Estimating uncertainties of hpv16/18 vaccination a dynamic modelling
Health technology assessment
|Updated
Key message
Background
The National Council on Quality and Prioritisation in Health Care asked the Norwegian Knowledge Centre for the Health Services to quantify how several unresolved issues regarding long term efficacy may affect the future burden of cervical cancer and health economics of HPV 16/18 vaccination.
Method
We used the dynamic natural history model and health economic approach as previously described in Report No. 12-2007. The analyses are made in a twostep approach:
i) a natural history model is used to estimate the development in pre-cancerous and cancerous lesions in the time period 2008-2060,
ii) health economic evaluations are performed based on the predicted burden of disease.
Outcomes are compared to a base case vaccination-screening scenario assuming 90 percent vaccine coverage among 12-year old girls, 90 percent vaccine efficacy and a mean 3-yearly screening coverage of 80 percent among 25-69 year old women.
Results
A more intensive screening strategy involving screening every year is comparable to the effectiveness of HPV 16/18 vaccination when compared to baseline screening. This increase in effectiveness comes with a high additional estimated health sector cost of NOK 9,5 billion. We compared baseline screening to the strategy with the smallest effect. This combined strategy
seems to be cost saving compared to baseline screening. Cost effectiveness of HPV 16/18 vaccination was relatively insensitive to reduced screening compliance among vaccinated women.
Conclusion
Cytological screening is effective in reducing the burden of cervical cancer, and continues to be important also after implementing HPV 16/18 vaccination. The intervention most likely to be selected given this evidence, is vaccination with increased screening intervals after 20 years. Cross protection and (slowly) waning vaccine efficacy is found to have minor positive and negative effects respectively on future incidence rates of cervical cancer and cost effectiveness of the vaccine. A potential caveat is strain replacement of HPV 16 and 18 by other oncogenic HPV types, which may reduce the impact of HPV 16/18 vaccination significantly in a 50 year perspective.
Summary
Background
Whether HPV16/18 vaccination should be introduced into the Norwegian childhood vaccination programme is currently being discussed. The National Council on Quality and Prioritisation in Health Care discussed the matter in December 2007, and asked the Norwegian Knowledge Centre for the Health Services to use a dynamic health economic model to quantify how several unresolved issues regarding long term efficacy may affect the future burden of cervical cancer and health economics of HPV 16/18 vaccination.
Method
We used the same dynamic natural history model and health economic approach as previously described in Report No. 12-2007. The analyses are made in a two step approach: i) a natural history model is used to estimate the development in pre-cancerous and cancerous lesions in the time period 2008-2060, ii) health economic evaluations are performed based on the predicted burden of disease. Outcomes are compared to a base case vaccination screening scenario assuming 90 percent vaccine coverage among 12-year old girls, 90 percent vaccine efficacy and a mean 3-yearly screening coverage of 80 percent among 25-69 year old women.
Results
The potential disease impact and cost-effectiveness of HPV 16/18 vaccination as a result of performing a series of additional scenarios to accompany Report No. 12-2007 are presented in this section.
Screening once a year without HPV 16/18 vaccination
A more intensive screening strategy involving screening every year is comparable to the effectiveness of HPV 16/18 vaccination (LYG and QALYs) when compared to baseline screening. This increase in effectiveness comes with a high additional estimated health sector cost of NOK 9,5 billion. When incremental analysis is performed between vaccination screening and the improved screening program, it turns out that the latter is dominated by the vaccination based strategy. From a health economic point of view, this means that the more intensive screening programme should not be further considered for implementation since the health improvements can be achieved with vaccination for only one ninth of the additional costs of screening.
Screening 5-yearly from 2028, 2038 or 2048 together with HPV 16/18 vaccination
We first compared baseline screening to the strategy with the smallest effect (vaccination with 5-yearly screening after 20 years). This combined strategy seems to be cost-saving compared to baseline screening. If the 5-yearly screening is assumed to be delayed to 2038 or 2048, this would give ICERs of respectively 582 000 and 850 000 NOK per QALY gained from the societal perspective. Finally, moving from a strategy with vaccination and 5-yearly screening in 2048 to vaccination and screening every 3 years for the entire time horizon 2008-2060 would create an ICER of 2.3 million NOK per QALY gained.
Reduced screening compliance among vaccinated women
Cost effectiveness of HPV 16/18 vaccination was relatively insensitive to reduced screening compliance among vaccinated women.
Booster vaccination
The incremental information shows that waning is dominated by booster because it averts less QALY loss at the same time as costing more compared to booster. Among booster options 50 percent coverage is more likely to be optimal since the ICER of 80 percent is as high as 1.4 million NOK per QALY from the societal perspective.
Strain replacement
Cost-effectiveness of HPV 16/18 vaccination was sensitive to strain replacement, both at a 50 percent and at a 55 percent rate. In comparison with the baseline vaccination screening scenario, the ICERs increased almost 5 times with a 55 percent rate of strain replacement (societal perspective).
Cross protection and strain replacement
Cost effectiveness was relatively insensitive to crossprotection, but when a combined scenario of cross protection and strain replacement was modelled only some of the negative effect seen by replacement could be counteracted. In comparison with the baseline vaccination screening scenario, the ICERs still increased almost 4 times with a 55 percent rate of strain replacement combined with cross protection of 20 percent (societal perspective).
Conclusions
Cytological screening is effective in reducing the burden of cervical cancer, and continues to be important also after implementing HPV 16/18 vaccination. The intervention most likely to be selected given this evidence, is vaccination with increased screening intervals after 20 years.
Cross protection and (slowly) waning vaccine efficacy is found to have minor positive and negative effects respectively on future incidence rates of cervical cancer and cost effectiveness of the vaccine. A potential caveat is strain replacement of HPV 16 and 18 by other oncogenic HPV types, which may reduce the impact of HPV 16/18 vaccination significantly in a 50 year perspective.