Chapter 1 
 Chapter 2
 Chapter 3
 Chapter 4
 Chapter 5
 Chapter 6
 Chapter 7
 References
CHAPTER 2.  ECONOMIC EVALUATIONS OF ALCOHOL AND DRUG ABUSE TREATMENT PROGRAMS

2.1    Economic Evaluations of Health Care Interventions

            There is a vast literature on economic evaluation in health care.  Studies have been conducted by economists, medical researchers, clinicians, and multidisciplinary teams containing several of these parties.  The studies have different focuses, and vary greatly in quality, as methodological critiques have shown.  In recent years, there have been efforts to standardize economic evaluations of health care interventions to raise their quality and to make them more comparable.  This chapter draws heavily from three major references on the economic evaluation of health care (Gold, et al., 1996; Drummond, et al. 1997; and Sloan 1996).

2.2    Types of Health Care Evaluations

            As Drummond, et al. (1997) explain, there are at least six general types of health care evaluations.  Table 2.1 uses two questions – (1) are there comparisons of two or more alternatives and (2) are both costs (inputs) and consequences (outputs) of the alternatives examined – to define a six-cell matrix for health care evaluations.

            Outcome description studies indicated in cell 1A only examine the consequences of a single service or program.  They do not compare services or programs.

            The “cost of illness” or “burden of illness studies indicated in cell 1B only describe the cost of the disease to society.  They are not full economic evaluations because alternatives are not compared.  The “cost of illness” studies of alcoholism and drug abuse will be discussed in Chapter 4.

Table 2.1.  Distinguishing Characteristics of Health Care Evaluation

Are both costs (inputs) and consequences (outputs) of the alternatives examined?

            Cost-outcome description studies examine the costs and outcomes of a single service or program.  While these types of studies are often called economic evaluations in the alcohol and drug abuse treatment literature, they are really not economic evaluations because they do not compare the cost-outcome relationship of two or more alternatives.

            “Efficacy” or “effectiveness” evaluations in cell 3A compare the consequences or outcomes of two or more alternatives, but they do not consider the costs of the alternative services or programs.  “Efficacy” studies address the question – does the service or program do more good than harm to those people to whom it is offered?  These studies take into consideration that patients do not always fully comply with recommended treatments.

            Cost analysis studies included in cell 3B only compare the costs of alternative services or programs without considering the consequences or outcomes.  While such studies are useful, they are not full economic evaluations and they cannot be used to answer efficiency questions.

            There are four general types of full economic evaluations, which simultaneously consider the consequences, and costs of two or more alternative services or programs, as shown in cell 4.  Economic analysis deals with both the costs and consequences of alternative services or programs.  It concerns itself with choices and relative efficiencies.  Resource scarcity necessitates that choices must be made in all areas of human activity.  As Drummond, et al. (1997, p. 8) explains:  “Economic analysis seeks to identify and to make explicit one set of criteria which may be useful in deciding among different uses for scarce resources.”  The basic tasks of any economic evaluation are to identify, measure, value, and compare the costs and consequences of the alternatives being considered.

            Figure 2.1 illustrates that an economic evaluation is usually formulated in terms of a choice between competing alternatives.  In the figure, Program A is being compared to Comparator B, which could be an alternative program or could be the alternative of doing nothing.  It is important to note that when assessing programs A and B, it is the difference in costs compared with the difference in consequences that matters.  There are four major types of economic evaluation, each dealing with costs but differing in the way that the consequences of health care programs are measured and valued.

Figure 2.1 Economic Evaluation

2.2.1  Cost-minimization analysis (CMA)

            When two or more programs generate the same outcome, cost-minimization analysis can be used to guide resource allocation decisions.  By estimating and comparing the costs of alternative programs, the analyst can identify which program costs less to achieve a given outcome.  Because outcomes are considered identical, cost-minimization analysis is actually a special form of cost-effectiveness analysis.  Because alcohol and drug abuse treatment programs involve multiple outcome measures, cost-minimization analysis is rarely used to evaluate such programs (French, 2001).

