WO2001009758A2

WO2001009758A2 - Computer-implemented methods and systems for estimating costs and benefits associated with disease management interventions

Info

Publication number: WO2001009758A2
Application number: PCT/US2000/021106
Authority: WO
Inventors: Zeba Mohammad Khan; Peggy Sue Olson; Teresa Lynn Young; Michael T. Halpern
Original assignee: Glaxo Group Limited
Priority date: 1999-08-02
Filing date: 2000-08-02
Publication date: 2001-02-08
Also published as: EP1198755A2; WO2001009758A8; AU6397100A; JP2003522994A

Abstract

Computer-implemented methods and systems for estimating costs and benefits associated with disease management interventions provide a graphical user interface through which a user inputs population-specific data for a population to be treated or offered treatment in a disease management intervention. Costs and benefits associated with the disease management intervention are projected over a time period for the population based on the user-specified data. The costs and benefits are output to the user to facilitate evaluation of the economic impact of the disease management intervention.

Description

COMPUTER-IMPLEMENTED METHODS AND SYSTEMS FOR

ESTIMATING COSTS AND BENEFITS ASSOCIATED WITH DISEASE

MANAGEMENT INTERVENTIONS

Technical Field The present invention relates generally to computer-implemented methods and systems for evaluating the costs and benefits of disease management interventions. More particularly, the present invention relates to computer-implemented methods and systems estimating direct medical and productivity costs and cost savings related to disease management interventions in order to facilitate user-evaluation of the economic impact of disease management interventions.

Related Art

In determining whether to implement a disease management intervention, such as a smoking cessation program or an influenza vaccination program, it is desirable for an employer or a health plan provider to have an accurate and efficient method for evaluating the economic impact of such a program. For example, in de'ciding whether to implement a smoking cessation program for employees, an employer may wish to evaluate the direct medical costs, productivity costs, and cost savings of such a program. Similarly, a health plan provider may wish to evaluate the direct medical costs and cost savings associated with a disease management intervention. Evaluating the costs and benefits of a disease management intervention may include complex calculations based on actuarial data. Due to the complexity of these calculations, employers and health plan providers are typically incapable of performing these calculations and accurately assessing the costs and benefits of a disease management intervention. Another scenario in which it may be desirable to have an efficient method of calculating the costs and benefits of a disease management intervention is in marketing such a program or product associated with the intervention to an employer or a health plan provider. For example, a pharmaceuticals manufacturer may wish to market a drug associated with a disease management intervention to employers or health plan providers. As part of the marketing campaign, the manufacturer may wish to present a cost- benefit analysis relating to the drug to the employer or health plan provider. This analysis may include calculating incremental costs and benefits of the intervention over a period of time using employer- or health plan-specific data. These calculations may be complex and typically require the hiring of an actuarial consulting firm. Hiring an actuarial consulting firm may be expensive and impractical because the firm would be required to be hired to perform the cost-benefit analysis for each employer or health plan provider.

In addition to the complexity of the calculations associated with evaluating a disease management program and the necessity of repeating the calculations for each employer or health plan provider, another problem is that conventional economic models created by actuarial firms may not indicate source data for performing the calculations associated with the models. For example, when a model is based on actuarial data, the actuarial data may be held as proprietary by the actuarial consulting firm. As a result, employers or health plan providers may be unable to evaluate the credibility of the calculations.

Accordingly, there exists a long-felt need for computer-implemented methods and systems for efficiently estimating the costs and benefits of disease management interventions in order to facilitate user-evaluation of the economic impact of disease management interventions.

Disclosure of the Invention The present invention provides user-friendly graphics-based computer software that allows employers or decision makers in a managed care organization to quickly and easily view and explore the impact of a disease management intervention on health and economic outcomes over designated periods of time. In one implementation, the software determines cost savings for a cohort of employees or health plan members from the start of the model through a specified age, such as retirement age (such as age 65), or death. As used herein, the term "model" refers to an economic model, including 5 mathematical algorithms and assumptions regarding a population, used to evaluate a disease management intervention. The model may be implemented as a set of computer-executable instructions, such as a program. The program may be embodied in a computer-readable medium, such as a magnetic disk, an optical disk, or a tape storage device. The program provides default values0 for the personnel and intervention characteristics, but allows users to modify these characteristics.

The program presents a variety of results, including the number of cases of illness, medical costs, and non-medical costs incurred over time with and without health-plan coverage of pharmaceutical products associated with the5 disease management intervention. Results are presented graphically and in numerical tables to facilitate interpretation. Results can be viewed on the computer screen or printed. According to an important aspect, the program displays data sources to the user and provides information about the program terminology and citations for the data sources. o The program also calculates and presents cost savings data to the user.

As used herein, the phrase "cost savings data" refers to data that indicates the amount of money saved over time as a result of implementing the disease management intervention. Cost savings data may include direct medical costs saved and indirect costs saved. 5 Accordingly, it is an object of the present invention to provide computer- implemented methods and systems for estimating costs and benefits associated with disease management interventions that facilitate user- evaluation of the economic impact of a disease management intervention. It is another object of the present invention to provide a graphics-based o user interface that allows users to input data specific to an employee or health plan population being treated in the disease management intervention. It is yet another object of the present invention to provide a computer interface for displaying source information associated with a disease management intervention to a user.

Some of the objects of the invention having been stated hereinabove, other objects will become evident as the description proceeds, when taken in connection with the accompanying drawings as best described hereinbelow.

Brief Description of the Drawings Figure 1 is a flow chart illustrating exemplary steps performed by a computer-implemented system for estimating costs and benefits associated with a disease management intervention according to an embodiment of the present invention;

Figure 2 is a block diagram illustrating an exemplary hierarchy of screens or windows that may be displayed on a computer display device to allow a user to enter data and view output relating to a disease management intervention;

Figure 3 illustrates an initial window of a computer-implemented system for estimating costs and benefits associated with a disease management intervention according to an embodiment of the present invention; Figure 4 illustrates a parameters window that allows a user to enter parameters associated with a disease management intervention;

Figure 5(A) illustrates an age/gender folder tab of a population demographics window according to an embodiment of the present invention;

Figure 5(B) illustrates an occupations folder tab of a population demographics window according to an embodiment of the present invention;

Figure 6 illustrates an additional model parameters window according to an embodiment of the present invention;

Figure 7 illustrates an intervention parameters window according to an embodiment of the present invention; Figure 8 illustrates a recidivism rates window according to an embodiment of the present invention; Figure 9 illustrates a disease-specific costs window according to an embodiment of the present invention;

Figure 10 illustrates a smoking status distributions window according to an embodiment of the present invention; Figure 11 (A) illustrates a cost savings graph window according to an embodiment of the present invention;

Figure 11(B) illustrates a disease savings graph window according to an embodiment of the present invention;

Figure 11 (C) illustrates an outcomes table window according to an embodiment of the present invention;

Figure 12 illustrates a benefit-cost analysis window according to an embodiment of the present invention;

Figure 13(A) illustrates a break-even graph window according to an embodiment of the present invention; Figure 13(B) illustrates a break-even analysis table window according to an embodiment of the present invention; and

Figure 14 illustrates a data sources window according to an embodiment of the present invention.

Detailed Description of the Invention

Figure 1 is a flow chart illustrating exemplary steps performed by a computer-implemented system for estimating costs and benefits associated with a disease management intervention according to an embodiment of the present invention. In step ST1, the user defines a model population for the disease management intervention. Defining the model population may include inputting, into a computer, the number of employees or health plan members, and characteristics associated with the employees or health plan members. The data may be input using any suitable method, for example, using a keyboard, a mouse, touch screen, microphone, or other suitable input device. In step ST2, the user specifies characteristics of the intervention. Exemplary characteristics that may be specified include costs of pharmaceutical products and the level of counseling associated with the intervention. In step ST3, the program projects costs and health outcomes for the population with and without the intervention. The projection is based on data from known populations extracted from published studies and actuarial databases. The calculations and assumptions involved in performing the projections are discussed in more detail below.

In step ST4, the program presents output on a computer display device indicating cost savings associated with the disease management intervention and data sources for the calculations performed by the program. The output is preferably displayed in a manner illustrating the costs and benefits of a disease management intervention. For example, graphs, tables, or any other type of data presentation format may be used to illustrate the costs and benefits of the disease management intervention. Because the user can enter population-specific data into the program, the program projects cost savings for the user-specified population. The data sources are viewable, and the program provides an effective and re-usable method for evaluating a disease management intervention.

Each of the steps performed by the program will be discussed in more detail below. The steps are discussed in the context of a smoking cessation program. However, the present invention is not limited to a smoking cessation program. For example, the steps illustrated in Figure 1 may be used to evaluate cost savings for any disease management intervention, including influenza vaccination programs, migraine headache programs, cardiac programs, diabetes programs, asthma programs, rhinitus programs, stroke programs, and irritable bowel programs.

Population Characteristics As stated above, the present embodiment preferably allows users to define the characteristics of the employed (or health plan) population. These characteristics may include the number of personnel by age and gender; the number (or proportion) of employees in each of several job classifications (e.g., clerical/administrative, labor, managerial); the type of industry (e.g., manufacturing, professional service, etc.); and a geographic region, such as a region or state of the U.S. (e.g., Northeast, Southeast, Midwest, West). Users are preferably required to specify their population size, industry or health plan type, and U.S. region. Default values preferably exist in the input fields for age, gender, and job classification breakdown of the workforce.

Based on the age and gender distribution of the model cohort, the type of organization, and the indicated geographic region, the model estimates the number of smokers in the workforce population. These calculations are preferably based on rates of smoking observed in population cohorts stratified by age group, gender, and health plan/job classification. While default values for the number and proportion of smokers in each age/gender group may exist in the model, users may input other values. The model provides default values for most inputs and can thus be run with minimal user input. However, all input values are preferably modifiable at the discretion of users based on the amount of information available.

Smoking Cessation Intervention After defining the model population, the user can specify the characteristics of the cessation intervention, such as the proportion of smokers attempting smoking cessation; the success rates of the intervention; the level of smoking cessation promotion; and the costs of the intervention. Default values may be included in the model, which are preferably modifiable by the user.

