Click for next page ( 456


The National Academies | 500 Fifth St. N.W. | Washington, D.C. 20001
Copyright © National Academy of Sciences. All rights reserved.
Terms of Use and Privacy Statement



Below are the first 10 and last 10 pages of uncorrected machine-read text (when available) of this chapter, followed by the top 30 algorithmically extracted key phrases from the chapter as a whole.
Intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text on the opening pages of each chapter. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

Do not use for reproduction, copying, pasting, or reading; exclusively for search engines.

OCR for page 455
19 Cost-Effectiveness Information that compares the net medical cost of treatment for a person with alcohol problems with the treatment effect or effects is necessary for judging the desirability of support for such treatment and of insurance coverage for it. This analysis of costs and effects furnishes "cost-effectiveness" data that can be used in decision making. Cost-effectiveness analysis as a decision aid is, at base, a comparative tool. No strategy is literally cost-effective in isolation; the most that can be said is that one strategy is more cost-effective than another. If, for instance, strategy A is more costly and no more effective than strategy B. one can say that B is cost-effective compared with ~ However, the information on costs and effectiveness of the two strategies cannot tell us that strategy B is cost effective in isolation or in comparison, let us say, to a strategy of no intervention. The comparative aspect of cost-effectiveness is an important point, especially when different treatment strategies are analyzed. Strictly speaking, cost-effectiveness analysis as such can only provide information about relative desirability and not about whether anything at all should be done. There is, however, one common and natural use of the term cost-e,ffectiverzess as an absolute concept that is relevant to the treatment of alcohol problems. If a costly medical treatment lowers medical costs elsewhere in the system (either simultaneously or in the future), if it does so to such an extent that the value of the cost savings more than offsets the cost of the treatment, and if the treatment at least does no harm, then it might legitimately be said that the treatment is Cost effective and obviously desirable in the sense that it leads to negative net medical costs compared with a No treatments alternative. The cost-effectiveness analysis of treatment for alcohol problems has, to a considerable extent, pursued this question of cost offset (IOM, 1989~. In addition to the question of whether treatment itself actually saves money, compared with no treatment, there is the additional particular question of which treatment to select. The net cost of treatments varies; some treatment strategies may have higher net costs than others but may also have greater effectiveness. This situation then poses the classic cost-effectiveness question: how does the additional cost compare with the additional effectiveness? The discussion that follows, examines both of these cost-effectiveness questions: the possibility of cost offsets (and a negative net cost) of some treatment compared with no treatment and the cost-effectiveness of one treatment compared with another. Studies of Costs and Cost Offsets Because alcohol problems and alcohol dependence are associated with positive medical costs-for example, from alcohol-related illness and from accidents and injuries- it might be expected that successful treatment would lower those costs and an offset would occur. Jones and Vischi (1979) carried out an early review that displayed the pattern shown in subsequent research. They reviewed 25 studies that examined whether treatment for mental illness, alcohol abuse, or drug abuse reduced subsequent utilization of health services. Twelve of the 25 studies involved alcohol abuse; Jones and Vischi concluded that these studies showed that reductions did take place in either medical care utilization measures (e.g., hospital days, outpatient visits) or surrogate measures (e.g., sick days, sickness and accident benefits paid). They also observed reductions in the subsequent use of health services that ranged from 26 percent to 69 percent, with a median reduction of 40 percent. Methodological problems were noted in all 12 studies (e.g., limited time spans, small samples, lack of appropriate comparison groups); in addition, none of the studies was a randomized clinical trial that could have served to suggest causality. A further 455

