Required information

The estimated return on investment for an asthma care quality improvement program depends on user responses to three basic questions which are explained in detail below.

Who will be eligible for the asthma program?

The user must define who is going to be eligible to participate in the asthma program. The calculator asks the user to describe the population eligible for the asthma program in terms of insurance coverage, age group, asthma severity level and geographic location. The options available in each category follow:

• Type of insurance
• Medicaid
• State employee's health plan
• Employer-sponsored insurance
• Age group
• Children
• Adults
• Children and adults
• Asthma severity (see Table 1 for detailed definitions of asthma severity levels)
• All asthma
• Persistent asthma
• Persistent asthma with acute visit
• Geographic location
• 50 States and DC
• Nation

This information will be used to estimate the population of the targeted group and number of participants in the asthma care quality improvement program.

What types of studies should be used to estimate the impact of the asthma program?

The user needs to select how to estimate the impact of the asthma program. The calculator provides three options for the impact estimates, each of which is based on studies with a particular type of design. The options are randomized controlled studies, statistically controlled studies, and studies without a control group.

Fifty-two studies that examined asthma care improvement programs were analyzed to develop impact estimates. The studies selected for analysis were conducted in the United States and focused on pediatric or adult asthma patients under 65; the programs studied worked either directly with patients or indirectly with patients through their physician. Excluded from the analysis were evaluations conducted outside the United States, studies of patients age 65 and over, drug trials of asthma medications, and studies that reported only the impact on quality of life and knowledge of asthma.

For each study, the impact of the program was measured as the percentage change in total medical and productivity costs attributable to the program. Because few studies reported total costs, the impact of the program on each component of total cost was also measured.

For each result reported by the study, the following types of information were recorded:

• Annual visit rate or cost before and after the intervention for the treatment and control groups (if applicable).
• The sample size related to the results (e.g., number of inpatient stays, number of emergency department [ED] visits).
• Measures of statistical significance (confidence intervals, standard errors, and p-values).
• Study design (e.g., controlled or uncontrolled studies, length of evaluation).
• Patient characteristics (e.g., age group, insurance or program coverage, and severity of asthma).

The impact of the program was calculated in three different ways based on the data reported by each study. All three calculations were performed for studies that reported all necessary information.

• For randomized controlled studies, postintervention comparison of treatment and control was used. This method compares the treatment and control groups after the intervention. For example, if the control group had an average of three ED visits per year at the end of the study and the treatment group had an average of two visits per year at the end of the study, then the postintervention estimate of the impact of the program would be a -33% change in visits.
• For statistically controlled studies, net change was used. This method compares the change over time in each group. For example, the control group had the same visit rate as the treatment group (four per year) before the program. After the program, the control group had three visits per year and the treatment group had two. In this case, the change in the control group was -25% and the change in the treatment group was -50%, for a net effect of the program of -25%.
• For studies without a control group, pre- and postintervention comparison was used. This method estimates the percentage change in the treatment group before and after the program. For example, before the program, patients had on average of four ED visits per year and after they had two ED visits per year. Thus, the impact of the program would be a -50% change.

• Note: This last type of estimate without a control group does not account for changes outside the program ("secular" influences) that might have affected the treatment group results over time. This type of calculation is provided here only as interim information. Before a careful evaluation, this information may help program administrators know how their returns compare to other asthma quality improvement programs that did not account for "secular" changes. It may help administrators assess their programs' early results and change direction, if need be. If early results are discouraging compared to non-controlled studies, they may want to reconsider the program design. If early results are much better than non-controlled studies, they may want to expand their program to less severely ill subpopulations. The calculator enables program managers to explore what might happen if they include other subpopulations.

What costs/savings should be included?

The user must make choices regarding which costs and savings will be included in the calculation.

Type of cost: Asthma-related medical costs or all costs?

• Asthma-related medical costs will reflect the costs of visits with asthma as a primary diagnosis. All costs will reflect all the visits made by patients with asthma, whether or not the visits were for asthma care.

• The default option is asthma-related medical costs. This is the recommended option to use with the results from our literature review because most of the studies in the review measure only asthma-related visits. However, a user may want to select all costs if they have their own estimate of the impact of the asthma program that measures all visits made by patients with asthma, regardless of the diagnosis. In this case, the user would also change the impact estimates on the program impact page.

Cost perspective: Program/plan costs or program/plan and patient costs?