2.2.2  Cost-Effectiveness Analysis (CEA)

            Cost-effectiveness analysis is the most common economic evaluation method employed in health care.  In this form of analysis, costs are measured in dollars and consequences or outcomes are measured in the most appropriate natural effects or physical units such as life-years gained, disability-days saved, etc.  In alcohol or drug treatment studies the outcome measure may be the degree of abstinence or the degree of drug or alcohol use.  Cost-effectiveness analysis can be performed on any alternatives that have a common effect.  As noted above, economic evaluations use incremental analysis that relates the differences in the cost between two alternatives (DC) to the differences in program outputs (DO).  The results of such comparisons may be stated either in terms of cost per unit of effect  or in terms of effects per unit of cost .

            The central purpose of CEA is to compare the relative efficiency of different interventions (e.g., programs) in creating better outcomes.  A cost-effectiveness analysis provides information that can help decision makers sort through alternatives to determine which ones best serve their programmatic and financial needs (Gold, et al. 1996).  CEA can also be used by program managers to determine the relative efficiency of the treatment components of their programs.  Such analysis can help them to redesign their program to improve its overall efficiency (NIDA, 1999).

            There are dangers in making cross-study comparisons of CEAs (Gold et al., 1996).  Ideally, one would wish to be able to array all CEAs  on a “league table,” where cost per health effect gained using one intervention is assessed side by side with the cost per health effect gained using others.  Decision makers could then make resource allocation decisions based on the relative efficiencies of different health investments.  The problem is that studies vary with respect to the analyst’s perspective (which will be discussed in the next chapter) and there are no uniform standards for CEAs.  For meaningful comparisons to be made, the studies must use the same methodology, have the same analysts perceptive, define and measure costs and outcomes similarly, and treat similar groups of patients.  According to NIDA (1999), it is easy to find an apparent difference in the cost-effectiveness of different program components or different programs, but it is much harder to show that the difference is real.  Gold, et al. (1996, p. 38) explains:

The ideal table of cost-effectiveness ratios would list all existing and potential programs at all feasible levels of program scale and intensity, for all population and patient groups, compared to all feasible alternatives.  This table would provide complete cost-effectiveness information for decision making from which a technically optimal allocation of resources could be identified, given a budget constraint.

As shall be discussed in subsequent chapters, the noncomparability of methods across evaluation studies of alcohol and drug abuse programs and treatments has complicated the original intent of this paper to identify the most efficient types of alcohol and drug abuse programs and treatments.  We shall base our estimates on the effectiveness of such programs on those studies employing the soundest methodology.

            CEA studies generally focus on a single outcome measure.  Furthermore, no attempt is made to value the outcome in terms of dollars or utility.  In the case of programs having multiple outcomes, CEA becomes problematic.  Some CEA health intervention studies present an array of output measures alongside cost and leave it to the reader to form his own view of the relative importance of these.  Some analysts have used the term “cost-consequences analysis” for the form of CEA (Gold, et al., 1996).

2.2.3  Cost-utility Analysis (CUA)

            Under cost-utility analysis the outcomes or consequences of health care interventions are valued in terms of health state preference            scores or utility weights.  In this sense utility refers to the preferences individuals or society may have for any particular set of health outcomes at a point in time or a profile of states through time.  There are several different methods to measure health state preferences (see Drummond, et al., 1997, chapter 6).

            CUA is a particularly useful technique because it can assess changes in the quality of life as well as changes in the length of life resulting from health care interventions.  The most common CUA outcome is quality-adjusted life-years (QALYs), which is calculated by adjusting the length of time affected through the health outcome by the utility value (on a scale of 0 to 1) of the resulting level of health status (Drummond, et al., 1997, pp. 15-17).