Model Outputs: Health Outcomes and Smoking-Related Costs Based on the characteristics of the model cohort and the cessation intervention, smoking-related costs and health outcomes for the population are projected with and without coverage of smoking cessation aids. The model preferably links the population characteristics with the incidence of smoking- related illnesses among current smokers versus former and never smokers over time. The model preferably provides specific information on common smoking-related illnesses, such as lung cancer, coronary heart disease, chronic obstructive pulmonary disease, cerebrovascular disease, and pregnancy complications, such as low birth weight and miscarriage. Rates of diseases among smokers and non-smokers over time are preferably determined using epidemiologic data on disease incidence from published studies and government reports. Model outcomes may include the number of cases of each of these conditions over time as well as the total number of cases and the change (decrease) with smoking cessation aid coverage under a health plan. In addition, the number of individuals successfully engaging in smoking cessation with and without coverage of cessation aids are presented. Users can view these health outcomes from the inception of the workforce or health plan cohort until a specified age, such as retirement age, a default age of 65, or until death.

Costs are presented as model outcomes. Cessation costs with and without aid coverage as well as medical care costs for overall healthcare and for each of the smoking-related conditions, and the change in costs associated with coverage of smoking cessation intervention are presented over time for the model population. These costs may be derived from information in published studies or analysis of government databases. In addition, for workforce cohorts, non-healthcare costs are projected over time for the modeled cohort with and without coverage of smoking cessation aids. These non-healthcare costs are associated with decreased workplace productivity and increased absenteeism for current smokers as compared to former and never smokers.

For non-workforce cohorts, such as health plan members, the projection of non-healthcare costs, such as lost productivity, may be omitted. The model preferably presents costs from the start of the model until a specified age, such as retirement age (such as age 65), or age 85, as total costs or healthcare costs only. The model determines the benefit-cost ratio for coverage of smoking cessation interventions and two break-even points: the increased investment in the smoking cessation interventions compared with direct medical cost savings alone, or with direct medical cost savings plus indirect cost savings.

Graphical Interface As discussed above, the model for estimating cost savings associated with a disease management intervention, such as a smoking cessation program, is preferably implemented as a computer program. Desirable functions performed by the program include providing easy-to-use graphical interfaces, supplying default values for most of the input data fields, and displaying output in a format that conveys costs and benefits of the disease management intervention to the user. Figure 2 is a block diagram illustrating an exemplary hierarchy of screens or windows that may be displayed on a computer display device to allow a user to enter data and view output relating to a disease management intervention. For example, each of the blocks illustrated in Figure 2 may represent a window in a windowed computing environment, such as WINDOWS^® 95, WINDOWS^® 98, WINDOWS^® 2000, or WINDOWS^® NT. Alternatively, each of the blocks may represent a screen in a non-windows based computing environment, such as DOS. In a preferred embodiment, each of the blocks illustrated in Figure 2 represents a window.

In order to initiate data input reception and output production, the program may include an initial window 200. The initial window 200 may allow the user to edit an existing scenario or to create a new scenario. As used herein, the term "scenario" refers to a computer file containing user-specified parameters or default parameters associated with the population being treated in a disease management intervention and parameters used to model the intervention. Once the user selects an existing scenario or chooses to create a new scenario, the user may be granted access to a parameters window 202, an outcomes window 204, a benefit-cost analysis window 206, a break-even analysis window 208, and a data sources window 210.

The parameters window 202 is adapted to receive user input relating to the population to be treated and relating to the design of the model. The parameters window 202 may include a button or other graphical user interface for displaying a population demographics window 212. The population demographics window 212 displays default age and gender distributions of the model population and is adapted to receive user input to change the default values. The population demographics window 212 may also be adapted to receive user input relating to job positions and the hourly rate of pay per position. Default population demographics parameters may be displayed and the user is preferably permitted to change the default parameters.

As stated above, the parameters window 202 preferably allows the user to input parameters relating to the model. For example, the parameters window 202 may allow the user to select between a cohort model and a model with replacement. The parameters window 202 may also include a button or other graphical user interface that allows the user to access an additional model parameters window 214. The additional model parameters window 214 allows the user to view and change values pertaining to additional absenteeism in days per year, decreased productivity levels, additional annual direct costs, and additional annual indirect costs. The additional model parameters window may also allow the user to view and input an annual discount rate. Annual medical expenses for current, former, and people who have never smoked may also be viewed and changed.

The parameters window 202 may include a button or other graphical user interface that allows the user to access an intervention parameters window 216. The intervention parameters window 216 allows the user to customize parameters of the disease management intervention, such as the level of promotion, the success rate, participation rates, and costs associated with the intervention. As with the other windows, default values may be displayed, and the user is preferably allowed to change the default values based on the specific parameters of the intervention being implemented. For example, one of the costs that the user may specify is the cost of a drug associated with the intervention. This cost may include the average wholesale price, the amount of co-payment, dispensing fee, etc. These costs may vary from one intervention to another. Thus, it may be desirable to allow the user to input different values. The intervention parameters window 216 may include a button or other graphical user interface that allows the user to access a recidivism rates window 218. The recidivism rates window 218 is particular to a smoking cessation program. The recidivism rates window 218 displays a schedule of 5 annual recidivism rates based on years since smoking cessation. Rates are specified with and without health plan coverage of smoking cessation aids. Alternatively, recidivism may be specified as a single annual rate specified by the user. In a model that is used to evaluate another type of disease management intervention, such as an influenza vaccination program, the0 recidivism rates window 218 may be omitted.

The parameters window 202 may include a button or other graphical user interface that allows the user to access a disease-specific costs window 220. The disease-specific costs window 220 displays costs associated with smoking-related medical complications of specified conditions. For smoking,5 exemplary medical conditions for which costs may be specified include chronic obstructive pulmonary disease, lung cancer, coronary heart disease, ischemic stroke, and pregnancy complications. Default values are preferably displayed and are modifiable by the user.

The parameters window 202 may include a button or other graphical 0 user interface for accessing a smoking status distributions window 222. The smoking status distributions window 222 presents the population breakdown by age group and gender according to smoking status, i.e., former smokers, current smokers, or persons who have never smoked (referred to herein as "never smokers"). Default percentages for the population are displayed and 5 are modifiable by the user.

The outcomes window 204 allows the user to view cost savings data associated with the disease management intervention. The outcomes are preferably presented in graphical format or tabular format to allow easy interpretation by the user. For example, the outcomes window may include a o button for accessing a cost savings graph window 224. The cost savings graph window 224 allows the user to view cost savings achieved through change in costs associated with covering smoking cessation aids at predetermined time periods since the smoking cessation plan was implemented. For example, increased cessation costs or the dollar amount that will be spent on covering smoking cessation aids may be presented. Healthcare savings representing direct medical costs avoided by smoking cessation may also be presented. For employee populations, healthcare and indirect savings may be presented. These savings represent direct medical costs avoided and lost productivity and absenteeism avoided by smoking cessation. For health plan populations, indirect savings figures may not be presented, i.e., only healthcare savings figures will be presented.

Another window which may be accessed from the outcomes window 204 is the disease savings graph window 226. The disease savings graph window 226 allows the user to view cost savings due to smoking-related medical conditions avoided by covering smoking cessation aids. The cost savings are preferably presented at fixed intervals since implementation of the disease management intervention. For a smoking cessation program, the cost savings may include the amount of money saved as a result of reduced expenditures on coronary artery disease, cerebrovascular disease, represented by ischemic stroke, chronic obstructive pulmonary disease, lung cancer, miscarriage, and low birth weight.

Another window which may be accessed through the outcomes window 204 to view cost savings associated with the disease management intervention is the outcomes table window 228. For a smoking cessation program, the outcomes table window 228 presents the number of cases of the above- mentioned conditions avoided as a result of smoking cessation for the population. For example, the number of coronary artery disease cases avoided may be presented at 2, 5, 10, and 20 years. Similar output may be displayed for the other conditions.

The benefit-cost analysis window 206 displays costs, benefits, benefit- cost ratios, incremental costs, and internal rate of return for the disease management intervention. The costs may include the monetary amount for health plan coverage of a drug, such as a smoking cessation aid, associated with the disease management intervention, for the population. Savings may include healthcare savings and indirect savings, such as savings for increased productivity and decreased absenteeism. The benefit-cost ratio represents the incremental amount saved for every monetary unit expended on the disease management intervention. The internal rate of return is defined as the discount rate that equates the net present value of a stream of cash outflows and inflows to zero. For a smoking cessation program, the internal rate of return is the discount rate for which the increased expenditure on smoking cessation would exactly equal the increased savings from direct or direct plus indirect costs.

The break-even analysis window 208 displays, in graphical and tabular format, the cost savings associated with the disease management intervention.

For example, the break-even analysis window 208 may include a first button for displaying a break-even graph 230. The break-even graph 230 presents the marginal cost associated with the disease management intervention and the cost savings associated with the disease management intervention. The point at which the cost savings exceed the marginal cost is the break-even point. For employer plan scenarios, the cost savings may include direct and indirect cost savings. For health plan scenarios, only direct cost savings may be presented.

The break-even analysis window 208 preferably also includes a button for displaying a break-even table 232. The break-even table 232 displays the results of the break-even analysis, including the time when smoking cessation cost savings exceed smoking cessation costs. As with the break-even graph, the break-even table preferably displays direct and indirect cost savings for an employer scenario and direct cost savings only for a health plan scenario.

In order to provide credibility for the calculations and outcomes presented by the model, the initial window 200 preferably includes a button or other interface for displaying a data sources window 210. The data sources window 210 displays parameter categories and citations of sources, such as studies and published reports, used to provide values for the parameter categories.

Each of the windows discussed with respect to Figure 2 will now be described in more detail. In the description below, each of the windows that comprise the graphical user interface will be explained in terms of buttons and input cells. Although the drawings do not show textual labels for each of the buttons or input cells, it is understood that each button or input cell may include a textual label that corresponds to its function as described herein. For example, referring to Figure 3, initial window 200 may include a parameters button 300, a ZYBAN^® outcomes button 304, a break-even analysis button 306, and a data sources button 308. Each of these buttons may respectively include textual labels, such as: "Parameters," "ZYBAN^® Outcomes," "Benefit/Cost Analysis," "Break-even Analysis," and "Data Sources." Buttons 300, 302, 304, 306, and 308 respectively allow the user to access the parameters window 202, the outcomes window 204, the benefit-cost analysis window 206, the break-even analysis window 208, and the data sources window 210. The buttons 300-308 may not be activatable from the initial window 200 because the user has not yet selected a scenario.

