would be saved from diagnosis of cervical cancer by vaccination of the cohort and almost two lives saved annually. These savings continue over the life of the cohort and were discounted. According to Goldie et al. (2004) the quality of life for those who are diagnosed with cervical cancer is about 60% compared to those who remain healthy. We decided to be conservative and to use the figure 65% instead as these are not Icelandic figures. The quality of life gained by vaccination for those women who avoided

3 Cost-effectiveness of HPV vaccination in Iceland 1413 cervical cancer was calculated over two years from presumed diagnosis. The quality of life gained for those women where vaccination saved their life was calculated over the lifespan of the cohort. Absence from work because of treatment for cervical cancer or precancerous lesions was not accounted for in the analysis. However, in the sensitivity analysis, absence from work due to the operative procedures was accounted for, two days after conization and two months after treatment for cervical cancer. Results Table I shows the number of women based on model calculations at each stage, unit and total treatment costs for each disease stage, in addition to unit and total costs of vaccination. Table II shows the costs of vaccination in excess of savings. Total costs of vaccination would be around e annually while yearly savings would be around e because of a decrease in cervical cancer and its precancerous lesions. Therefore, vaccination costs in excess of savings would be about e each year. In 2006, the life expectancy of Icelandic 12-yearold girls was 83 years (24) and five-year survival of Icelandic women diagnosed with cervical cancer was 80.2% (9). Vaccination of each cohort of 12-year-old girls would reduce five-year death rates and increase the quality of life of those women who otherwise would have been diagnosed with cervical cancer (14). The death rate would be reduced by 50% by vaccination and accounting for QALY reduction for cervical cancer survivors, vaccination of each cohort of 12-year-old girls would reduce the number of women diagnosed with cervical cancer by almost 9 women each year and prevent the death of 1.7 women annually (14). By the same assumption vaccination would result in 16.9 QALY gained annually. This resulted in a calculated ICER of e for every QALY saved. A sensitivity analysis assuming official registered wholesale price showed that the incremental costeffective ratio would worsen from about e to If a vaccination booster dose was required after 10 years, the ICER would worsen from about e to , given the same wholesale costs. A 6% depreciation rate instead of the conventional 3% would result in a worsened ICER from about e to Broader coverage (cross-protection) of the vaccine against oncogenic HPV types, such as assuming increased cross-protection from 35/ 40/40/60 to 35/57/57/70 for LG, HG, CIN 2/3 and cervical cancer, respectively, would improve the ICER from about e to Administrative costs and absence from work did not have any substantial effect on the ICER calculation. Discussion The aim of this study was to examine the costeffectiveness of HPV vaccination against cervical cancer in Iceland, a small but defined population. Vaccination of every cohort of 12-year-old girls would in 33 years reduce the annual number of women with cervical cancer by 8.7 and deaths by 1.7. The annual costs of vaccination was estimated as e without VAT with savings of e due to reduction of cervical cancer and its precancerous lesions and a gain of 16.9 QALY. The cost-effectiveness of vaccination calculated as costs per QALY was therefore e18.547, slightly lower than in Denmark and Norway (15,16), but similar to that of Canada and the United States (14,18,26). The difference between cost-effectiveness in this study and the results in other countries might be due to different reasons, for example, vaccine prices, whether herd immunity is included or not, coverage rate used, inclusion of booster or not, the inclusion of productivity losses or not and different treatment costs of cervical cancer and its precancerous lesions. Most vaccination programs will give immediate results in lowering the incidence of diseases, while the benefits of HPV vaccination will take years or decades to take effect. After the introduction of general HPV vaccination program it would take on average 12 years to notice any reduction in LG lesions, 16 years for HG and CIN 2/3 lesions and 33 years for cervical cancer. As long as current HPV vaccination does not prevent all cervical cancer it is of paramount importance that screening programs for cervical cancer continue. A recent audit of the Swedish cervical screening program has shown that non-adherence to screening intervals was the major reason for cervical cancer morbidity (27). In Belgium, it has been estimated that a 10% reduction in screening attendance would eliminate the benefits of HPV vaccination (19). However, the effect of HPV vaccination might call for changes in the screening program in the future. The calculated cost-effectiveness in this study was based on the best available data at the time and the results represent the gain of HPV vaccination based on the criteria used. Possible cross protection was not accounted for, but both vaccines probably prevent HPV infection by other oncogenic HPV viruses than HPV16/18 (28,29). Cross-protection would improve the cost-effectiveness. In addition, both vaccines are likely to prevent other types of cancer related to HPV16/18 infection. Gardasil Ò also prevents most