            One might think that CUAs of alcohol and drug abuse treatment programs would be commonplace since the technique provides a generic outcome measure, which could be summed for multiple outcomes and used for comparison of cost and outcomes in different programs.  Our literature search revealed that few if any CUA of alcohol and drug abuse treatment programs has been conducted or published.  There are several possible explanations for this.  Whereas the analyst’s perspective in most health care intervention evaluations is from the perspective of the patient or society as a whole, the analyst’s perspective in most economic evaluations of alcohol and drug abuse treatment programs is from the perspective of the government or taxpayer.  The health care studies focus on the changes in the value of outputs to the individual patient.  The alcohol and drug abuse program evaluations focus on the cost-savings to tax payers.  The health care evaluations use micro data and examine the long-term or lifetime effects of interventions.  The alcohol and drug abuse studies frequently use aggregate data and examine the short-term effects per episode of treatment.

2.2.4 Cost-benefit analysis (CBA)

            Cost-benefit analysis evaluates changes in the outcomes of alternative health care interventions in money terms.  The most common methods of assigning dollar value to health consequences are willingness to pay and human capital (Gold et al. (1996, p. 40).  Willingness to pay can be derived from a survey approach known as “contingent valuation, or it can be inferred from decisions actually made that involve tradeoffs between health and money.  Human capital essentially values health in terms of the productive value (i.e., earnings) of people in the economy.  Many health care researchers have avoided CBA on the grounds that the methods used to monitize the value of outcomes raise ethical concerns; they prefer CEA, albeit at some sacrifice of generalizability (Gold, et al., 1996).

            In theory, CBA can be used to ascertain whether the beneficial consequences of a program justify the costs and to determine how government funds should be allocated to health care programs and non-health care programs to obtain the highest rate of return on investments.  CBA provides information on the absolute benefit of programs, in addition to information on their relative performance.  Cost-benefit analysis provides an estimate of the value of resources used up by each program compared to the value of resources the program might save or create.  Often CBA implicitly assumes that each program is being compared to a do-nothing alternative that entails no costs and no benefits, even when this is not true (Drummond et al. 1997, p. 13).  Under CBA all relevant costs and consequences are supposed to be valued in money terms.  In practice, many of the cost-benefit analyses published to date are limited to a comparison of those costs and consequences that can easily be expressed in money terms.

            Alcohol and drug abuse treatment programs lend themselves to some form of CBA since they produce measurable monetary outcomes like increased days of legitimate employment and decreased job absences.  In addition, such programs may reduce patients’ use of food stamps, public health services, and other public assistance – a potentially huge cost savings.  Substance abuse programs also produce beneficial indirect or secondary effects on crime related costs.  In Chapter 3, we will review the so-called cost-of-illness studies dealing with the problems of alcohol and drug abuse and the potential cost-savings associated with programs to treat their problems.

            The difficulty associated with estimating the dollar value of program outcomes such as abstinence, reductions in alcohol and drug use, and improved family life has resulted in relatively few CBA studies in the alcohol and drug abuse treatment literature (French, 1995 and 2001).  There have been, however, a number of cost-offset analysis studies which examine dollar changes in income, reduced social services and associated costs, reduced expenditures for other services for alcohol and drug abusers or their families, and increased economic productivity resulting from alcohol and drug abuse treatment programs (Holder, 1987 and Holder, Longabough, Miller, and Rubonis, 1991).  In reality, cost-offset analysis is a partial cost-benefit analysis because it compares the cost of a program with the dollar value of avoided future health care costs (French 2001). 

            One final point should be made with respect to cost-benefit analysis.  CBA results are often expressed as a ratio (Dbenefits/Dcosts).  This approach should be avoided because the costs of averted illness are sometimes viewed as a benefit and placed in the numerator or sometimes viewed as a negative cost and placed in the denominator.  It is preferable to calculate the net benefit of a program by subtracting the Dcosts from the Dbenefits, to make studies more comparable (Gold et al. 1996).