Limits...
Economic evaluation of pneumococcal conjugate vaccination in The Gambia.

Kim SY, Lee G, Goldie SJ - BMC Infect. Dis. (2010)

Bottom Line: We extended the base-case results for PCV9 to estimate the cost-effectiveness of PCV7, PCV10, and PCV13, each compared to no vaccination.Under base-case assumptions ($3.5 per vaccine), compared to no intervention, a PCV9 vaccination program would cost $670 per DALY averted in The Gambia.Assuming a cost-effectiveness threshold of three times GDP per capita, all PCVs examined would be cost-effective at the tentative Advance Market Commitment (AMC) price of $3.5 per dose.

View Article: PubMed Central - HTML - PubMed

Affiliation: Department of Health Policy and Management, Harvard School of Public Health, Center for Health Decision Science, Boston, MA, USA. sykim@hsph.harvard.edu

ABSTRACT

Background: Gambia is the second GAVI support-eligible country to introduce the 7-valent pneumococcal conjugate vaccine (PCV7), but a country-specific cost-effectiveness analysis of the vaccine is not available. Our objective was to assess the potential impact of PCVs of different valences in The Gambia.

Methods: We synthesized the best available epidemiological and cost data using a state-transition model to simulate the natural histories of various pneumococcal diseases. For the base-case, we estimated incremental cost (in 2005 US dollars) per disability-adjusted life year (DALY) averted under routine vaccination using PCV9 compared to no vaccination. We extended the base-case results for PCV9 to estimate the cost-effectiveness of PCV7, PCV10, and PCV13, each compared to no vaccination. To explore parameter uncertainty, we performed both deterministic and probabilistic sensitivity analyses. We also explored the impact of vaccine efficacy waning, herd immunity, and serotype replacement, as a part of the uncertainty analyses, by assuming alternative scenarios and extrapolating empirical results from different settings.

Results: Assuming 90% coverage, a program using a 9-valent PCV (PCV9) would prevent approximately 630 hospitalizations, 40 deaths, and 1000 DALYs, over the first 5 years of life of a birth cohort. Under base-case assumptions ($3.5 per vaccine), compared to no intervention, a PCV9 vaccination program would cost $670 per DALY averted in The Gambia. The corresponding values for PCV7, PCV10, and PCV13 were $910, $670, and $570 per DALY averted, respectively. Sensitivity analyses that explored the implications of the uncertain key parameters showed that model outcomes were most sensitive to vaccine price per dose, discount rate, case-fatality rate of primary endpoint pneumonia, and vaccine efficacy against primary endpoint pneumonia.

Conclusions: Based on the information available now, infant PCV vaccination would be expected to reduce pneumococcal diseases caused by S. pneumoniae in The Gambia. Assuming a cost-effectiveness threshold of three times GDP per capita, all PCVs examined would be cost-effective at the tentative Advance Market Commitment (AMC) price of $3.5 per dose. Because the cost-effectiveness of a PCV program could be affected by potential serotype replacement or herd immunity effects that may not be known until after a large scale introduction, type-specific surveillance and iterative evaluation will be critical.

Show MeSH

Related in: MedlinePlus

Model schematic. This figure presents the schematic of the Markov model. Solid arrows represent transition probabilities between health states, which are differentiated depending on immunization status. Dashed arrows represent disease-specific deaths. Curved arrows represent that individuals stay in the same health states during the next cycle. Deaths due to other causes occur at every stage according to the background mortality rate for Gambian children but are not shown.
© Copyright Policy - open-access
Related In: Results  -  Collection

License
getmorefigures.php?uid=PMC2944347&req=5

Figure 1: Model schematic. This figure presents the schematic of the Markov model. Solid arrows represent transition probabilities between health states, which are differentiated depending on immunization status. Dashed arrows represent disease-specific deaths. Curved arrows represent that individuals stay in the same health states during the next cycle. Deaths due to other causes occur at every stage according to the background mortality rate for Gambian children but are not shown.