With regard to ZYBAN^® outcomes button 302, ZYBAN^® is a product name for a drug (bupropion hydrochloride) used in smoking cessation programs. The model described herein projects cost savings for the smoking cessation program based on the cost of providing ZYBAN^® as a covered health benefit. However, the present invention is not limited to evaluating cost savings based on ZYBAN^®. For example, other smoking cessation products may be substituted for evaluating cost savings for the smoking cessation program. For other types of disease management interventions, outcomes may be presented for the drug or any other aspect associated with the intervention. For example, for an influenza vaccination program, the program may be evaluated based on the costs and benefits of the vaccination program. The initial window 200 also includes a tool bar including a scenario tool

312, a mode tool 314, and a reset tool 316. The scenario tool 312 allows the user to access a drop-down window for creating a new scenario or for accessing an existing scenario. Additional functions that may be included in the scenario drop-down menu include delete, edit, name, import, export, and exit. The delete function can be used to permanently remove a scenario from 5 the program. The edit scenario function allows the user to edit existing scenarios. The export command allows the user to save a scenario under a user-specified name and location. The import command allows a user to open an exported file containing a scenario. The exit command terminates the program. 0 The mode tool 314 on the tool bar allows the user to execute the program in edit or demo mode. Edit mode allows the user to view and specify parameters associated with a disease management intervention. Demo mode is not accessible until parameters have been saved in the parameters window using the default edit mode. 5 The reset button 316 on the tool bar returns values entered by the user to default values at any point in the program. The reset button 316 preferably allows the user to select between resetting all of the parameters or only some of the parameters. For example, the reset tool 316 may allow access to a dropdown menu permitting the user to select categories of parameters to return to o default values. For a smoking cessation program, the categories of parameters may include disease-specific costs, additional model parameters, smoking status distributions, intervention parameters, or population characteristics.

Figure 4 illustrates the parameters window 202. The parameters window 202 may be displayed when the user chooses to create a new scenario5 or edit an existing scenario using the scenario tool 312 from the initial window 200. When program displays the parameters window 202, the buttons 300 - 308 and the tools 314 and 316 become activatable. The buttons 300 - 308 and tools 312 - 316 preferably remain visible in each of the windows 200 - 210 of the program to allow the user to access major program functions from any 0 window of the same hierarchical level. In the illustrated embodiment, the parameters window 202 includes input cells that allow the user to view and specify population- and model-related parameters. Each input cell may include a default parameter. Some of the input cells may include a pull down list of alternate selections. For example, a type of organization input cell 400 may allow the user to select the organization type for which the disease management intervention is being implemented. The type of organization input cell may include a pull-down list of industry and health plan types. The user preferably selects a category that most closely matches the type of organization of interest. An example of parameter that may be input by the user in cell 400 is "mining" to indicate a mining organization. A region input cell 402 allows the user to select a geographic region in which the organization of interest is located. For the United States, geographic regions provided for the user may include the Northeast, Midwest, South, and West. A state input cell 404 allows the user to view and specify the state in which the intervention is being implemented. This input cell is optional and may be omitted. The present invention is not limited to evaluating the costs and benefits associated with a disease management intervention in the United States. For example, an additional input cell may be included to select a country in which the disease management intervention is being implemented. Default parameter values for the country may be selected once the country is specified. An employee population input cell 406 allows the user to input the number of employees or health plan members in the population for which the disease management intervention is being implemented. A dependent population input cell 408 allows the user to view and specify the number of dependents associated with each member in the population. A total population input cell 410 includes the sum of the values entered in the employee population input cell 406 and the dependent population input cell 408. This value is preferably automatically calculated by the program. A population demographics button 412 allows the user to access the population demographics window 212. Figure 5(A) illustrates the population demographics window 212. Referring to Figure 5(A), the population demographics window 212 includes an age/gender folder tab 500 and an occupations folder tab 502. The age/gender folder tab 500 includes a plurality of input cells 504 that allow the user to view and change population demographics by age and gender. Table 1 shown below illustrates exemplary values that may be input into input cells 504.

Table 1 : Distribution By Age and Gender

The values in Table 1 that may be specified by the user through input cells 504 are the numbers of employees in each age group. The age group ranges and the table headers may be part of, i.e. displayed to the user, through employee demographics window 212. The "totals" in the last row of Table 1 may be calculated and displayed by the program. A reset button 506 returns the values in the input cells to the default values for any selected input cell in the window 212. An average family size input cell 508 displays the average size of a population family. A default value of 2.60 can be provided for the average family size. This value is modifiable by the user to fit the population being treated. An annual turnover rate input cell 510 allows the user to view and specify the average number of members or employees leaving the population during a year.

Figure 5(B) illustrates the occupations tab 502 of the population demographics window 212. The occupations tab 502 includes a table 507 that displays occupation types, the percentage of employees in each occupation type, and the average hourly salary associated with each occupation type. Table 2 shown below is an example of table 507 that may be displayed when a user selects occupations tab 502.

Table 2: Distribution by Occupation

The values in Table 2 can be changed to reflect those of the company of interest. Each of the values can be returned to the default values by clicking the reset button 506. Once the user has selected all of the values, the user can click "Ok" or "Cancel" to return to the parameters window 202 illustrated in Figure 4. Clicking "Ok" saves changes made by the user, and clicking "Cancel" does not save changes made by the user. The occupations tab 502 may be omitted for health plan scenarios. Referring back to Figure 4, the parameters window 202 includes input cells that allow the user to view and specify characteristics of the model. For example, a model mode input cell 414 allows the user to choose between a model with replacement or a cohort model. As used herein, a cohort model includes personnel who are part of the starting model population, and new individuals do not enter the model as members of the starting population leave or die. A model with replacement allows new individuals to enter the population to replace members that leave or die. In an exemplary model, the replacement individuals have the same age and gender as determined by the original model population. The replacement individuals may or may not have the same smoking status as determined by the original model population.

An additional model parameters button 416 allows the user to access the additional model parameters window 214. Figure 6 illustrates an exemplary embodiment of the additional model parameters window 214. The additional model parameters window 214 allows the user to view and change values pertaining to increased absenteeism per year, decreased productivity levels, additional annual direct costs, and additional annual indirect costs. For example, with regard to additional absenteeism per year, the additional model parameters window 214 includes input cells 600 that allow the user to enter absenteeism in days per year of current and former smokers. Input cells 602 allow the user to view default decreased productivity values and specify decreased productivity values. Input cells 604 and 606 allow the user to view and specify additional annual direct and indirect costs associated with male and female smokers and non-smokers. These additional costs are employer- or health plan-specific and may include any costs not already included in the model. An input cell 608 allows the user to select an annual discount rate so that the costs and benefits of a smoking cessation program can be presented using current monetary values.

Medical expense allocation input block 610 allows the user to view and specify medical expense allocation by family, gender, or smoker status. The selection can be made by clicking on the circle adjacent to the appropriate textual label (not shown).

A mean annual medical expenses table 612 is displayed in the window 214 including mean annual medical expense input cells 614 that allow the user to input mean annual medical expenses by smoker category for the subject population. Table 3 shown below is an example of mean annual medical expenses table 612.

Males Females

Current Former Never Current Former Never Smokers Smokers Smoked Smokers Smokers Smoked

2528 1416 3076 2099

Table 3: fv lean Annu al Medical Expenses

In Table 3, mean annual medical expense are calculated for current, former, and never smokers. The expenses for each category are calculated for males and females.

By default, the model calculates mean annual medical expenses for former smokers using a formula based on a weighted average of years since cessation. Alternatively, the user may designate a single annual value for former smokers by selecting an input cell 616 and entering the value in the input cells 614. However, this option is less accurate than the formula because the medical costs of former smokers are not constant. Rather, these values decrease as the years since cessation increase. Once the user has completed the additional model parameters window 214, the user may select "Ok" or "Cancel" to return to the parameters window 202. Clicking "Ok" saves changes made by the user, arid clicking "Cancel" does not save changes made by the user.

Referring back to Figure 4, the parameters window 202 includes a level of counseling input cell 418 that allows the user to select an appropriate level of counseling associated with the disease management intervention. Exemplary values that may be selected include "high," "low," and "medium," to indicate the level of counseling to the program. An intervention parameters button 420 allows the user to view and modify the parameters associated with the selected level of counseling through the intervention parameters window 216. Figure 7 illustrates an exemplary embodiment of the intervention parameters window 216 associated with a high level of counseling. A first group of input cells allows the user to view and specify parameters associated with the level of promotion. For example, a cost level input cell 700 allows the user to select a cost associated with the promotion of the smoking cessation program. A second input cell 702 allows the user to select whether ZYBAN^® is promoted as part of the smoking cessation program. A third input cell 704 allows the user to view and specify the percentage of physician visits for smoking cessation only. For example, some members may receive counseling regarding the disease management intervention during a regular physician visit, such as an annual physical. The value specified in input cell 704 indicates physician visits for purposes of the disease management intervention program only. An input cell 706 allows the user to view and specify the number of hours away from work for the physician visit.

A table 708 allows the user to view and specify intervention costs and success rates associated with selected intervention products. Table 4 shown below illustrates an example of Intervention Costs and Success Rates table 708 that may be displayed in intervention parameters window 216.

Table 4: Intervention Costs and Success Rates

In table 4, the column labeled "Success Rate" corresponds to input cells 710, 712, 714, and 715 that allow the user to view and specify success rates for ZYBAN^® alone, ZYBAN^® plus a nicotine patch, other aids, and no aids. The column labeled "Non-ZYBAN^® Cost" corresponds to input cells 716, 718, 720, and 722 that allow the user to view and specify non-ZYBAN^® costs associated with each of the products. Non-ZYBAN^® costs may include counseling costs, promotional costs, and other costs for each type of cessation aid. The column labeled "Total Cost Per Quit Attempt" includes calculated values indicative of the total cost per quit attempt associated with each of the cessation aids listed in table 708.

