1. Costs and Expected Benefits of Investment in Cervical Cancer Prevention
Nicole G. Campos, PhD

2. Background
Despite the availability of both primary and secondary prevention, ~528,000 women worldwide developed cervical cancer in 2012.
The WHO recommends
HPV vaccination for girls aged 9-13 years, prior to initiation of sexual activity.
Screening programs prioritizing women aged years, with HPV testing when resources are available; for settings with fewer resources, visual inspection with acetic acid (VIA) is recommended.
While momentum for HPV vaccination programs is increasing, scale-up of screening programs in low- and lower-middle income countries has not kept pace.

3. Objective
To evaluate any policy intervention, decision makers require information on feasibility, acceptability, financing, and value for money.
We sought to estimate the health impact, financial costs, and cost-effectiveness of scaling up coverage of HPV vaccination (young girls) and cervical cancer screening (women of screening age) in countries that will likely need donor assistance.
At this critical juncture for cervical cancer prevention efforts…

4. Modeling Approach

5. I’ll summarize our methods here, but since time is brief I refer you to the special issue on cervical cancer prevention, which is available through open access at the link on the slide, for further details on our model-based approach.

6. Cervical Cancer Prevention Strategies
The Excel-based CERVIVAC model (developed for PAHO’s ProVac Initiative) was used to estimate the costs and lifetime health impact of
HPV-16/18 vaccination
2 doses
10 yo girls
Screening once in a lifetime
HPV DNA testing or VIA, with treatment if necessary
35 yo women

7. Target Population and Time Horizon
50 low- and lower-middle income (LI and LMI) countries
populations >1 million persons
Gross national income per capita < US$2,585
Intervention period:
Time horizon: lifetime of birth cohorts who received either HPV vaccination or screening based on age during 10-year intervention period ( )

8. Model Inputs Epidemiologic inputs Costs Technology effectiveness
GLOBOCAN 2012
Statistical analyses
Individual-based microsimulation model of HPV infection and cervical carcinogenesis
Costs
Literature review
Extrapolation methods
Technology effectiveness
Model inputs included epidemiologic data inputs, which were derived using GLOBOCAN 2012, as well as statistical methods and a peer-reviewed individual-based microsimulation model of HPV infection and cervical cancer.
Country-specific unit costs were based on published literature and extrapolation methods.
The effectiveness of HPV vaccination and screening were based on literature reviews and the microsimulation model of cervical cancer.

9. Scale-Up Scenarios

10. Regarding scale-up scenarios, we assumed vaccination coverage in both low- and lower-middle-income countries would roll out with linear increases of 10% per year during the 10-year intervention period, beginning at 10% coverage in 2017 and reaching 100% coverage in 2026.

11. For screening coverage in low-income countries, we assumed that screening with VIA (depicted by the blue bars) would roll out with linear increases of 20% per year, beginning at 20% coverage and reaching 100% coverage in We then assumed a gradual replacement with HPV-based screening (depicted by the green bars), increasing linearly to cover 50% of the screening eligible population by 2026.

12. For screening coverage in lower-middle-income countries, we assumed that screening with both VIA and HPV testing would roll out with linear increases of 10% per year for each test, beginning at 20% total screening coverage in 2017 and reaching 100% coverage in We then assumed a gradual replacement with HPV-based screening (the green bars), which increased linearly to cover 100% of the screening-eligible population by 2026.

13. Model Outcomes Health Outcomes Cost Outcomes Efficiency (Value)
Cervical cancer cases
Cervical cancer deaths
Disability-adjusted life years (DALYs)
Costs (2013 US$)
Incremental cost- effectiveness ratios (ICER; US$ per DALY averted)*
Model health outcomes included…
Cost outcomes were in 2013 US$.
We considered incremental cost-effectiveness ratios, or ICERs, as the measure of efficiency, or value. ICERs were calculated separately for HPV vaccination and screening, relative to no intervention, and while there is no universal criterion that defines a threshold cost-effectiveness ratio, we consider the heuristic that an intervention with an ICER < per capita Gross Domestic Product is “very cost-effective.”
* ICERs were calculated separately for HPV vaccination and screening, relative to no intervention. While there is no universal criterion that defines a threshold cost-effectiveness ratio, we consider the heuristic that an intervention with an ICER < per capita GDP is “very cost-effective”.