• The user must also select which cost perspective to use in estimating the ROI. The default is set to "Program/Plan & Patient," which means that costs will include payments made by the patient (including copayments and deductibles) and those made by the employer or Medicaid. This is appropriate if the user wants to estimate the total savings to society of the asthma program. However, if the user wants to estimate the financial impact on the Medicaid program only (or the private health plan only), then the user should select "Program/Plan."

Type of savings: Health care savings or health care savings and productivity gains?

• Finally, the user needs to decide on the type of savings to be calculated. The user can choose to include productivity gains in the calculation of the Net Present Value (NPV) and ROI in addition to health care savings. Another option is to include only health care savings. If the user wants to take a societal point of view, then productivity can be included.

Once the features of the program have been defined, the user is asked to review and edit the program features and data.

Population data

The user should review the demographic information for the population for which the asthma care quality improvement program is being designed. The user should review:

• Size of population
• Age and gender distribution of population
• Race distribution of population (for Medicaid populations only)

The calculator contains default data on the characteristics of enrollees in each State Medicaid program, enrollees in State employee health plans, and the population in each State with employer-sponsored health insurance. It automatically populates itself with default demographic data based on the selections the user makes on the required information page. The user can accept these defaults or change the supplied estimates. For example, a commercial health plan could select employer-sponsored health insurance and then change the age-gender profile to match enrollees in the plan.

Participant data

The user needs to review the information on the estimated number of participants in the asthma care quality improvement program. The number of participants is based on the estimated number of asthma patients eligible for the program and the estimated percentage of eligible patients who will participate in the program. The user should review:

Asthma patients eligible for program

• The calculator automatically determines the number of asthma patients eligible for the program based on the demographics of and asthma prevalence rates for the targeted population. Prevalence rates, which are stored in the calculator, are chosen based on the selections for type of asthma and population that the user makes on the required information page.

• The prevalence rates for Medicaid are calculated using the MarketScan Medicaid Database 2004, which includes medical claims and enrollment data from millions of Medicaid beneficiaries in eight States. For Medicaid, calculated age-, gender-, and race-specific asthma prevalence rates are provided. The average prevalence rates from the eight States are used and are assumed to approximate asthma prevalence for other State Medicaid program populations. State Medicaid agencies would have more accurate calculator results for their State if they entered their own prevalence data.

• The prevalence rates for employer-sponsored health insurance and State employees are calculated using the MarketScan Commercial Claims Database 2004, which includes medical claims and enrollment data from millions of employees and their dependents. We calculated age-, gender-, and State-specific prevalence rates from the MarketScan data so that the calculator can more accurately predict the number of patients with asthma based on information about the age-gender mix and State where the asthma program will take place. (We could not adjust for race/ethnicity because these elements were not available in the MarketScan Commercial Database.)

• If the user knows the number of asthma patients eligible for the program, it can be entered at this point.

Percentage of eligibles who will participate

• The percentage of eligibles who will participate in the program is set at 25 percent but can be changed by the user.

Expected patients who will participate

• The expected number of participants cannot be changed by the user, except by changing the prior items: percentage of eligibles who will participate or the number of asthma patients eligible for the program.

Baseline data

Next, the user needs to review and, if desired, change the baseline cost and utilization data which appear on the baseline data page. This page displays average annual medical care utilization, productivity, and costs per eligible patient with asthma. If the user has data specific to his or her population of interest, he or she should input the data here. Otherwise, the baseline data on health care utilization and costs are gathered from the MarketScan Databases.

If the user selects Medicaid on the required information page, utilization and cost data are filled in that represent the average experience of patients in the eight States that contribute data to MarketScan, adjusted for the age-gender-race characteristics shown on the population page. The utilization and cost data also come from the patients who meet the asthma severity selected on the required information page. For example, if user selects persistent asthma, he or she will see a higher ED visit rate on the baseline page than if they select all asthma, because patients with persistent asthma are seen in the ED more often. Similarly, if the user selects Employer-Sponsored Insurance or State Employee Health Plan coverage on the required information page, utilization and cost data are shown that represent the average experience of commercial patients from MarketScan.

The table below defines each field on the baseline data page.

Program impact

The user has the opportunity to review the program impact estimates presented on this page in terms of percentage reduction or increase in utilization and costs due to asthma for the targeted participants. The program impact is based on analysis of 52 published studies. The calculator populates this page with impact estimates based on the type of studies that the user selected on the required information page. If the user has custom impact estimates to enter, he or she should do so on this page.