Another table 723 allows the user to view and specify ZYBAN^® costs. Table 5 shown below illustrates an example of ZYBAN^® cost table 708 that may be displayed in intervention parameters window 216.

Table 5: ZYBAN^® Cost

Input cell 724 in Figure 7 corresponds to the row labeled "AWP" in Table 5 and allows the user to view and specify the average wholesale price of ZYBAN^®. Input cell 725 in Figure 7 corresponds to the row labeled "No. of Weeks" in Table 5 and allows the user to view and specify the number of weeks of ZYBAN^® treatment. Input cell 726 in Figure 7 corresponds to the row labeled "Copayment" in Figure 7 and allows the user to view and specify the co- payment that a member of the population would pay for each ZYBAN^® prescription. Input cell 728 in Figure 7 corresponds to the row labeled "No. Prescriptions" in Table 5 and allows the user to view and specify the number of prescriptions. Input cell 730 in Figure 7 corresponds to the row labeled "Discount" in Table 5 and allows the user to view and specify the discount for each prescription. Input cell 732 in Figure 7 corresponds to the row labeled "Dispensing Fee" in Table 5 and allows the user to view and specify a dispensing fee associated with the prescription. Calculation cell 734 in Figure 7 corresponds to the row labeled "Total" in Table 5 and indicates the total cost per employee for the number of weeks selected in the input cell labeled "No. of Weeks". Yet another table 735 includes input cells that allow the user to view and specify participation rates with and without ZYBAN^® coverage. Table 6 shown below is an example of participate rates table 735 that may be displayed in intervention parameters window 216.

Table 6: Participation Rates

Input cell 736 in Figure 7 corresponds to the row labeled "Without coverage" in Table 6 and allows the user to view and specify the percentage of the population that would participate in the disease management intervention without cessation aids coverage. Input cell 738 in Figure 7 corresponds to the row labeled "With coverage" in Table 6 and allows the user to view and specify the percentage of the population that would participate in the disease management intervention with cessation aids coverage. In Table 6, the percentages of the population that would participate in the disease management intervention with and without cessation aids coverage are equal because there is no promotion of the smoking cessation program or ZYBAN^®.

Yet another table 739 allows the user to view success rates and costs associated with a disease management intervention with and without cessation aids coverage. Table 7 shown below is an example of summary of success rates and costs table 739 that may be displayed in intervention parameters window 216.

The column labeled "Success Rate" in Table 7 corresponds to cells 740 and 5 741 in Figure 7, which allow the user to view the success rates for the program with and without cessation aids coverage. The columns labeled "Cost per Quit Attempt" in Table 7 correspond to cells 742, 743, 744, and 746 in Figure 7, which allow the user to view the covered and total cost per employee per quit attempt when cessation aids are covered and not covered. 0 The intervention parameters window 216 may also include a "view recidivism rate" button 748 that allows the user to access the recidivism rates window 218. Figure 8 illustrates an exemplary embodiment of the recidivism rates window 218. The recidivism rates window 218 allows the user to view and specify percentages of population members that quit and later begin5 smoking again. In Figure 8, the recidivism rates window 218 includes input cells 800 and 802 to allow the user to view and specify an annual rate or to use a rate schedule at specified years since cessation. If the user selects "Use a Rate Schedule," a schedule of annual recidivism rates displays the percentage of smokers that return to smoking based on fixed time periods since cessation. o Table 8 shown below illustrates an example of a schedule of annual recidivism rates that may be displayed in window 218.

The first row in Table 8 contains values for no cessation aid coverage. This row corresponds to input cells 804 illustrated in Figure 8. The second row in Table 8 contains values for cessation aid coverage. This row corresponds to input cells 806 illustrated in Figure 8. Once the user has selected or viewed the rates displayed in the recidivism rates window 218, the user can return to the intervention parameters window 216 by actuating the "Ok" or "Cancel" buttons. Once all of the parameters in the intervention parameters window 216 have been viewed or specified, the user can click "Ok" or "Cancel" to return to the parameters window 202. Clicking "Ok" saves changes made by the user, and clicking "Cancel" does not save changes made by the user.

Referring back to the parameters window 202 illustrated in Figure 4, an intervention availability input cell 422 allows the user to specify whether the disease management intervention will be available continuously or at specified time periods. For example, the intervention may provide one-time-only coverage of smoking cessation aids or continuous coverage of smoking cessation aids. In either case, the costs and outcomes with smoking cessation aids coverage are compared to the costs and outcomes with no smoking cessation aids coverage for the model. A disease-specific costs button 424 allows the user to access the disease-specific costs window 220. Figure 9 illustrates an exemplary embodiment of the disease-specific costs window 220. In Figure 9, the disease-specific costs window 220 includes input cells for displaying and allowing the user to view and specify the average total and annual costs of treatment for various conditions associated with the disease or diseases being treated. For example, input cell 900 includes the annual cost of treating chronic obstructive pulmonary disease. Similarly, input cell 902 displays lifetime costs for treating lung cancer in an individual of the population. With regard to coronary heart disease, input cells 904 display age- and gender-specific costs.

Table 9 shown below illustrates exemplary disease-specific costs for coronary heart disease that may be displayed in window 220.

Table 9: Lifetime Costs for Coronary Heart Disease

Input cells 904 correspond to the values in Table 9 for each age group/gender. For example, the program preferably displays default values in the input cells. These values can be determined based on actuarial data and are preferably modifiable by the user. For cerebrovascular disease, represented by ischemic stroke, input cells 906 display age- and gender- specific costs. Table 10 shown below illustrates exemplary disease-specific costs for cerebrovascular disease that may be displayed in window 220.

Table 10: Lifetime Costs for Cerebrovascular Disease Input cells 908 display episode-specific costs of adverse pregnancy outcomes. Input cells 906 correspond to the values in Table 10 for each age group/gender. For example, the program preferably displays default values in the input cells. These values can be determined based on actuarial data and are preferably modifiable by the user. Table 11 shown below illustrates exemplary costs for pregnancy complications.

Table 11 : Costs for Pregnancy Complications The values in Table 11 are default values for each specific complication and may be determined based on actuarial data.

Once the user has completed viewing or changing the parameters illustrated in Figure 9, the user may select "Ok" or "Cancel" to return to the parameters window 202. Clicking "Ok" saves changes made by the user, and clicking "Cancel" does not save changes made by the user.

Referring back to Figure 4, the parameters window 202 may include a smoking characteristics button 426 that allows the user to access the smoking status distributions window 222. Figure 10 illustrates an exemplary embodiment of the smoking status distributions window 222. In Figure 10, the smoking status distributions window includes a plurality of input cells 1000 that display values representing the population break-down by age group and gender according to smoking status, i.e., current smoker, former smoker, or person who has never smoked. This distribution is applied to the total model population, including both employees or members and adult dependents. Default values are specified, and the user can change the default values by accessing the appropriate cell. Table 12 shown below illustrates exemplary smoking status distribution values that may be displayed in window 222.

Once the user has completed viewing or changing the values in the smoking status distributions window 222, the user can click "Ok" or "Cancel" to return to the parameters window 202. Clicking "Ok" saves changes made by the user, and clicking "Cancel" does not save changes made by the user. Referring back to Figure 4, once the user has completed entering all desired parameters relating to the population, the user can select one of the buttons 302, 304, and 306 to view cost savings data associated with the disease management intervention. For example, in order to view outcomes associated with the disease management intervention, the user may select the outcomes button 302. Selecting the outcomes button 302 causes the program to display an outcomes window 204. Figures 11 (A) - 11 (C) illustrate exemplary embodiments of the outcomes window 204. In Figure 11 (A), the cost savings graph window 224 is displayed. The cost savings graph window 224 illustrates healthcare costs and healthcare savings over predetermined time periods associated with the disease management intervention. For example, in Figure

11 (A), the costs and savings are displayed at 2, 5, 10, and 20 year intervals.

In Figure 11(A), shading 1103 represents the 2-year model, 1104 represents the 5-year model, 1106 represents the 10-year model, and 1108 represents the

20-year model. Bars 1110 represent increased cessation costs, bars 1112 represent health care savings, and bars 1114 represent health care plus indirect savings. Healthcare savings include direct medical costs avoided by smoking cessation. Healthcare plus indirect savings include direct medical costs avoided, lost productivity avoided, and absenteeism avoided by smoking cessation. The amount of healthcare savings may be calculated by the difference in healthcare costs between former smokers and current smokers in the two model scenarios (with and without coverage of smoking cessation aids). The values on the vertical axis in Figure 11 (A) are in U.S. dollars. In Figure 11 (A), a "view cost savings graph" button 1100 is actuated in order to select display of the cost savings graph. In order to access a disease savings graph, the user may select a "view disease savings graph" button 1101. Figure 11(B) illustrates an exemplary embodiment of the disease savings graph window 226. In Figure 11 (B), the abscissa axis represents various diseases associated with smoking. For example, bars 1116 represent coronary artery disease, bars 1118 represent cerebrovascular disease, bars 1120 represent COPD, bars 1122 represent lung cancer, bars 1124 represent miscarriage, and bars 1126 represent low birth weight.

The ordinate axis represents the number of dollars saved associated with each disease. As in Figure 11(A), shading 1103, 1104, 1106, and 1108 respectively represent data for 2, 5, 10, and 20-year models. The user can select the view outcomes table button 1102 to view the outcomes table window 228. Figure 11 (C) illustrates an exemplary embodiment of the outcomes table window 228. Table 13 shown below illustrates an exemplary outcomes table that may be displayed in window 228.

Table 13: Improved Outcomes with ZYBAN^® The data illustrated in Table 13 allows the user to view improved outcomes achieved through smoking cessation aids. In the table, outcomes are expressed as the number of cases avoided. For example, the entry at row 1 , column 1 indicates that 3.068 times as many population members will quit smoking as a result of using ZYBAN^® than would quit without using ZYBAN^®. Similar data is presented for the diseases associated with smoking and the number of deaths postponed by ZYBAN^®.