14. Results

15. Costs, Health Outcomes, and Cost-Effectiveness of HPV Vaccination
Number eligible (millions)
Number reached (millions)
Cases averted (millions) U
DALYs averted (millions) D
Program cost (millions)
Program cost per case averted
Program cost per DALY averted
Vaccine LI
112 63
1.7
5.0
$768
$643
$443
$129
LMI
181
100
1.6
4.5
$1538
$1291
$984
$287
Total
293
163
3.3
9.5
$2306
$1934
$700
$204
This table displays selected model outcomes for HPV vaccination in the 50 countries, stratified by income tier.
An HPV vaccination program targeting 10-year old girls and scaling up at a rate of 10% per year would reach an estimated 163 million girls in the countries analyzed.
163 million girls reached
U: undiscounted
D: discounted at an annual rate of 3%.

16. Costs, Health Outcomes, and Cost-Effectiveness of HPV Vaccination
Number eligible (millions)
Number reached (millions)
Cases averted (millions) U
DALYs averted (millions) D
Program cost (millions)
Program cost per case averted
Program cost per DALY averted
Vaccine LI
112 63
1.7
5.0
$768
$643
$443
$129
LMI
181
100
1.6
4.5
$1538
$1291
$984
$287
Total
293
163
3.3
9.5
$2306
$1934
$700
$204
HPV vaccination during the 10-year intervention period would avert an estimated 3.3 million cervical cancer cases, and 9.5 million DALYs over the lifetimes of the vaccinated cohorts.
3.3 million cases averted
9.5 million DALYS averted
U: undiscounted
D: discounted at an annual rate of 3%.

17. Costs, Health Outcomes, and Cost-Effectiveness of HPV Vaccination
Number eligible (millions)
Number reached (millions)
Cases averted (millions) U
DALYs averted (millions) D
Program cost (millions)
Program cost per case averted
Program cost per DALY averted
Vaccine LI
112 63
1.7
5.0
$768
$643
$443
$129
LMI
181
100
1.6
4.5
$1538
$1291
$984
$287
Total
293
163
3.3
9.5
$2306
$1934
$700
$204
The estimated cost of the HPV vaccination program scaling up over 10 years would be $1.93 billion.
Program cost: $1.93 billion
U: undiscounted
D: discounted at an annual rate of 3%.

18. Costs, Health Outcomes, and Cost-Effectiveness of HPV Vaccination
Number eligible (millions)
Number reached (millions)
Cases averted (millions) U
DALYs averted (millions) D
Program cost (millions)
Program cost per case averted
Program cost per DALY averted
Vaccine LI
112 63
1.7
5.0
$768
$643
$443
$129
LMI
181
100
1.6
4.5
$1538
$1291
$984
$287
Total
293
163
3.3
9.5
$2306
$1934
$700
$204
The program cost per case averted was an estimated $700…
$700 per case averted
U: undiscounted
D: discounted at an annual rate of 3%.

19. Costs, Health Outcomes, and Cost-Effectiveness of HPV Vaccination
Number eligible (millions)
Number reached (millions)
Cases averted (millions) U
DALYs averted (millions) D
Program cost (millions)
Program cost per case averted
Program cost per DALY averted
Vaccine LI
112 63
1.7
5.0
$768
$643
$443
$129
LMI
181
100
1.6
4.5
$1538
$1291
$984
$287
Total
293
163
3.3
9.5
$2306
$1934
$700
$204
…while the program cost per DALY averted was $204.
This figure is comparable to the estimated cost per DALY averted of expanding eligibility for ART to all HIV+ adults in several lower-middle income countries.
$204 per DALY averted
U: undiscounted
D: discounted at an annual rate of 3%.

20. Costs, Health Outcomes, and Cost-Effectiveness of Cervical Cancer Screening
Number eligible (millions)
Number reached (millions)
Cases averted (millions) U
DALYs averted (millions) D
Program cost (millions)
Program cost per case averted
Program cost per DALY averted
Screening LI 61 50
0.6
3.9
$247
$207
$448
$52
LMI
148
120
1.3
8.6
$1265
$1072
$964
$124
Total
209
170
1.9
12.6
$1512
$1279
$811
$102
This table displays selected model outcomes for cervical cancer screening programs in the 50 countries, stratified by income tier.
A cervical cancer screening program reaching full scale within the first 5 years of the intervention period would reach an estimated 170 million women in the countries analyzed.
170 million women reached
U: undiscounted
D: discounted at an annual rate of 3%.

