Human Papillomavirus Vaccination

The government has developed a simple model to estimate the cost-effectiveness of HPV vaccination in the context of current cervical cancer screening in the United States. We found that the cost per QALY gained by adding routine vaccination of 12-year-old girls to existing screening practices ranged from $3,906 to $14,723 under base-case parameter values (depending on the model version we applied) and ranged from <$0 (cost-saving) to $122,976 in the sensitivity analyses when several key parameter values were varied. Our results were consistent with results of published studies based on more complex models, particularly when key assumptions (e.g., vaccine duration, efficacy, and cost) were similar.

The simplicity of the approach offers advantages and disadvantages. The main advantage is that it requires substantially fewer assumptions than the more complex Markov and transmission models. For example, there is no need to model the probability of HPV acquisition, the possible progression from HPV infection to disease, the mixing of sex partners, the probability of HPV transmission, and so forth. There also is no need to model cervical cancer screening and sexually transmitted disease prevention activities because these activities are reflected in the incidence rates of HPV-related health outcomes that we applied.

Because we do not model cervical cancer screening directly, however, we are unable to use our model to examine how changes in cervical cancer–screening strategies can affect the cost-effectiveness of HPV vaccination, and vice versa. For example, HPV vaccination is expected to reduce the positive predictive value of abnormal Papanicolaou (Pap) test results (38). However, our analysis did not include the loss in quality of life attributable to the initial distress associated with receiving an abnormal Pap result (39), regardless of whether it is a false positive. This omission of the lost QALYs due to abnormal Pap test results underestimates the benefits of HPV vaccination because vaccination is expected to offer moderate reductions in the number of abnormal Pap results overall (38,40). Future changes in screening strategies, such as delayed screening, could also possibly improve the cost-effectiveness of HPV vaccination .

Another disadvantage of our approach is that it offers only a rough approximation of the cost-effectiveness of HPV vaccination and is not suitable for examining strategies such as vaccination of boys and men. In addition, although many of the parameter values and assumptions in our model can be modified with ease, changing the assumption of lifelong duration of protection or examining vaccination at older ages would require the incorporation of assumptions about the incidence and natural history of HPV to account for the probability of acquiring HPV (before vaccination or after vaccine immunity wanes) and the subsequent probability of adverse HPV-attributable health outcomes. However, we can address the issue of waning immunity by assigning a higher cost per vaccination series (as in the sensitivity analyses) to reflect the cost of a booster.

Another limitation of our approach is the uncertainty in the key parameter values, such as the cost and loss in quality of life associated with HPV-related health outcomes, the percentage of health outcomes attributable to each type of HPV targeted by the vaccine, and the incidence of CIN and Genital Warts. However, our results were fairly robust in response to changes in these key parameter values. For example, when simultaneously varying the costs of HPV-related health outcomes and the loss in QALYs associated with HPV-related health outcomes, we found that the estimated cost per QALY gained by vaccination ranged from $3,262 to $21,779.

Our adjustments for the effect of herd immunity were arbitrary; we simply assumed an additional effect of vaccination in the nonvaccinated population. However, our results did not vary substantially (in absolute terms) when the assumed effect of herd immunity was varied. For example, the estimated cost per QALY gained by quadrivalent vaccination (including herd immunity and excluding the benefits of preventing cancers other than cervical) was $5,336 in the base case and ranged from $3,423 to $7,596 when the adjustments for the effects of herd immunity (including the impact on Genital Warts in males) were varied. We also note that the benefits to nonvaccinated persons were assumed to occur only in nonvaccinated persons of similar ages to those vaccinated. This restriction may have understated the potential benefits of herd immunity.

Our analysis did not address all of the potential costs and benefits of vaccination. For example, the cost-effectiveness estimates would have been more favorable to vaccination if we had included the potential for cross-protection against high-risk HPV types besides 16 and 18 ; the prevention of anal, vaginal, and vulvar cancer precursor lesions ; the prevention of other cancers not included in this analysis (such as anal cancer and oropharyngeal cancers in male patients); and the prevention of other HPV-related health outcomes such as recurrent respiratory papillomatosis. Conversely, the cost-effectiveness estimates would have been less favorable to vaccination if we had included the potential for HPV type replacement (i.e., an increase in HPV types not protected against by vaccination), waning immunity, and the possible costs and loss in quality of life associated with adverse side effects of vaccination.

A key finding from this analysis was that the choice of discount rate and time horizon has a substantial influence on the estimated cost-effectiveness of vaccination. Because the costs of HPV vaccination begin to accrue immediately but the full benefits of vaccination are not realized for many years, the cost-effectiveness of vaccination becomes less favorable when higher discount rates are applied or when shorter time horizons are examined.

Another key finding was that the potential benefits of preventing anal, vaginal, vulvar, and oropharyngeal cancers offer nontrivial improvements in the estimated cost-effectiveness of HPV vaccination. The inclusion of these additional benefits decreased the cost per QALY gained by vaccination by ?$2,200 (or 21%) in the population model (without herd immunity), by ?$1,400 (or 27%) in the population model (with herd immunity), and by ?$2,200 (or 25%) in the cohort model. Future studies that develop better estimates of the cost and loss in quality of life associated with these cancers could more accurately estimate the effects of these additional benefits on the cost-effectiveness of HPV vaccination. Despite the limitations discussed above, our simplified model provides useful estimates of cost-effectiveness of HPV vaccination in the United States. Our results were consistent with previous studies based on more complex models. This consistency is reassuring because models of various degrees of complexity will be essential tools for policy makers in the development of optimal HPV vaccination strategies.