Additional cost savings data that may be viewed for a disease management intervention includes a benefit-cost analysis. The benefit-cost analysis may be accessed through the benefit-cost analysis button 304. For example, when a user selects the benefit-cost analysis button 304, the benefit- cost analysis window 206 will be displayed. Figure 12 illustrates an exemplary embodiment of the benefit-cost analysis window 206. Tables 14, 15, and 16 illustrate exemplary benefit/cost data that may be displayed in window 206.

Table 14: Health, Indirect, and Cessation Costs

Table 15: Benefit-Cost Ratios

Table 16: Internal Rate of Return

As illustrated in Tables 14-16, the benefit-cost analysis window 206 displays costs with and without cessation aids coverage, benefit-cost ratios, incremental costs per employee per year or per member per month for a health plan, and internal rate of return. For example, under the heading "Costs without Cessation Aids Coverage," $57,746,150 indicates the total healthcare costs for a population without smoking cessation aids coverage. This number is based on an employee population of 1 ,000 with 600 dependents and the other parameters entered in Figures 4-10. Under "Costs with Cessation Aids Coverage," the number $57,661 ,119 indicates total healthcare costs with cessation aids coverage. Under the heading "Savings with Cessation Aids Coverage," the number $85,031 indicates the difference between the total healthcare costs with and without cessation aids coverage. The benefit-cost ratio of 2.76 indicates that for every additional dollar spent on smoking cessation, $2.76 are saved in healthcare expenditures. Additional data that may be displayed in window 206 is the incremental cost of pharmaceutical coverage per employee per year. For health plan models, an incremental cost per member per month is determined for the first year, which incorporates the number of beneficiaries per member, as specified in the population characteristics box in the parameters screen. The internal rate of return is calculated over a 20 year period. Similar calculations are performed for healthcare plus indirect costs. Another manner in which cost savings associated with the disease management intervention may be presented is through a break-even analysis. In order to display results from the break-even analysis, a user may select the break-even analysis button 306. The breakeven analysis button 306 causes display of the break-even analysis window 208. Figures 13(A) and 13(B) illustrate exemplary embodiments of the breakeven analysis window 208. For example, Figure 13(A) illustrates the break- even analysis graph window 230 that is displayed when the user selects a "view break-even graph" button 1300. Figure 13(B) illustrates the break-even table window 232 that is displayed when the user selects "view break-even table" button 1302. Referring to Figure 13(A), the break-even graph illustrates the number of years until cost savings exceed smoking cessation costs. The abscissa represents time in years. The ordinate axis represents cost in dollars.

Line 1303 represents smoking cessation costs averaged over a 20 year period. Line 1304 represents healthcare cost savings with cessation aids coverage. Line 1305 represents healthcare plus indirect cost savings with cessation aids coverage. The intersection of lines 1303 and 1304 indicates that the break-even point for healthcare cost savings occurs in about 7 years. The intersection of lines 1303 and 1305 indicates that the break-even point for healthcare plus indirect cost savings occurs in about 3 years. Specific data points can be viewed by selecting the break-even table button 1302.

Figure 13(B) illustrates an exemplary embodiment of the break-even table window 232. An example of a break-even analysis table that may be displayed in window 232 is shown below.

Table 17: Break-Even Analysis Table In Table 17, the columns indicate, respectively, model year, annual cessation costs with coverage, cumulative increased cessation costs with coverage, cumulative healthcare savings, return on investment for healthcare costs only, cumulative health and indirect savings, and return on investment for healthcare and indirect costs. As the table indicates, a positive return on investment for healthcare costs only occurs at approximately 7 years. A positive return on investment occurs for healthcare and indirect costs between 3 and 4 years.

According to an important aspect of the invention, the program preferably allows users to view data sources associated with the model. In order to access the data sources window 210, the user selects the data sources button 308. Figure 14 illustrates an exemplary embodiment of the data sources window 210. In the illustrated embodiment, a parameter categories input cell 1400 allows the user to select categories of parameters for which the user desires to view source information. The category selected by the user controls display of parameters and data sources from which default values for the parameters were derived. Area 1404 displays a table including a parameter list and sources for default values for each parameter. Table 18 shown below is an example of a data sources table that may be displayed in area 1402.

Table 18: Data Sources Area 1404 includes citations for the parameter category selected using area 1400. For example, if disease costs is selected as the parameter category. Low birth weight costs is one of the parameters under the disease costs category. The full citations from which the default value for the low birth weight parameter was derived are displayed in area 1404. An example of a citation that may be displayed in area 1404 is as follows: Halpern et al 1996:

Halpern et al., Cost-Effectiveness of Preconception Counseling in the United States. Report to the U.S. Centers for Disease Control and Prevention, 1996. Thus, because the user can access sources from which default parameter values were derived, the user can evaluate the credibility of the model calculations.

According to another aspect of the invention, the parameters window 210 may also allow the user to select and view definitions associated with the model. For example, the parameter categories input cell 1400 may include a definitions option. When the user selects the definitions option, definitions for model parameters may be displayed in the locations occupied by citations in the illustrated embodiment. Allowing the user to access and view definitions facilitates interpretation of model outcomes.

Model Inputs Table 19 shown below illustrates exemplary model inputs and data sources that may be used to estimate costs and benefits for a smoking cessation program. The left column of the table illustrates exemplary model input parameters. The right column of the table illustrates exemplary sources for default values for the model parameters. Sources that indicate "Glaxo Wellcome Data" were derived from data collected by Glaxo Wellcome, Inc. Sources that indicate "Towers Perrin Data" were derived from data collected by Towers Perrin. Full citations for each of the sources listed in the right column are included in a bibliography section below.

Table 19: Model Inputs

Table 20 shown below illustrates additional model inputs by category. The left column illustrates parameter types. The center column illustrates input parameters associated with each parameter type. The right column illustrates exemplary sources for default values for each parameter. Full citations for each of the remaining sources for default parameters are listed below in the bibliography section.

Table 20: Other Model Parameters Assumptions Assumptions relating to model input parameters that were obtained from government databases will now be explained in more detail.

Population Distribution by Age, Gender, and Smoking Status The information for this parameter was abstracted from the 1994 National Health Interview Survey database. The data reflects specific groups/categories for each of the four regions of the country and industry/health plan type. Population Distribution by Occupation Type This information was obtained from the 1997 Current Population Survey, specific for each of the four regions of the country and industry type.

Adult Dependents The model assumes that there are 0.6 adult dependents for each employee/health plan member (based on values supplied by Towers Perrin). Adult dependents are assumed to be the same age and opposite gender of the employee/health plan member. The smoking status distribution for the adult dependents is the same as specified in the current population smoking status distribution for employees/health plan members of the specified age cohort and gender. Adult dependents do not incur any indirect costs and may change smoking status independent of the employee/health plan member. Mean Hourly Salary by Occupation Type This information was obtained from the 1997 Current Population Survey, specific for each of the four regions of the country or the state (if specified).

Workforce or Health Plan Turnover For the model with replacement, the turnover rate has a default value of 10% annually; this value incorporates job termination and retirement. For the cohort model, a schedule for turnover rate with years of continuous employment is used. Values for both the model with replacement and the cohort model were supplied by Towers Perrin. The rate schedule for turnover in the cohort model is as follows:

In the model with replacement, the program assumes that employees/health plan members who leave prior to age 65 will be replaced by new employees/health plan members with the same age and gender. The smoking status of the replacement employee/health plan member will be determined based on the distributions from the original model population (i.e., the smoking status distribution in the overall population, based on age, gender, occupation mix, and region of country), not the current distribution in the workforce or health plan when the employee/health plan member leaves. When an employee/health plan member leaves, a proportional number of adult dependents also leave (based on the proportion of the employees/health plan members to adult dependents). The departing adult dependents will have the same age and opposite gender of the departing employees/health plan members. The smoking status of the departing adult dependents will be based on the proportion of never/former/current smokers currently in the workforce/health plan for that gender and age cohort. Employees dying prior to age 65 will be replaced in a similar fashion.

Probability of Using Smoking Cessation Aids Based on data from Towers Perrin, if smoking cessation aids are covered by an employer or health plan, 14% of smokers attempting to quit will use ZYBAN^® (alone or with nicotine patches) while the remaining 86% will use other aids (such as nicotine replacement therapy only) or no aids. If smoking cessation aids are covered and ZYBAN^® is promoted, 17% will use ZYBAN^® while the remaining 83% will use other aids or no aids. If smoking cessation aids are not covered, only 7% of smokers attempting to quit will use ZYBAN, while the remaining 93% will use other aids or no aids.

Among smokers using ZYBAN^® (with or without coverage), 97% will use ZYBAN^® alone while 3% will use ZYBAN^® plus nicotine patches. Among smokers not using ZYBAN^® (again, with or without coverage), 30% will use other aids and 70% will use no aids.

Effectiveness of Smoking Cessation Interventions by Level of Counseling Cessation rates for ZYBAN^® alone and ZYBAN^® plus nicotine patches were based on data from a Glaxo Wellcome clinical trial with five minutes of brief counseling. The efficacy of other aids was based on rates from Tonnesen et al. Efficacy rates of the no aids option was based on Cromwell et al. The change in effectiveness of each of these cessation aids in conjunction with low or no/minimal levels of counseling was estimated using data from Cromwell et al. (1997). For example, Cromwell et al. report that comparing cessation using nicotine patch therapy with full counseling (equivalent to the high counseling level in the model) to brief counseling (equivalent to the low counseling level in the model), decreases effectiveness by 7.6%. Therefore, the model assumes that going from the clinical trial data with full counseling to a low level of counseling, effectiveness for ZYBAN^® alone, ZYBAN^® plus nicotine patch, or other aids will decrease by 7.6%. Similarly, comparing use of nicotine patch therapy with full counseling versus minimal counseling (equivalent to the no counseling level in the model), Cromwell et al. report that cessation effectiveness decreases by 9.3%. For smokers attempting to quit using no aids, Cromwell et al. report that the success rate decreases by 4.34% going from full counseling to brief counseling (with no aids) and by 6.2% going from full counseling to no counseling (with no aids).