4 1414 K. Oddsson et al. Table I. Number of women based on model calculations at each stage, unit and total treatment costs for each disease stage, in addition to unit and total costs of vaccination. Stage of disease Time (year) Number of women a Unit costs b Total costs c Vaccination (3 injections) LG lesions HG lesions CIN 2/3 lesions Cervical ccancer Consultation Stage 1A Stage higher than 1A 33 7 Total hysterectomy Internal and external radiation and chemotherapy a Number of females vaccinated or diagnosed with a disease annually. b Cost per female in euro, price level of c Total costs in euro, price level of Table II. Costs of vaccination in excess of savings. Stage of disease t (year) Percentage Savings due of 16/18 a Costs b to vaccination c Costs in excess of savings d Vaccination (3 injections) LG lesions HG lesions CIN 2/3 lesions Cervical cancer Consultation Stage 1A Stage > 1A Total hysterectomy Internal and external radiation and chemotherapy Total a Percentage of HPV16/18 related disease (4). b Total costs in euro, price level of c Calculated savings related to percentage of HPV that causes disease at each stage, 90% vaccination participation rate and 95% efficacy of vaccine. Savings depreciated by 3% annually for t years. d Total costs of vaccination minus depreciated total savings. infections due to HPV6/11, which are responsible for about 90% of genital warts (30). Including this broad effect of the vaccines in the analysis would improve the cost-effectiveness, especially due to additional savings in the first few years after initiation of vaccination (18,26). Different criteria used in the calculations can have a substantial effect on cost-effectiveness, as observed in the sensitivity analysis. If the discount rate increased from 3 to 6%, the ICER would worsen by 464%. With high general interest rates it might be argued that a discount rate of 6% or higher should have been used instead of 3%, which would have worsened the costeffectiveness potentially even more. However, a 3% rate was chosen to facilitate comparison between countries (31). Studies have shown that catch-up vaccination and the vaccination of boys worsened the cost-effectiveness of HPV vaccination (15,18). Catch-up vaccination of girls would prevent cervical cancer of more women, but at a proportionally greater cost. Some uncertain factors may influence the costeffectiveness calculations. This study was based on a static model, a model that assumes a follow-up of

5 Cost-effectiveness of HPV vaccination in Iceland 1415 cohorts of women and does not take into account transmission of virus between individuals like socalled dynamic models do. The latter models are in theory superior to static models, as they account for herd immunity. However, it has been pointed out that dynamic models rely on even more assumptions and uncertainties than static models do. Herd immunity, due to widespread vaccination, may improve the estimated cost-effectiveness. However, replacement of HPV types by types not included in the vaccine may worsen the estimated cost-effectiveness of immunization. There is no consensus for what should be considered cost-effective in terms of cost per QALY gained. In Canada and UK a cost under Canadian dollars and pounds, respectively, has been regarded as cost-effective (31,32). In conclusion, HPV vaccination of 12-year-old girls in Iceland seems to be cost-effective, but the calculations are sensitive to various parameters in the analysis, mainly the final price of the vaccine and the need for a booster dose. The final decision on general HPV vaccination in Iceland has not been made. Acknowledgements Kristjan Sigurdsson and Anna Salvarsdottir, Icelandic Cancer Society, contributed with valuable information on the screening. Disclosure of interest: No financial support was received for this project. There are no conflicts of interests. References 1. Parkin DM, Bray F. Chapter 2: The burden of HPV-related cancers. Vaccine. 2006;24:S Sigurdsson K. The Icelandic and Nordic cervical screening programs: trends in incidence and mortality rates through Acta Obstet Gynecol Scand. 