OCR for page 455
456 BROADENING THE BASE OF TREATMENT FOR ALCOHOL PROBLEMS methodological problem was that a majority of the reviewed studies were conducted in HMOs or employer-based programs. This factor led the reviewers to raise questions about the generalizability of the findings as well as about the specific cause of the observed subsequent reductions in utilization (i.e., the cause may have been a characteristic of the treatment setting rather than of the treatment itself). Jones and Vischi discussed the implications of the findings for three areas of policy concern: (1) the setting in which treatment takes place, (2) the linkage of treatment for alcohol and drug abuse and mental disorders with general health services, and (3) health insurance coverage. They made no specific recommendations because of the limitations of the studies they reviewed. Their review, however, set the stage for continued studies of cost offsets in subsequent health services utilization and in insurance costs as a major strategy for obtaining improved coverage for treatment of persons with alcohol or drug problems or mental disorders. ~ ~ coworkers (1983), Holder (1987), and Holder and colleagues (1988). ~ . come or these more recent studies are reviewed by Saxe and These analysts attempted to determine whether treatment for alcohol problems provides any cost offsets and, if it does, whether the offsets are large enough to produce a negative net cost for ~, treatment. The types of cost offset studies that have been conducted and the nature of the results differ by the form of insurance coverage for treatment, the characteristics of the population at risk, and the process by which people are induced to seek treatment. Because there is reason to believe that the cost offset may vary depending on the process that stimulates treatment, it will be useful to treat each group of studies separately. Early studies of cost offsets looked at the experience of employee assistance programs (EAPs). latest programs identified workers who were problem drinkers and assisted and encouraged them in seeking treatment. A number of studies of (EAPs) have compared medical costs, disability costs, and sick days before and after an employee was successfully referred for treatment (e.g., Alander and Campbell, 1975~. There was consistent and unequivocal evidence of a change in outcome after treatment, either compared with the employee's previous behavior or compared with trends among a control group of persons with alcohol problems who did not seek treatment. Amone the treated employees, sick days and injury days fell, and inpatient costs dropped. _ ~ ,^ Sometimes the savings in the medical costs over the one- to two-year follow up period exceeded (in present value terms) the cost of the program; sometimes the offset was not complete. However, these studies typically followed small groups over a limited time period and did not measure all medical costs. A later set of studies used data from insurance and HMO plans and permitted longer follow-up periods and larger samples of employees and dependents. Holder and colleagues (Holder and Hallan, 1976, 1986; Holder and Those, 1986) examined several large samples (in one case, more than 1,500 observations) of persons with alcohol problems. Using sophisticated statistical models, they compared the total medical cost levels of the participants before and after treatment initiation. The Holder team found a universal pattern of decreases in hospital admission rates and average total medical costs, compared with past trends and with trends in a control group. Over a two- to three-year period, the full cost of the treatment was more than offset. Other studies (Brock and Boyajy, 1978; Sherman et al., 1979) showed similar results. The third group of studies involved were people who were receiving treatment from publicly funded programs (both the Veterans Administration and Medicaid) (Magruder-Habib et al., 1985; Calkins et al., 1986~. These studies failed to find a cost offset and may in fact have found a cost increase. This finding is in contrast to those from earlier studies of public clients (JWK International Corporation, 1976; Gregory et al., 1982; Becker and Sanders, 1984~; however, those studies projected substantial cost offsets but were not based on actual total expense data. Luckey (1987) conjectures that the differences

OCR for page 455
COST-EFFECTIVENESS 457 among the populations in these studies of publicly funded programs, compared with the EAP and insurance company studies, may explain the difference in results; that is, lower income people may have more chronic health problems before they initiate treatment and may have fewer incentives (or less opportunity) to maintain recovery. Thus, the overall picture regarding cost offsets is one of cost declines after treatment for people who are not poor, declines which are frequently large enough to offset the cost of treatment (Luckey, 1987~. Do these studies conclusively demonstrate that there is a negative net cost? Does the remarkable consistency of documented reductions in health care costs among insured groups strongly suggest the decreases are real? Unfortunately, despite the consistency of the results and the large sample sizes, there is still a methodological Achilles heel to the findings. In all of the studies, treatment recipients selected treatment for themselves (self-selection) and were not randomly assigned. That is, the studies do not tell us how those individuals would have fared had they not obtained treatment, information that could have been provided through the use of random assignment to experimental and control groups. The comparisons that were actually used in the studies were based either on the prior experience of those who sought treatment or on the experience of other control groups. As a result, the studies cannot report with confidence that treatment caused a difference in outcome (i.e., enabled later cost offsets) because the studies use two different groups of people for comparison: those who sought treatment and those who did not. Yet a question remains: If the treatment did not necessarily cause the decline, what did? One possible answer is that the people who sought treatment were ready to stop drinking anyway, with or without help. The other possibility is the statistical phenomenon called "regression to the mean," in which periods of unusually high levels of anything (whether it be rainfall, temperature, or the Dow-Jones index) are most likely to be followed by a return to the average: because an episode of treatment for alcohol problems is accompanied by high costs for all medical care, it is to be expected that low costs would follow. Some reviewers nevertheless conclude that there is some evidence to support the hypothesis that treatment for alcohol problems is cost-beneficial (Saxe et al., 1983; Holder, 1987; Luckey, 1987~. The benefits of treatment for alcohol problems are cautiously seen to be in excess of the cost of providing such treatment even if they fall short of what may be claimed. The caution stems from the subjective judgments that the reviewers must conjecture about the degree of spurious causation in the studies they reviewed because neither the studies nor the reviews provide any objective basis for determining the seriousness of the problem of spurious causation. One way out of this impasse would be to use the kind of reasoning suggested earlier. There are no controlled trials of cost offsets in real world settings, but there are the many before-and-after studies reviewed (Holder, 1987; IOM, 1989~. There are the many controlled trials of the effectiveness of treatment in clinical settings reviewed for this report. Can one link the two sets of results to come to a reasonable conclusion? There are two reasons why the answer may still be negative. One answer is brief and theoretical; the other is longer and is based on empirical fact. The theoretical problem is that the effectiveness demonstrated in controlled trials of treatment alternatives (there have been no controlled trials of cost-effectiveness) may not carry with it the kind of cost offset (at least in terms of size) suggested by the uncontrolled studies. This objection might be dismissed by appeal to the ethical and practical difficulties of randomizing large numbers of persons with alcohol problems to a long-term no-treatment trial. However, the strength of such an appeal is lessened by the fact that there have been some additional large sample studies that do use a type of real-world random assignment and that failed to find any cost offset (e.g., Hayami and Freeborn, 1981; Manning et al., 1986~.