Cost of Smoking Cessation Interventions by Level of Counseling Individuals using any aids for smoking cessation were assumed to have at least minimal counseling associated with the aids. Therefore, the cost for cessation using any aid(s) in the model with no counseling includes the cost reported by Cromwell et al. (1997) for minimal counseling. Costs in the model associated with a low level of counseling include the Cromwell et al. cost for brief counseling, and model costs for a high level of counseling include the Cromwell et al. cost for full counseling. Smokers in the model using no aids incur no cessation costs when no counseling is present, and incur brief or full counseling costs for the low and high levels, respectively. Physician Costs for Smoking Cessation with ZYBAN^® For the individuals attempting smoking cessation using ZYBAN^® when cessation aids are covered, 9% will have a physician visit solely to receive the ZYBAN^® prescription. The other 91 % of individuals attempting to quit smoking using ZYBAN^® will receive their prescription at a physician visit primarily involving other reasons. The cost of the physician for this 9% and the time off of work to go to the physician (4 hours) will be included in the cost of the covered cessation attempt. The 9% and 4 hour values were supplied by Towers Perrin. Length of Therapy and Dosing for ZYBAN^®

Efficacy rates for ZYBAN^® were based on patients receiving 9 weeks of 150 mg ZYBAN^® twice a day in the clinical trial utilized for this model. However, pricing was based on 7 weeks of therapy.

Pricing for ZYBAN^® The starting price for ZYBAN^® is based on average wholesale price

(AWP) for 7 weeks of therapy. This model assumes that the pharmacy is reimbursed by the third-party payor at 13% off of AWP plus a dispensing fee of $2.50. Smokers attempting to quit using ZYBAN^® (alone or with nicotine patches) receive the medication in two prescriptions, each having a co- payment of $8.00.

This model assumes for Medicaid plans that the pharmacy is reimbursed by the third-party payor at a 10% discount and 12.3% rebate from AWP. There are two prescription co-payments of $1.00 each and two dispensing fees of

$4.00 each. There is then a federal match of 50% of the total cost, resulting in states being responsible for the remaining 50%. , Smoking Cessation Promotion Employers and health plans which cover cessation aids may optionally choose to promote cessation. No promotion results in no additional costs and a rate of attempting smoking cessation of 34%, equal to the participation rate with no coverage of cessation aids. Low promotion costs $200 and results in an increased cessation attempt rate calculated by multiplying the no promotion rate by 1.5% and adding the product to the no promotion rate. Thus, the rate of attempting smoking cessation for low promotion is 34.5%. Medium promotion costs $500 and results in an increased cessation attempt rate obtained by multiplying the no participation rate by 3% and adding the product to the no promotion rate. For a no promotion rate of 34%, the cessation attempt rate for medium promotion is 35.02%. High promotion costs $1000 and results in an increased cessation attempt rate obtained by multiplying the no promotion rate by 4.5% and adding the product to the no promotion rate. For a no promotion rate of 34%, the cessation attempt rate resulting from high promotion is 35.53%. The costs and effects of promotion were obtained from data supplied by Towers Perrin and Glaxo Wellcome.

Disease and Mortality Risk by Age, Gender, and Smoking Status

The impact of smoking status on disease risks and overall mortality was assessed using relative risk values from two sources: Smoking-Attributable

Mortality, Morbidity, and Economic Costs (SAMMEC ) II for coronary artery disease (CAD) (International Disease Classification, Ninth Revision (ICD-9) codes 410-414), cerebrovascular disease (CVD) (430-438), chronic obstructive pulmonary disease (COPD) (496) and lung cancer (162); and the American

Cancer Society Cancer Prevention Study-2 (Current Population Survey (CPS)-

2) for overall mortality. "SAMMEC II" is computer software developed for the Office on Smoking and Public Health Service to permit rapid calculation of deaths, years of potential life lost, direct healthcare costs, indirect mortality costs, and disability costs associated with cigarette smoking. Although the

SAMMEC relative risk values are for mortality, not morbidity, other age/gender/smoking status disease rates or relative risk values could not be identified. Following consultation with a number of experts in this field,

Applicants elected to use the SAMMEC values for disease values.

Separate relative risk values for current and former smokers (both compared to never smokers) and males versus females within each smoking status group were used. In addition, separate relative risk values for CAD and CVD for current and former smokers ages 35-64 versus age greater than 65 were also used. For individuals less than age 35, the relative risk value from the 35-64 age group was also used.

To determine the annual rates of CAD, CVD, and COPD by age, gender, and smoking status, the overall population size and population disease rate were determined using the 1993 National Health Interview Survey (NHIS) for males and females in six age cohorts: 18-24, 25-34, 35-44, 45-54, 55-64, and 65+. The proportion of never, current, and former smokers in each of these age/gender stratum was determined using the Year 2000 supplement from the 1993 NHIS. The disease rate among never smokers in a given age/gender stratum, R, was then determined as: NUM where R _{NS + CS} * _{cs +FS} * β

NUM = the total number of individuals in the age/gender stratum developing the specified disease OR

= the rate of disease in the age/gender stratum times the total population of the stratum NS = the number of never smokers in the age/gender stratum CS = the number of current smokers in the age/gender stratum RRcs = the relative risk for the disease condition among current smokers FS = the number of former smokers in the age/gender stratum RRfs = the relative risk for the disease condition among former smokers To determine the disease rate among former or current smokers, the rate among never smokers were multiplied by the appropriate relative risk value. For lung cancer, a similar calculation was performed; however, lung cancer rates among males and females by age group were obtained from the National Cancer Institute's Cancer Statistics Review. A similar calculation was also performed for overall mortality, using annual mortality rates by age and gender from the 1993 National Center for Health Statistic's Vital Statistics of the United States.

For evaluating pregnancy complications, the rate of spontaneous abortions among smokers versus non-smokers (former plus never) was derived from DiFranza and Lew, 1995, while the rate of low birth weight infants was obtained from Marks et al., 1990. It was assumed that former and never smokers experienced the same rate of pregnancy complications. Annual pregnancy rates by age were determined using the 1993 NHIS, and an assumption was made that no pregnancies occurred after age 44. Based on Marks et al., 1990, Applicants also assumed that 20% of female smokers temporarily stop smoking when pregnant and thus experience the same rate of pregnancy complications as non-smokers; these women resumed smoking following conclusion of the pregnancy. Age-specific rates of induced (elective) abortion were also included to determine the actual number of live births and the proportion of live births (by smoking status) of low birth weight.

Length of Time After Quitting Before Smoker Presents

Medically as a Non-Smoker Based on data from Towers Perrin, former smokers' overall healthcare costs equal those of never smokers at 16 years following cessation. However, former smokers' risks for specific conditions evaluated in the model (COPD,

CAD, CVD, and lung cancer) as well as excess mortality risk remains elevated to that of never smokers through the model, as the relative risks for these conditions and for overall mortality are based on data comparing all former smokers to never smokers.

Annual Medical Care Costs for Current and Never Smokers Annual medical care costs were derived from values reported by Hodgson (1992). Hodgson presented costs over a period of several years, and presented separate costs for smokers/non-smokers who survived for the specified time period or died during the period. Applicants determined weighted average annual costs for current and never smokers by evaluating the proportion in each age/gender/smoking status cohort who survived vs. died in each period and the number of years of survival among the population dying in each period. These costs were then inflated to 1997 values using the medical care component of the consumer price index (CPI). Overall annual medical care costs for male current smokers, male never smokers, female current smokers, and female never smokers are determined by weighting the annual costs in each age/gender/smoking status cohort by the proportion of individuals in each age group (e.g., the annual medical care costs for male smokers is determined by summing the proportion of male smokers in each age cohort multiplied by the annual medical care costs for male smokers in that age cohort). Annual Medical Care Costs for Former Smokers Values were supplied by Towers Perrin determining the proportion of excess medical costs incurred by current smokers which former smokers experience annually, based on years since cessation. After 16 years of 5 cessation, former smokers have the same costs as never smokers. However, the data supplied by Towers Perrin was a step function (discrete annual values), which led to costing abnormalities (e.g., costs for individuals who quit 10.99 years ago were substantially different from costs for individuals who quit 11 years ago). To smooth these values and avoid such abnormalities, o Applicants used curve fitting software to determine a function fitting the Towers

Perrin data. The function was created using quadratic curve fitting, where the percent of incremental smokers' total healthcare costs applicable to former smokers after X years of cessation equals (0.87075-0.09084804*X +0.00247549*X²) for X<18 and equals zero for X=>18. The R-squared value5 for the quadratic curve fitting of the Towers Perrin data is 0.9933143. The years since cessation used in the regression equation is a weighted average of all former smokers in each age/gender cohort. Years since cessation for individuals who are former smokers at the start of the model (i.e., quit smoking prior to the start of the model) was determined for each age/gender cohort o using data from the NHANES, 1988-1994.

Recidivism Rate An exemplary schedule of annual recidivism rates used by the model is as follows:

The source for the values in the recidivism rates schedule is a smoothed function of 1990 Surgeon General's Report data.

Productivity Average lost productivity cost in employer charts was calculated only for the employee covered lives (not dependents). It is assumed that smokers have an additional cost of 1.4% of payroll due to greater time off and workplace breaks for smoking.

Smoking Cessation Success Rates Success rates with ZYBAN^® are based on clinical trial results after one year. Towers Perrin assumes medical and productivity savings do not accrue to the plan or the employer unless a smoker has been abstinent for a full year. Other than therapy with ZYBAN^®, success rates are the weighted average success rates of nicotine replacement therapy and no cessation aids. Fifty- two-week continuous abstinence data was used. Continuous abstinence is defined as the percentage of patients who were continuously smoke free from their quit day to the day of follow-up.

BIBLIOGRAPHY Full citations for each of the data sources described above are as follows:

CDC. Smoking cessation during previous year among adults - United States,

1990-1991.

MMWR. 1994; Vol. 42, No. 50., pp. 925-930.

Cromwell J, Bartosch WJ, Fiore MC, Hasselblad V, Baker T. Cost- effectiveness of the clinical practice recommendations in the AHCPR guideline for smoking cessation.

JAMA. 1997;278:1759-1766.

DiFranza JR, Lew RA. Effect of maternal cigarette smoking on pregnancy complications and sudden infant death syndrome. J Fam Pract. 1995; 40:385- 394.