21. Costs, Health Outcomes, and Cost-Effectiveness of Cervical Cancer Screening
Number eligible (millions)
Number reached (millions)
Cases averted (millions) U
DALYs averted (millions) D
Program cost (millions)
Program cost per case averted
Program cost per DALY averted
Screening LI 61 50
0.6
3.9
$247
$207
$448
$52
LMI
148
120
1.3
8.6
$1265
$1072
$964
$124
Total
209
170
1.9
12.6
$1512
$1279
$811
$102
Screening efforts during the 10-year intervention period would avert an estimated 1.9 million cervical cancer cases, and 12.6 million DALYs over the lifetimes of the screened cohorts.
1.9 million cases averted
12.6 million DALYS averted
U: undiscounted
D: discounted at an annual rate of 3%.

22. Costs, Health Outcomes, and Cost-Effectiveness of Cervical Cancer Screening
Number eligible (millions)
Number reached (millions)
Cases averted (millions) U
DALYs averted (millions) D
Program cost (millions)
Program cost per case averted
Program cost per DALY averted
Screening LI 61 50
0.6
3.9
$247
$207
$448
$52
LMI
148
120
1.3
8.6
$1265
$1072
$964
$124
Total
209
170
1.9
12.6
$1512
$1279
$811
$102
The estimated cost of the screening program for women aged 35 years, reaching full coverage after 5 years, was $1.28 billion.
Program cost: $1.28 billion
U: undiscounted
D: discounted at an annual rate of 3%.

23. Costs, Health Outcomes, and Cost-Effectiveness of Cervical Cancer Screening
Number eligible (millions)
Number reached (millions)
Cases averted (millions) U
DALYs averted (millions) D
Program cost (millions)
Program cost per case averted
Program cost per DALY averted
Screening LI 61 50
0.6
3.9
$247
$207
$448
$52
LMI
148
120
1.3
8.6
$1265
$1072
$964
$124
Total
209
170
1.9
12.6
$1512
$1279
$811
$102
The program cost per case averted was an estimated $811…
$811 per case averted
U: undiscounted
D: discounted at an annual rate of 3%.

24. Costs, Health Outcomes, and Cost-Effectiveness of Cervical Cancer Screening
Number eligible (millions)
Number reached (millions)
Cases averted (millions) U
DALYs averted (millions) D
Program cost (millions)
Program cost per case averted
Program cost per DALY averted
Screening LI 61 50
0.6
3.9
$247
$207
$448
$52
LMI
148
120
1.3
8.6
$1265
$1072
$964
$124
Total
209
170
1.9
12.6
$1512
$1279
$811
$102
… while the program cost per DALY averted was $102.
Again, this estimate is comparable to the estimated cost per DALY averted of expanding eligibility for ART to all HIV+ adults in several lower-middle income countries.
$102 per DALY averted
U: undiscounted
D: discounted at an annual rate of 3%.

25. Costs, Health Outcomes, and Cost-Effectiveness of HPV Vaccination and Screening
Number eligible (millions)
Number reached (millions)
Cases averted (millions) U
DALYs averted (millions) D
Program cost (millions)
Program cost per case averted
Program cost per DALY averted
Combined LI
173
113
2.3
8..9
$1014
$850
$444
$95
LMI
329
220
2.9
13.1
$2803
$2363
$975
$180
Total
502
333
5.2
22.0
$3817
$3213
$740
$146
Combining these results, we see that a comprehensive cervical cancer prevention program involving scale-up of HPV vaccination and screening could avert 5.2 million cases and 22 million DALYs over the lifetimes of the intervention cohorts, for a total 10-year program cost of $3.2 billion.
5.2 million cases averted
22.0 million DALYs averted
Program cost: $3.2 billion
U: undiscounted
D: discounted at an annual rate of 3%.