Program cost

On the program cost page, the user needs to review four items, all of which the user can modify.

Length of operation planned for the program

• The user can choose the time horizon of the program and of the evaluation of benefits and costs to be performed by the calculator. The default is 3 years, which means that the costs and benefits of the program will be evaluated for 3 years. This time period is appropriate for a program that is planned to last 3 years and then will be reevaluated to determine if it will be continued. The time horizon can be adjusted to any integer from 1 to 10.

Time until the full impact of the program is expected

• The user also decides the number of years (1 through 10) until the program achieves full impact. If a 1 is in the field, then year 0 (2007) has the baseline outcome. The first year (2008) has the full impact of the program. For example, if the program reduced ED visits by 40 percent, then ED visits in 2008 will be 40 percent of the rate in 2007. If the model is set so that it takes 2 years for the program to achieve full impact, then half of the reduction will be experienced in year 1 (2008) and the full impact will be achieved in year 2. This can be seen if the user selects "Detailed Results" on the results page.

Cost of the program per person per year (consider entering your own data)

• The average cost per patient per year of asthma disease management programs that reported this information in the literature was just under $400, in inflated 2008 dollars. This is the calculator's default setting for program cost. Because the estimates varied widely, users should consider using their own costs for this component of the calculator.

• Enter your own program costs in place of the default value. The cost can be entered as cost per patient per year or total program cost per year. Total program cost per year can be estimated for all people in the program for the time horizon of the program and then divided by the number of years of the program. To calculate program cost, sum all costs related to staffing, equipment, and materials, assuming that capital expenditures (such as buildings) are not needed to implement the program. For staffing, include design of the intervention, recruitment of participants (patients, physicians, and facilities, if relevant), interactions with program participants (throughout the program), and dissemination of results. These factors also can be estimated by considering the number of full-time employees who will be working on the program, multiplied by their annual salaries. For equipment and supplies, include educational materials, any medications (if provided as part of the program and not purchased out of patients' pockets or by third-party purchasers), equipment for monitoring (if relevant), and any other expenses related to the program.

• If the program is to be purchased through a vendor, the user can consult vendors for asthma programs to obtain information on the annual cost of a program.

Discount rate for valuing savings and costs that occur in different years

• The user can also specify the discount rate. A 3 percent discount rate means that $100 saved next year is worth $97 this year and $100 dollars saved 2 years from now is worth $94 this year. This allows savings and costs realized over multiple years to be valued in the present, as the net present value. Essentially, the sooner the savings, the more valuable is the return from the program. If immediacy of results is essential to your program, you can raise the discount rate.

Results

On the results page, the calculator displays the impact of the program on health care costs and productivity costs, depending on the selections the user made on the required information page. Savings are shown as positive numbers and costs are shown as negative numbers. The results page shows the overall impact of the program in terms of the following:

• NPV shows the net savings due to lower medical expenditures and/or fewer missed days minus the cost to implement the program, all in present value terms. Present value discounts future dollars by the discount rate and sums up the total discounted amounts for each of the years in the time horizon. The NPV is the difference between the present value of savings and the present value of costs.

• The estimated expenditures per person at baseline before program implementation are subtracted from the estimated expenditures per person under the asthma intervention program, as specified by the user, to estimate the savings (assuming expenses are lower) per person resulting from the use of the asthma management program. These per participant savings are multiplied by the estimated number of people with asthma expected to participate in the program to obtain total program savings. The total savings of the program is then divided by the total cost of the program to calculate the return on investment.

• Break-even cost per participant is the point where program costs exactly equal program savings.

• You can save your results from each run of the calculator to an Excel file. For each scenario you save, two worksheets are added to the file. One contains the data you entered into the calculator (or the defaults you accepted) and the other contains the results of running the data through the calculator. When you are ready to download the Excel file, select the button labeled Download Saved Scenarios and open and save the file.

Patient selection criteria for each asthma definition in medical claims data

Prescription asthma medications^a

a. Asthma medications from: National Asthma Education and Prevention Program. Expert panel report 2: guidelines for the diagnosis and management of asthma - update on selected topics 2002. Bethesda, MD: National Insitutes of Health; June 2003. NIH Publication No. 02-5074.