To arrive at an estimate of the eligible population size, we recommend a process of
“funneling down” from the total population in the jurisdiction of interest to those eligible for the new drug. This funnel-down approach may be accomplished through a series of calculations using data such as the following:
1. Total jurisdiction population size
2. Age and sex distribution in the population
3. Annual age- and/or sex-specific incidence for an acute condition or age- and/or sex-specific incidence and prevalence for a chronic condition
4. Percentage of the incident or prevalent population with a diagnosis of the condi- tion and who are under a physician’s care
5. Percentage of the incident or prevalent population with a diagnosis of the condi- tion and who are under a physician’s care who are included in the marketing indication for the new drug (eligible population)
6. Percentage of the eligible population who are not restricted for reimbursement of the new drug by additional criteria imposed by the reimbursement decision maker or health plan such as failure of previous treatments or prior authorization (reimbursement-eligible population)
Using these data, typically derived from a mixture of published, Internet, and other data sources, and starting from the jurisdiction or health plan total population, we can estimate the size of the incident and prevalent populations that might need to be included in the budget-impact analysis.
Although both incident and prevalent populations should be considered in a budget- impact analysis for a chronic condition, it is sometimes possible to estimate the eligible population size by combining these two subgroups into a single estimate of the annual treated prevalence. By doing this, the modeler is making the assump- tion that the treatment efficacy and mix will be the same for both the incident and prevalent populations.
In Box 3.2, we present two examples of the funnel-down approach. The example for asthma illustrates an approach combining the incident and prevalent populations into a single estimate of annual treated prevalence, whereas the example for HIV infection illustrates the approach when estimating the size of the incident and preva- lent populations separately.
Box 3.2 Funneling Down to Estimate Eligible Population Size
Identifying Patients Within a Health Plan Who May be Eligible for a First-line Inhaled Corticosteroid
Assume that a new inhaled corticosteroid has been approved in the USA for adolescents and adults (≥12 years of age) for prophylactic, maintenance treatment in asthma. Health plans want to understand the potential impact that including this new asthma drug on the formulary will have on their budgets.
We can create the funnel-down approach for identifying patients who would
S. Earnshaw and J. Mauskopf
currently be eligible for treatment with this new drug out of the total health plan population.
1. We start with the total number of people within the health plan.
2. Since the drug is approved for adolescents and adults, the proportion of the members of the health plan that are adolescents and adults is estimated.
3. From these individuals, those with asthma are identified based on esti- mates of the national or local prevalence of asthma by age group.
4. Among these individuals, we further identify those who are on any main- tenance controller treatment and those who are eligible for maintenance treatment with monotherapy with an inhaled corticosteroid depending on their disease severity using data from published studies or health plan data.
The calculation to derive the number of patients taking ICS monotherapy then is as follows:
Number of patients taking ICS monotherapy
= total health plan population × percentage of population who are ≥12 years of age
× percentage of patients who are ≥12 years of age with a diagnosis of asthma
× percentage of patients with asthma on any controller × percentage of patients with asthma on any controller who are taking an ICS as monotherapy
Identifying Patients Within a Health Plan Who May be Eligible for Salvage Treatment Regimens for HIV Infection.
Assume that an antiretroviral treatment regimen has been approved in the USA for treatment for those with HIV infection who are highly treatment experienced and for whom no fully suppressive treatment regimens are cur- rently available. Health plans want to understand the potential impact that
Total health plan population
Adolescents and adults (patients aged ≥ 12 years)
Diagnosis of asthma
Patients with asthma and on any controller
Patients taking ICS
mono- therapy
Funnel down to eligible asthma patients. ICS inhaled corticosteroid
40
including this new HIV treatment regimen on the formulary will have on their budgets. We can create the funnel-down approach for identifying patients who would be eligible for treatment with this new regimen out of the total health plan population (the prevalent population) as well as those who would become newly eligible each year of the analysis time horizon (incident population).
For those who would be eligible for the new regimen now (prevalent population):
1. We start with the total number of people within the health plan.
2. Since the treatment is approved for those with HIV infection, the propor- tion of the members of the health plan that are living with HIV infection is estimated based on local or national prevalence data.
3. From these individuals, the proportion with a diagnosis and who are treated with antiretroviral therapy is estimated based on national or local data.
4. Among diagnosed and treated individuals, we further identify those for whom three lines of treatment have failed and/or who have no fully suppres- sive regimens remaining. These estimates are based on published estimates of the prevalence of multiclass-resistant HIV from observational data cohorts or from estimates of the life expectancy after initiating first- line treatment and the average duration on the first three lines of treatment from modeled data.
For those who would become newly eligible each year (incident populations):
5. We divide those whose third regimen has failed and/or those who do not have a fully suppressive regimen (the prevalent population) by their average life expectancy to estimate the number of people who are newly eligible for the new regimen each year.
1 million members
Persons with HIV
Persons diagnosed and treated
Persons with no fully suppressive
regimen
Funnel down to HIV patients eligible for a new fully suppressive regimen. HIV human immunodeficiency virus
S. Earnshaw and J. Mauskopf