26. Limitations
Our assumption of scale-up to 100% coverage of the target populations (with perfect compliance) is not realistic, but provides an upper bound on costs.
Due to lack of country-specific data in some settings, we relied on statistical techniques to extrapolate data from other settings.
We did not include the cost of building capacity to provide HPV vaccination and screening.
Cost-effectiveness analyses evaluating HPV vaccination and screening also need to be conducted at the country level for local decision-making.
As with any modeling analysis, this analysis had a number of limitations.

27. Cost-Effectiveness of Cervical Cancer Prevention in HIV+ Women
While the analysis that I have just described can provide important aggregated information for decision makers and donors who wish to address health disparities and efficiently allocate resources, it is also critical to estimate the comparative and cost-effectiveness of specific cervical cancer prevention strategies at the country level and in high-risk populations, particularly women with HIV.
We have published a number of cost-effectiveness studies on HPV vaccination and cervical cancer screening in low-resource settings, and have recently conducted an analysis of screening HIV-infected women in South Africa. Results are preliminary, but given the topic of this morning’s panel I wanted to highlight some of our preliminary findings.

28. Cost-Effectiveness Analysis: Preliminary Findings
Screening HIV+ Women in South Africa
A test-and-treat approach with HPV testing every 3 years is effective.
Cancer risk reduction: ~50%.
A test-and-treat approach with HPV testing every 3 years is very cost-effective.
ICER < per capita GDP
Price reductions in HPV tests and improving compliance with recommended treatment for screen-positive women would further improve health benefits and value.
Cost-effective: provides good value for public health dollars
We estimate an ICER value less than South Africa’s per capita GDP, and within the range of ICERs suggested by an economic evaluation of expanding ART access to all HIV+ adults in South Africa.

29. Key Findings
Both HPV-16/18 vaccination of young girls and once-in-a- lifetime screening and treatment of precancer for women of screening age are estimated to be very cost- effective.
While HPV vaccination may avert more cases and deaths than screening at comparable coverage levels, the benefits of vaccination will not be realized for decades to come.
Comprehensive cervical cancer prevention will require both HPV vaccination and screening.
Screening women with HIV appears to be very cost- effective.
To summarize, we are able to provide supporting evidence for several key insights.

30. Putting the Numbers in Context
A comprehensive cervical cancer prevention program in 50 LI and LMI countries could avert 5.2 million cases, 3.7 million deaths, and million DALYs for a total 10-year program cost of US$ 3.2 billion.
US$ 664 million per year at full-scale
In 2015, US$ 36.4 billion in development assistance for health was disbursed1:
US$ 10.8 billion for HIV/AIDS
US$ 6.5 billion for child and newborn health
US$ 3.6 billion for maternal health
The ICERs for HPV vaccination and cervical cancer screening programs (~$100-$200/DALY averted) are comparable to ICERs for expanding ART access to all HIV+ adults in several LMI countries2.
We estimate that a comprehensive cervical cancer prevention program in 50 LI and LMI could avert 5.2 million cases, 3.7 million deaths, and 22 million DALYs for a total 10-year program cost of $3.2 billion, or $664 million per year at full scale.
To put these numbers in context - In 2015, $36.4 billion in development assistance for health was disbursed. Of this, $10.8 billion was for HIV/AIDS, $6.5 billion was for child and newborn health, and $3.6 billion was for maternal health.
1 Dieleman JL et al. Lancet 2016;387:
2 Eaton JW et al. Lancet Glob Health 2014;2:e23-34.

31. Acknowledgements
Co-authors: Monisha Sharma, Andrew Clark, Kyueun Lee, Fangli Geng, Catherine Regan, Jane J. Kim, Stephen Resch.
This work was based on research funded by the American Cancer Society (16300). The findings and conclusions contained within are those of the authors and do not necessarily reflect positions or policies of the Bill & Melinda Gates Foundation. The funders had no role in study design or data collection, analysis, and interpretation; in the writing of the manuscript; or in the decision to submit the paper for publication.
We thank the American Cancer Society Cervical Cancer Global Cost of Action Advisory Group and staff and Cervical Cancer Action for helpful discussions on the scope of the research question and policy options to consider.
We gratefully acknowledge American Cancer Society for research funding that supported this work, and the Cervical Cancer Global Cost of Action Advisory Group and Cervical Cancer Action.