Mentions: The model was structured as a static, aggregate-level, state-transition model with a cycle length of one month (TreeAge Pro 2008). The model simulates the natural history of three different diseases--pneumonia, meningitis, and sepsis--known to be caused by S. pneumoniae. For pneumonia, we dichotomized all cases into 'primary endpoint' and 'non-primary endpoint' pneumonias based on the standards recommended by WHO [25,26]. While the present study is concerned with pneumonia due to S. pneumoniae, the most comprehensive data available for local incidence of pneumonia measure incidence of all-cause pneumonia, largely since etiology can be confirmed only in very few cases. We therefore chose to use a "vaccine probe" approach in a broad sense [31]; that is, we simulate the occurrence of all-cause pneumonia of both primary and non-primary endpoints for both vaccination and no vaccination strategies, assuming that the net difference in the pneumonia disease burden reflects the burden caused by all serotypes of S. pneumoniae that is averted under a PCV intervention. For meningitis and sepsis, we modeled disease events by S. pneumoniae (all-serotype) only. We applied the model to a hypothetical Gambian birth cohort (N = 60,000) and estimated the clinical and economic consequences of a PCV intervention over the first 5 years of life, based on the following assumptions: (1) repeat infection by S. pneumoniae of the same serotype would not occur over the 5-year time horizon (mainly due to insufficient data on partial immunity following natural infection), although diseases may re-occur due to different pneumococcal serotypes or different etiologic agents (e.g., pneumonia due to Hib or meningococcal meningitis); (2) pneumonia might be treated in either inpatient or outpatient settings depending on severity, but, given the general severity of the diseases, all meningitis and sepsis cases would require hospitalization; and (3) a portion of pneumococcal meningitis patients would experience neurological sequelae such as vision loss, hearing loss, motor delay, and seizures upon recovery from the acute illness [32]. Figure 1 shows the schematic of the model.


Economic evaluation of pneumococcal conjugate vaccination in The Gambia.

Kim SY, Lee G, Goldie SJ - BMC Infect. Dis. (2010)

Model schematic. This figure presents the schematic of the Markov model. Solid arrows represent transition probabilities between health states, which are differentiated depending on immunization status. Dashed arrows represent disease-specific deaths. Curved arrows represent that individuals stay in the same health states during the next cycle. Deaths due to other causes occur at every stage according to the background mortality rate for Gambian children but are not shown.
© Copyright Policy - open-access
Related In: Results  -  Collection

License
Show All Figures
getmorefigures.php?uid=PMC2944347&req=5

Figure 1: Model schematic. This figure presents the schematic of the Markov model. Solid arrows represent transition probabilities between health states, which are differentiated depending on immunization status. Dashed arrows represent disease-specific deaths. Curved arrows represent that individuals stay in the same health states during the next cycle. Deaths due to other causes occur at every stage according to the background mortality rate for Gambian children but are not shown.
Mentions: The model was structured as a static, aggregate-level, state-transition model with a cycle length of one month (TreeAge Pro 2008). The model simulates the natural history of three different diseases--pneumonia, meningitis, and sepsis--known to be caused by S. pneumoniae. For pneumonia, we dichotomized all cases into 'primary endpoint' and 'non-primary endpoint' pneumonias based on the standards recommended by WHO [25,26]. While the present study is concerned with pneumonia due to S. pneumoniae, the most comprehensive data available for local incidence of pneumonia measure incidence of all-cause pneumonia, largely since etiology can be confirmed only in very few cases. We therefore chose to use a "vaccine probe" approach in a broad sense [31]; that is, we simulate the occurrence of all-cause pneumonia of both primary and non-primary endpoints for both vaccination and no vaccination strategies, assuming that the net difference in the pneumonia disease burden reflects the burden caused by all serotypes of S. pneumoniae that is averted under a PCV intervention. For meningitis and sepsis, we modeled disease events by S. pneumoniae (all-serotype) only. We applied the model to a hypothetical Gambian birth cohort (N = 60,000) and estimated the clinical and economic consequences of a PCV intervention over the first 5 years of life, based on the following assumptions: (1) repeat infection by S. pneumoniae of the same serotype would not occur over the 5-year time horizon (mainly due to insufficient data on partial immunity following natural infection), although diseases may re-occur due to different pneumococcal serotypes or different etiologic agents (e.g., pneumonia due to Hib or meningococcal meningitis); (2) pneumonia might be treated in either inpatient or outpatient settings depending on severity, but, given the general severity of the diseases, all meningitis and sepsis cases would require hospitalization; and (3) a portion of pneumococcal meningitis patients would experience neurological sequelae such as vision loss, hearing loss, motor delay, and seizures upon recovery from the acute illness [32]. Figure 1 shows the schematic of the model.