1999;78: Bosch FX, Lorincz A, Munoz N, Meijer CJ, Shah KV. The causal relation between human papillomavirus and cervical cancer. J Clin Pathol. 2002;55: Sigurdsson K, Taddeo FJ, Benediktsdottir KR, Olafsdottir K, Sigvaldason H, Oddsson K, et al. HPV genotypes in CIN 2-3 lesions and cervical cancer: a population-based study. Int J Cancer. 2007;121: Munoz N, Bosch FX, de Sanjose S, Herrero R, Castellsague X, Shah KV, et al. Epidemiologic classification of human papillomavirus types associated with cervical cancer. N Engl J Med. 2003;348: Koutsky L. Epidemiology of genital human papillomavirus infection. Am J Med. 1997;102: Feoli-Fonseca JC, Oligny LL, Brochu P, Simard P, Falconi S, Yotov WV. Human papillomavirus (HPV) study of 691 pathological specimens from Quebec by PCR-direct sequencing approach. J Med Virol. 2001;63: Smith EM, Swarnavel S, Ritchie JM, Wang D, Haugen TH, Turek LP. Prevalence of human papillomavirus in the oral cavity/oropharynx in a large population of children and adolescents. Pediatr Infect Dis J. 2007;26: Icelandic Cancer Registry [homepage on the Internet]. Reykjavik: [cited 2007 Nov 1]. Available at: krabbameinsskra.is/index.jsp?id=a. 10 Icelandic Cancer Society [homepage on the Internet]. Reykjavik: [cited 2007 Nov 1]. Available at: PageID= Newall AT, Beutels P, Wood JG, Edmunds WJ, MacIntyre CR. Cost-effectiveness analyses of human papillomavirus vaccination. Lancet Infect Dis. 2007;7: Elbasha EH, Dasbach EJ, Insinga RP. Model for assessing human papillomavirus vaccination strategies. Emerg Infect Dis. 2007;13: Taira AV, Neukermans CP, Sanders GD. Evaluating human papillomavirus vaccination programs. Emerg Infect Dis. 2004;10: Goldie SJ, Kohli M, Grima D, Weinstein MC, Wright TC, Bosch FX, et al. Projected clinical benefits and cost-effectiveness of a human papillomavirus 16/18 vaccine. J Natl Cancer Inst. 2004;96: Sundhedsstyrelsen [National Board of Health]. Copenhagen: [cited 2007 Nov 1]. Available at: Publ2007/MTV/HPV/HPV_vaccination.pdf. 16. Nasjonalt kunnskapssenteret for helsetjenesten [Norwegian Knowledge Centre for the Health Services] Oslo: [cited 2007 Nov 1]. Available at: no/filer/rapport_5_07_hpv_vaksiner.pdf. 17. Kulasingam S, Connelly L, Conway E, Hocking JS, Myers E, Regan DG, et al. A cost-effectiveness analysis of adding a human papillomavirus vaccine to the Australian National Cervical Cancer Screening Program. Sex Health. 2007;4: Brisson M, Van de Velde N, De Wals P, Boily MC. The potential cost-effectiveness of prophylactic human papillomavirus vaccines in Canada. Vaccine. 2007;25: Thiry N, Lambert MR, Cleemput I, Huybrechts M, Neyt M, Hulstaert F. HPV Vaccination for the Prevention of Cervical Cancer in Belgium: Health Technology Assessment. 2007;64C: Icelandic Cancer Registry [homepage on the Internet]. Reykjavik: [cited 2007 Nov 14]. Available at: Social Insurance Administration [homepage on the Internet]. Reykjavik: [updated 2009 Jan 1; cited 2007 Nov 14]. Available at: serfraedinga_09.pdf. 22. Landspitali University Hospital [homepage on the Internet]. Reykjavik[cited 2002 Nov 14]. Available at: Clifford G, Franceschi S, Diaz M, Munoz N, Villa LL. Chapter 3: HPV type-distribution in women with and without cervical neoplastic diseases. Vaccine. 2006;24: S Hagstofan [Statistics Iceland]. [homepage on the Internet]. Reykjavik: [updated 2009 Apr 1; cited 2007 Nov 14]. Available at: Dialog/varval.asp?ma=MAN00000%26ti=Lykilt%F6lur+ mannfj%f6ldans+1703%2d2009+%26path=../database/ mannoldi/yfirlit/%26lang=3%26units=fjöldi.

6 1416 K. Oddsson et al. 25. Icelandic Cancer Society [homepage on the Internet]. Reykjavik: [cited 2002 Nov 1]. Available at: lisalib/getfile.aspx?itemid= Sanders GD, Taira AV. Cost-effectiveness of a potential vaccine for human papillomavirus. Emerg Infect Dis. 2003;9: Andrae B, Kemetli L, Sparen P, Silfverdal L, Strander B, Ryd W, et al. Screening-preventable cervical cancer risks: evidence from a nationwide audit in Sweden. J Natl Cancer Inst. 2008;100: Harper DM, Franco EL, Wheeler CM, Moscicki AB, Romanowski B, Roteli-Martins CM, et al. Sustained efficacy up to 4.5 years of a bivalent L1 virus-like particle vaccine against human papillomavirus types 16 and 18: follow-up from a randomised control trial. Lancet. 2006;367: Szarewski A. Prophylactic HPV vaccines. Eur J Gynaecol Oncol. 2007;28: Greer CE, Wheeler CM, Ladner MB, Beutner K, Coyne MY, Liang H, et al. Human papillomavirus (HPV) type distribution and serological response to HPV type 6 virus-like particles in patients with genital warts. J Clin Microbiol. 1995;33: Devlin N, Parkin D. Does NICE have a costeffectiveness threshold and what other factors influence its decisions? A binary choice analysis. Health Econ. 2004;13: Baldwin JR B, Maynard JP. Insights on the Canadian Economy: Interprovincial differences in GDP per capita, labour productivity and work intensity: StatCan Catalogue No MIE-No. 2005;011:2005.