Gold MR, Siegel JE, Russell LB, Weinstein MC. Cost-effectiveness in Health and Medicine. New York, NY: Oxford University Press; 1996.

Halpern M, Sorenson S, Elixhauser A, Haddix A, Hatziandru E, Tanzi V, Shikiar

R. Cost-Effectiveness of Preconception Counseling in the United States. Report to the US Centers for Disease Control and Prevention; 1996.

Hodgson TA. Cigarette smoking and lifetime medical expenditures. Millbank

Quarterly. 1992; 70(1 ):81-125. Marks JS, Koplan JP, Hogue CJ, Dalmat ME. A cost-benefit/cost-effectiveness analysis of smoking cessation for pregnant woman. Am J Prev Med. 1990;

6:282-289.

Oster G, Colditz GA, Kelly NL. The Economic Costs of Smoking and Benefits of Quitting. Lexington, Massachusetts: Lexington Books; 1984.

Riley GF, Potosky AL, Lubitz JD, Kessler LG. Medicare payments from diagnosis to death for elderly cancer patients by stage of diagnosis. Medical

Care. 1995; 33:828-841.

Strauss MJ, Conrad D, LoGerfo JP, Hudson LD, Bergner M. Cost and outcome of care for patients with chronic obstructive lung disease. Analysis by physician specialty. Medical Care. October 1986; 24(10):915-924.

Taylor TN, Davis PH, Tomer CJ, Holmes J, Meyer JW, Jacobson MF. Lifetime cost of stroke in the United States. Stroke. 1996; 27:1459-1466.

Tonnesen P, Norregaard J, Simonsten K, Sawe U. A double-blind trial of a 16- hour transdermal nicotine patch in smoking cessation. N Engl J Med. 1991 ;

325:311-315.

Warner KE, Smith RJ, Smith DG, Fries BE. Health and economic implications of a work-site smoking cessation program: a simulation analysis. J Occup

Environ Med. 1996; 38(10):981 -992. 7997 Physicians Fee <& Coding Guide. HealthCare Consultants of America,

Inc.; 1997.

7998 Drug Topics Red Book. Medical Economics Company; 1998.

The disclosure of each of the above-cited sources is incorporated herein by reference in its entirety. The present invention is not limited to the sources listed above for default parameters. For example, as new sources are published, values for default parameters in the model can be updated to reflect data in the new sources. In addition, because the model allows the user to change default parameters, the user may use sources of which the user has knowledge, enter parameters from these sources into the model, and calculate cost savings based on these sources. Because the model allows the user to modify default parameter values, the model is capable of adapting to changes in the population for which the disease management intervention is being implemented.

Calculations In the software according to the present invention, users input the number of employees/health plan members, number of adult dependents, type of organization (workplace or health plan), geographic region, and model mode (model with replacement versus cohort model). Based on the type of organization and geographic region, the combined employees/health plan members and adult dependent population is allocated into male or female cohorts in six age groups (18-24, 25-34, 35-44, 45-54, 55-64, 65+). This proportion of model people in each age/gender group is determined using parameters from the NHIS.

Each age/gender cohort is separated into never, current, or former smokers, also using data from the NHIS. For models involving workplace programs, the workforce (number of employees) is separated into 12 different occupation classes, based on the specified employer type and geographic region using data from the Current Population Survey. Each occupational class is assigned an average hourly wage based on the geographic region and employer type, again using data from the Current Population Survey. An overall average hourly wage is determined as a weighted average of the hourly wages for each occupation type multiplied by the proportion of the workforce in that occupation type.

To determine costs and outcomes for the model, calculations are performed separately for each of the six age cohorts. However, these calculations are identical for each cohort group; only the parameters used in the calculation change. Within each age cohort, separate calculations are performed with coverage of smoking cessation aids (i.e., ZYBAN^®) versus without coverage of smoking cessation aids. The calculations with ZYBAN^® coverage versus without are also identical; only the parameters related to smoking cessation change. Within each age cohort, calculations are performed on six groups: male never smokers; male former smokers; male current smokers; female never smokers; female former smokers; and female current smokers. Separate calculations are performed for each year of the model (until the age cohort reaches age 85). Calculations within each age cohort begin by 5 assigning the mean age of the cohort. For example, for the 25-34 cohort, the mean age for the first year of the model is assigned as 29. This increases by one for each subsequent year of the model until age 85.

For male never smokers in a given cohort, calculations begin by determining the number of male never smokers in that cohort in the model. For0 the first year of the model, this is simply the proportion of the overall model population (employees/health plan members plus adult dependents) who are in the specified age/gender/smoking status group. For each subsequent year of the model, the number of male never smokers in that age cohort who died or left the firm/health plan in the previous year are subtracted, while the number5 of new male never smokers of that age cohort are added. The number of male never smokers developing coronary artery disease each year is determined by multiplying the age-gender-smoking status specific incidence of coronary artery disease by the number of male never smokers in that age cohort. The cost attributable to coronary artery disease each year is determined by multiplying o the number of male never smokers developing coronary artery disease by the appropriate cost per case of coronary artery disease. Costs occurring after the first year of the model are discounted at a default annual rate of three percent.

Parallel calculations are performed to determine the number and attributable cost of cerebrovascular disease, chronic obstructive pulmonary diseases, and5 lung cancer for male never smokers in each age cohort.

Total annual healthcare costs for male never smokers in each age cohort are determined by multiplying the number of male never smokers in each age cohort each year by the average total health care cost for that age- gender-smoking status group. Total health care costs occurring after the first o year of the model are discounted at a default annual rate of three percent. If additional direct or indirect costs are specified by the model users, these costs are also totaled and discounted for each model year. The number of deaths occurring each year among male never smokers is determined by multiplying the number of male never smokers in each age cohort by the age-gender- smoking status specific mortality rates. For models involving workplaces, the number of male never smokers leaving the firm (quitting or retiring) or health plan (disenrolling) each year is determined by multiplying the number of employees in each male never smoker age cohort by the turnover rate (which is either a user-supplied fixed value or a default set of values based on year of employment/health plan membership). The number of male never smoker employees dying or leaving the firm/health plan each year is summed, and a proportional number of female adult dependents of the same age cohort will also leave the model. In the cohort model, individuals leaving the model are not replaced. In the model with replacement, the total number of males (employees/health plan members and adult dependents) leaving the model each year are replaced with new males of the same age cohort. The smoking status of the replacements is determined using the initial smoking status breakdown of the model populations (which is based on age-gender- geographic region standardized values), not the smoking status of the individuals leaving the model. Calculations for male former smokers are similar. For the first model year, the number of male former smokers in each age cohort is determined by the proportion of the overall model population (employees/health plan members plus adult dependents) who are in the specified age/gender/smoking status group. The number of male former smokers present in each subsequent year's age cohort is determined by subtracting the number of male former smokers of that age cohort who died or left the firm/health plan in the previous year and adding the number of replacement male former smokers. In addition, the number of male current smokers who quit smoking in the previous year is added, and the number of male former smokers who resumed smoking (recidivism) in the previous year is subtracted. The average age at smoking cessation is determined each year for each age cohort of former smokers. For the first year of the model, this age is assigned for each age cohort using data from the National Health and Nutrition Examination Survey (NHANES). In each subsequent year, the age at smoking cessation is determined as a weighted average using two components: (1 ) the number of male former smokers still in the model from the previous year times their average age at cessation in the previous year; and (2) the number of new male former smokers (current smokers who quit smoking in the previous model year) times their age when they quit smoking. The number of years since smoking cessation for male former smokers is then calculated as the current age of that age cohort minus the weighted average age at smoking cessation.

The annual number of cases of coronary artery disease, cerebrovascular disease, chronic obstructive pulmonary disease, and lung cancer and their attributable costs in each age cohort are determined for male former smokers in the same manner as described for male never smokers. Total annual health care costs and additional annual direct or indirect costs are also determined in the same manner as described for male never smokers. For former smokers, total annual health care costs are based on age, gender, and years since smoking cessation. Additional annual costs for male former smokers due to increased absenteeism or decreased productivity attributable to their former smoker status are also calculated each year. Annual absenteeism costs are calculated as the increased number of days of work missed by male former smokers compared to male never smokers multiplied by average daily wages (average hourly wages times eight hours/day) times the number of male former smoker employees in each age cohort. Annual productivity costs are calculated as the proportional decrease in productivity for former smokers as compared to never smokers times the average annual salary (average hourly wages times eight hours/day times five days/week times 50 weeks/year) times the number of male former smokers employees in each age cohort. Absenteeism and productivity costs are calculated only for workplace-based models, not for health plan models, and are discounted annually. Using default model parameters, absenteeism and productivity costs due for former smokers are zero.

The number of male former smokers dying or leaving the firm (quitting or retiring) or health plan in each age cohort is determined in the same manner as for male never smokers. As for male never smokers, the number of male former smoker employees dying or leaving the firm/health plan is determined and a proportional number of female dependents in the same age cohort also leave the model. In the model with replacement mode, male former smokers leaving the model are replaced with an identical number of new males in the same age cohort; smoking status for these new males is determined as described for male never smokers. In addition, each year a proportion of male former smokers resumes smoking. The probability of recidivism is based on years since smoking cessation. The number of male former smokers resuming smoking is subtracted from the male former smoker population in the subsequent model year and added to the male current smoker population in the subsequent year.

Calculations for male current smokers in each age cohort are similar to those for male never and former smokers. For the first model year, the number of male current smokers in each age cohort is determined by the proportion of the overall model population (employees/health plan members plus adult dependents) who are in the specified age/gender/smoking status group. The number of male current smokers present in each subsequent year's age cohort is determined by subtracting the number of male current smokers of that age cohort who died or left the firm/health plan in the previous year and adding the number of replacement male current smokers. In addition, the number of male current smokers who quit smoking in the previous year is subtracted, and the number of male former smokers who resumed smoking (recidivism) in the previous year is added.

The annual number of cases of coronary artery disease, cerebrovascular disease, chronic obstructive pulmonary disease, and lung cancer and their attributable costs in each age cohort are determined for male current smokers in the same manner as described for male never smokers. Total annual health care costs and additional annual direct or indirect costs are also determined in the same manner as described for male never smokers. Additional annual costs for male current smokers due to increased absenteeism or decreased productivity attributable to their current smoker status are also calculated each year in each age cohort as described for former smokers.