Bottom Line: We extended the base-case results for PCV9 to estimate the cost-effectiveness of PCV7, PCV10, and PCV13, each compared to no vaccination.Under base-case assumptions ($3.5 per vaccine), compared to no intervention, a PCV9 vaccination program would cost $670 per DALY averted in The Gambia.Assuming a cost-effectiveness threshold of three times GDP per capita, all PCVs examined would be cost-effective at the tentative Advance Market Commitment (AMC) price of $3.5 per dose.

View Article: PubMed Central - HTML - PubMed

Affiliation: Department of Health Policy and Management, Harvard School of Public Health, Center for Health Decision Science, Boston, MA, USA. sykim@hsph.harvard.edu

ABSTRACT

Background: Gambia is the second GAVI support-eligible country to introduce the 7-valent pneumococcal conjugate vaccine (PCV7), but a country-specific cost-effectiveness analysis of the vaccine is not available. Our objective was to assess the potential impact of PCVs of different valences in The Gambia.

Methods: We synthesized the best available epidemiological and cost data using a state-transition model to simulate the natural histories of various pneumococcal diseases. For the base-case, we estimated incremental cost (in 2005 US dollars) per disability-adjusted life year (DALY) averted under routine vaccination using PCV9 compared to no vaccination. We extended the base-case results for PCV9 to estimate the cost-effectiveness of PCV7, PCV10, and PCV13, each compared to no vaccination. To explore parameter uncertainty, we performed both deterministic and probabilistic sensitivity analyses. We also explored the impact of vaccine efficacy waning, herd immunity, and serotype replacement, as a part of the uncertainty analyses, by assuming alternative scenarios and extrapolating empirical results from different settings.

Results: Assuming 90% coverage, a program using a 9-valent PCV (PCV9) would prevent approximately 630 hospitalizations, 40 deaths, and 1000 DALYs, over the first 5 years of life of a birth cohort. Under base-case assumptions ($3.5 per vaccine), compared to no intervention, a PCV9 vaccination program would cost $670 per DALY averted in The Gambia. The corresponding values for PCV7, PCV10, and PCV13 were $910, $670, and $570 per DALY averted, respectively. Sensitivity analyses that explored the implications of the uncertain key parameters showed that model outcomes were most sensitive to vaccine price per dose, discount rate, case-fatality rate of primary endpoint pneumonia, and vaccine efficacy against primary endpoint pneumonia.

Conclusions: Based on the information available now, infant PCV vaccination would be expected to reduce pneumococcal diseases caused by S. pneumoniae in The Gambia. Assuming a cost-effectiveness threshold of three times GDP per capita, all PCVs examined would be cost-effective at the tentative Advance Market Commitment (AMC) price of $3.5 per dose. Because the cost-effectiveness of a PCV program could be affected by potential serotype replacement or herd immunity effects that may not be known until after a large scale introduction, type-specific surveillance and iterative evaluation will be critical.

Show MeSH
Related in: MedlinePlus