A new frontier of cervical cancer prevention began with use of human papillomavirus (HPV) vaccines, but the question is: which vaccine to choose? There are now two highly efficacious vaccines approved in many developed countries, the Merck quadrivalent HPV 6/11/16/18 (Gardisil) and the GSK bivalent HPV 16/18 (Cevarix). Many countries have committed to HPV immunisation programmes, and most of them, like the USA, published recommendations before the approval of the bivalent vaccine.1 2 The UK has now chosen the bivalent product in contrast to the rest of Europe that has gone with the quadivalent product, although this decision has led to criticism.3 4 Presumably the decision would be based on the economic consequences, since this is the major determinant for policymakers. Vaccine cost played a significant role in the initial rejection to add HPV vaccine to the formulary of the Australian National Immunization Program under the recommendation of the Pharmaceutical Benefits Advisory Committee, although this was soon overturned, resulting in a successfully launched national vaccination programme.1 5 The Irish HPV immunisation plan was interrupted due to the outright cost of any HPV vaccination programme, even though HPV 16/18 vaccination showed a favourable cost-effectiveness compared with other countries.6 7 In the Ireland cost-effectiveness study, they did not try to determine the benefit of the quadrivalent vaccine, which may have been because the cost of genital warts was not known. Dee et al (see page 402) try to define the cost of genital warts treatment in Irish genitourinary medicine clinics (GUM).8 Among five Irish GUM clinics, one-quarter of all visits and time were genital-wart-related at a cost of €355 per incident episode.

The cost derived by Dee et al did not include hospital costs but was similar to the costs found in Australia and the USA.9 10 A recent paper in STI from British Columbia, Canada estimated that a genital wart episode would cost C$190 on average and C$110 for physician visits, which is significantly less than that found by Dee et al.10 The discrepancy may be due to methods regarding unit versus proportional cost calculations. In British Columbia, 76% of genital wart visits were to physicians that bill a flat visit fee regardless of diagnosis, such that the paid amount for genital wart treatment is the same regardless of the clinic time used. The number of visits per episode may also be underestimated because of visits missed under a general ICD 9 code (099/788/789) or a code for a concurrent condition where only one code is usually reported. In the British Columbia STD Clinic, the estimate for cost is not well explained but is not likely based on prospectively measured clinic time.

Another Canadian study comparing the cost–utility of the quadrivalent and bivalent vaccines used an estimate of C$338 per episode of genital warts and predicted a C$10 000 per quality-adjusted life year (QALY) advantage for quadrivalent vaccines if all vaccines were of equal cost (C$400), and to achieve equivalence the bivalent cost would need to be 25% cheaper.11 This did not include recurrent respiratory papillomatosis, a condition that in most instances is vertically transmitted from mothers where the infant develops long-term papillomatosis of the larynx requiring extensive surgeries.12 A recent publication has done this additional calculation in the USA scenario.13 Chesson et al used the estimate of 0.7 cases per 100 000 children per year with treatment cost estimated at US$200 000 per case. The authors calculated a 14–21% reduction in the cost per QALY gained using the quadrivalent vaccine.

Considering the UK decision, the analysis done by Jit et al suggested that the bivalent vaccine would need to cost only £13–21 less to be equivalent in cost-effectiveness with a quadrivalent vaccine (22–25% of £60–80.50). This is based on a conservative cost of genital warts of £134 per episode from a small cohort studied by Woodhall et al, who also compared health-related quality of life among clients of GUM in York, UK.14 Woodhall et al, in their STI paper, had over 50% of clients with genital warts report a significant or very significant impact on their quality of life, and on average there was a statistically significant lower quality of life score by about 16%. The impact on QALY based on Woodhall et al is greater than in the Canadian paper, and this may account for the UK and Canadian papers both arriving at similarly results. However, to further adjust for recurrent respiratory papillomatosis (RRP), as suggested by Chesson et al, the cost of the bivalent would need to be 35.5–40.75% less than quadrivalent. Then, in the UK case, if the cost estimate for genital warts is more than the amount used by Jit et al by a factor of up to 2–3 (suggested by Dee et al) and if one includes RRP, the actual cost for the bivalent vaccine potentially may need to be more than 50% less than the quadrivalent vaccine.

Although the prevention of cervical cancer is of primary importance in the choice of a HPV vaccine, there are also very good reasons, as listed above, to include HPV strains 6 and 11 in a combined product for a vaccine campaign. The caveats to this choice are that adding these additional components has no deleterious effects, none of which have been found, and that the cost of those additional components does not prevent any vaccine programme from existing. Of course, basic economic principles require a competitive market to reduce the costs of the vaccine, and so the more choices of HPV vaccine, the better, because it is also of critical importance that the costs of the vaccine eventually come down or are made available at a discounted price to developing countries where the impact of a vaccine would be greatest because of poor cervical screening programmes.15 The choice of vaccine for any country or state will always need to consider cost, but the rationale for a vaccine that includes HPV 6 and 11 is compelling.