The number of smoking cessation attempts made annually by male current smokers in each age cohort are determined each year as the product of the number of male current smokers times the probability of attempting smoking cessation. This probability will be different in calculations performed with coverage of smoking cessation aids versus without if promotion of smoking cessation is included. The annual number of successful smoking cessation attempts is determined as the number of smoking cessation attempts times the average success rate. The annual cost for smoking cessation is determined as the number of smoking cessation attempts times the average cost per smoking cessation attempt, and is discounted annually. The smoking cessation success rate and the cost per cessation attempt will differ if smoking cessation aids are covered versus not covered and if promotion of ZYBAN^® is included in the model.

The number of male current smokers dying or leaving the firm (quitting or retiring) or health plan in each age cohort is determined in the same manner as for male never smokers. As for male never smokers, the number of male current smoker employees dying or leaving the firm/health plan is determined and a proportional number of female dependents in the same age cohort also leave the model. In the model with replacement mode, male current smokers leaving the model are replaced with an identical number of new males in the same age cohort; smoking status for these new males is determined as described for male never smokers. Male current smokers who were successful in smoking cessation are subtracted from the number of male current smokers in the following model year and added to the number of male former smokers. Calculations for female never, former, and current smokers are identical to those for their male smoking status except that outcomes and costs related to pregnancy and the impact of smoking status on pregnancy are included. For female never smokers in each age cohort, the annual number of pregnancies is determined by multiplying the number of female never smokers by the age- specific pregnancy rate. Pregnancy rates after age 44 are assumed to be zero. The annual number of miscarriages (spontaneous abortions) for never smokers is determined by multiplying the annual number of pregnancies by the never smoker miscarriage rate. The annual cost for never smoker miscarriages is the product of the number of never smoker miscarriages and the average cost per miscarriage. The annual number of never smoker-induced abortions is also determined as the product of the annual number of pregnancies and the age-specific induced abortion rate. The annual number of never smoker live births is the number of pregnancies minus the number of miscarriages and induced abortions. The annual number of never smoker low birth weight infants is the product of the annual number of live births and the rate of low birth weights. The annual cost for never smoker low birth weight infants is the product of the annual number of low birth weights times the average cost per low birth weight infants. Costs for never smoker miscarriages and low birth weights are discounted annually. Pregnancy outcomes and costs for former smokers and current smokers are determined as described for never smokers using event rates applicable to former and current smokers, respectively.

Identical calculations are performed for each age cohort with and without coverage of smoking cessation aids. The only difference in these calculations using default parameters is the rate of successful smoking cessation and the cost per smoking cessation attempt. Differences in the rate of successful smoking cessation leads to differences in numbers of former smokers versus current smokers in each age cohort and thus to the number of smoking related events and costs. The number of events and costs are summed across the six groups

(male or female by three smoking statuses) each year in each age cohort separately for calculations with coverage of smoking cessation aids versus without. The total number of events and costs for males, females, and overall are also determined until retirement age (from the start of the model to age 65) and for the entire model period (from the start of the model to age 85) in each age cohort separately for calculations with coverage of smoking cessation aids versus without. The differences in events and costs with smoking cessation aid coverage versus without are then determined for each age cohort. Number of events and costs are also summed across the six age groups separately for calculations with smoking cessation aid coverage versus without, and the overall differences in the two sets of calculations are determined.

Outputs are presented in the model using the results of the above calculations. Outcomes include summed differences in costs, specific disease events, smoking cessations, and deaths for 2, 5, 10, and 20 model years for calculations with coverage of smoking cessation aids versus without. The Benefit-Cost Analysis presents health care costs, indirect costs (absenteeism and productivity), and cessation costs summed to age 65 or to age 85 for calculations with coverage of smoking cessation aids versus without and the differences between the two sets of calculations. Benefit-cost ratios are calculated as the savings in health care costs or health care plus indirect costs with coverage of smoking cessation aids divided by the increased cessation costs with coverage of smoking cessation aids. For workplace models, the incremental cost of smoking cessation aid coverage per employee per year (PEPY) is the increased cost of smoking cessation with coverage of cessation aids in the first model year divided by the model population for the first model year. For health plan models, the incremental per member per month (PMPM) cost is the increased cost of smoking cessation with coverage of cessation aids in the first model year divided by the model population for the first model year, then this value divided by twelve. Internal rate of return (IRR) is a calculation performed using a function supplied with WINDOWS^® applications to determine the average annual discount rate which would result in identical costs with coverage of smoking cessation aids versus without coverage of smoking cessation aids. The Break-Even Analysis presents the cumulative increased cost for smoking cessation by year due to coverage of smoking cessation aids as compared to no coverage. The cumulative savings in either health care costs only or health care plus indirect costs due to coverage of cessation aids are also presented each year. Return on investment (ROI) calculated with this analysis is the ratio of the cumulative net cost savings with coverage of cessation aids (health care savings or health care plus indirect cost savings minus increased expenditures on smoking cessation) divided by the cumulative expenditures on smoking cessation. Until cumulative savings are greater than cumulative increased cessation expenditures, ROI is negative.

It will be understood that various details of the invention may be changed without departing from the scope of the invention. Furthermore, the foregoing description is for the purpose of illustration only, and not for the purpose of limitation-the invention being defined by the claims.

Claims

CLAIMS What is claimed is:

1. A computer-implemented system for estimating costs and benefits associated with a disease management intervention, the system comprising computer-executable instructions embodied in a computer-readable medium for performing steps comprising:

(A) providing a first graphical user interface on a computer display device for receiving population-specific data for a population to be offered treatment in a disease management intervention;

(B) projecting costs and benefits associated with the disease management intervention over a time period for the population based on the population-specific data; and

(C) outputting data indicative of the costs and benefits.

2. The system of claim 1 , wherein providing a first graphical user interface includes displaying a first window including a plurality of population- related input cells for selecting the population-specific data.

3. The system of claim 2, wherein the first window includes a plurality of model-related input cells for selecting model data for modeling the disease management intervention.

4. The system of claim 2, wherein at least some of the population- related input cells include default values for the population.

5. The system of claim 2, wherein displaying a first window includes displaying a plurality of control buttons actuated by an input device for displaying the cost savings data.

6. The system of claim 1 , wherein projecting costs and benefits includes projecting direct medical cost savings.

7. The system of claim 1 , wherein projecting costs and benefits comprises projecting direct and indirect cost savings.

8. The system of claim 1 , wherein projecting costs and benefits includes projecting disease cost savings.

9. The system of claim 1 , wherein outputting data indicative of the costs and benefits includes displaying cost savings in tabular format.

10. The system of claim 1 , wherein outputting data indicative of the costs and benefits includes displaying cost savings in graphical format.

11. The system of claim 1 , comprising performing a benefit-cost analysis associated with the disease management intervention and displaying results from the benefit-cost analysis.

12. The system of claim 11 , wherein performing the benefit-cost analysis includes estimating direct medical costs for population with the disease management intervention and without the disease management intervention.

13. The system of claim 1 , comprising performing a break-even analysis associated with the disease management intervention and displaying results of the break-even analysis.

14. The system of claim 13, wherein performing the break-even analysis includes determining a time period when estimated savings or benefits associated with the disease management intervention exceed the estimated costs.

15. The system of claim 13, wherein performing the break-even analysis includes estimating percentages of healthcare costs saved resulting from the disease management intervention at each of a plurality of time intervals measured since implementation of the disease management intervention.

16. The system of claim 13, wherein displaying the results of the cost-benefits analysis includes displaying the results in tabular format.

17. The system of claim 13, wherein displaying the results of the cost-benefits analysis includes displaying the results in graphical format.

18. The system of claim 1 , wherein the disease management intervention is a smoking cessation program.

19. The system of claim 1 , wherein the disease management intervention is an influenza vaccination program.

20. A method for estimating costs and benefits associated with a disease management intervention, the method comprising:

(A) providing a graphical user interface for receiving population- specific data for a population to be offered treatment in a disease management intervention;

(C) outputting data indicative of the costs and benefits.

21. The method of claim 20, wherein receiving population-specific data includes receiving a number of employees to be offered treatment in the disease management intervention.

22. The method of claim 20, wherein receiving population-specific data includes receiving a number of health plan members to be offered treatment in the disease management intervention.

23. The method of claim 21 , wherein receiving the population-specific data includes receiving salaries and job types for the employees to be offered treatment in the disease management intervention.

24. A computer-implemented system for estimating costs and benefits associated with a disease management intervention, the system comprising:

(A) means for displaying, on a computer display device, default parameters for a population to be treated or offered treatment in a disease management intervention;

(B) means for modifying at least some of the default parameters to reflect population-specific values entered by a user; and

(C) means for estimating costs and benefits associated with the disease management intervention based on the default parameters and the population-specific parameters entered by the user.

25. The system of claim 24, wherein the means for estimating costs and benefits calculates an internal rate of return for the disease management intervention.

26. The system of claim 24, comprising means for displaying data sources for at least some of the default parameters to the user.

27. The system of claim 26, wherein the means for displaying data sources displays bibliographic citations for the data sources.

28. The system of claim 24, comprising means for displaying definitions associated with the disease management intervention to the user.

29. A computer program product comprising computer-executable instructions embodied in a computer-readable medium for performing steps comprising:

(A) displaying a plurality of windows on a computer display device, at least some of the windows being adapted to receive population-specific data relating to a population to be treated in a disease management intervention;

(B) calculating costs and benefits associated with the disease management intervention over time for the population based on the population-specific data; and

(C) outputting data indicative of the costs and benefits to a user.

30. The computer program product of claim 29, wherein outputting data indicative of the costs and benefits includes printing the data using a printing device.

31. The computer program product of claim 29, wherein outputting data indicative of the costs and benefits includes displaying the data on a computer display device.

32. The computer program product of claim 29, wherein calculating costs and benefits over time comprises calculating cost savings data for the disease management intervention.

33. The computer program product of claim 29, wherein calculating costs and benefits over time comprises calculating an internal rate of return for the disease management intervention.