OCR for page 455
458 BROADENING IlIE BASE OF TREATMENT FOR ALCOHOL PROBLEMS It is usually not easy to assign people who seek treatment to a no-treatment control group, nor is it possible to assign people who do not seek treatment to treatment. Considering the difficulties involved in randomized controlled trials, it will necessarily be impossible to get a reliable estimate of the cost offset for treatment of an "average" person with alcohol problems. In any case, because there is no way to compel that person to seek treatment, such an estimate does not really answer the relevant policy question: whether there is a cost offset for the type of person who can be induced to enter treatment by feasible policy interventions (e.g., better insurance coverage, the use of an EAP, court orders, an especially powerful way of marketing treatment). This question can be explored using a different methodology from that of randomized trials. Although one cannot assign individuals to treatment and no treatment study conditions, one can assign populations containing persons with alcohol problems to policy interventions that are thought to provide a stronger stimulus to such treatment than is provided in other environments. An example of such a study is provided by the Hayami and Freeborn (1981) analysis of different levels of out-of-pocket payment. This study raises serious questions about the existence of a relevant cost offset. In the Hayami and Freeborn study, more than 20,000 members of the Kaiser-Portland health plan were randomly assigned to different levels of out-of-pocket payment for subsequent treatment for alcohol problems. Half of the population was ~assigned" to receive coverage for such care free of charge (i.e., full insurance coverage); the others received the standard benefit of 50 percent coverage. Regardless of the level of insurance coverage, only a small percentage of the plan's members who were eligible for treatment of alcohol problems actually sought such treatment. The free care group used significantly more treatment services than the copayment group and also showed a modest improvement in condition. However, there was no difference between the two groups in the use of medical services in the posttreatment period. That is, there was no evidence of a cost offset associated with more generous insurance coverage and the higher level of use associated with it. A study that examined a wider range of insurance coverage but for mental health services in general and not just for alcohol problems treatment, is the RAND Corporation health insurance study carried out by Manning and coworkers (1986~. This study employed a randomized design that assigned 5,800 people to levels of insurance coverage ranging from free care to very high copayments. Compared with those who had free care, those who had copayments were less than half as likely to use mental health services. The follow-up period in this study (up to five years) was longer than in the Hayami-Freeborn study, but again no posttreatment cost offset was found. Yet these studies also have certain limitations that may affect the meaning one ascribes to their results. The follow-up period in the Hayamai-Freeborn study was short (although the follow-up period in the Manning study was not); the Manning team's study was not limited to treatment for alcohol problems. These limitations mean that one can only say that these studies failed to find any cost offset, not that a cost offset does not exist. Nevertheless, these results raise serious questions about the inevitability of cost offsets from more extensive insurance coverage, and they surely do not prove that better coverage saves enough money to offset its cost. It would be possible to reproduce these studies, even in private insurance settings, by phasing in more generous benefits for treatment of alcohol problems on a random or nonsystematic basis. Alternatively, it would be possible to examine the experience of people with other, exogenous influences on their use of services (e.g., the distance to or availability of treatment facilities, mandated referrals by the criminal justice system) to carry out population-based real world analysis of cost offsets. In the current state of knowledge, however, there is still some doubt that the net cost of treatment for alcohol problems would surely be negative for some population. "Probably" is a reasonable adverb to attach

OCR for page 455
COST-EFFECTIVENESS 459 to cost offsets, but reasonable people can differ. One should hasten to add that suspecting that treatment might have a net positive medical cost hardly suggests that treatment is undesirable. It only means that attention must De rocusea on ant; owner Do; [U ~O gained from reducing alcohol problems. Those benefits, in terms of increased productivity, reduced danger to others, and the value placed by the person with an alcohol problem and his or her family on the reduction or elimination of that problem, are all surely positive. It is on the value of these benefits that one should rest the case for the advantage of treatment for alcohol problems (Harwood et al., 1984; Harwood, 1988~. Cost-Benefit and Cost-Effectiveness Analysis The discussion above indicated that there is some, although not conclusive, evidence that spending money on treatment today lowers future medical costs. Whether there is, in any case, sufficient cost savings in the future to more than offset current costs is not known. Compared with no treatment, the use of some treatment (i.e., brief intervention) has a chance of lowering cost. But what about treatment beyond the level of brief interventions? More intensive treatment, even if it should be discovered to have some cost offset, almost surely will have a positive net cost. Yet a more costly treatment may still be appropriate if it is more effective, if it provides more overall person and social benefit. Conversely, a treatment that provides some additional benefit may still not be desirable if it costs too much. For effective judgements, what we need to know is the cost effectiveness of alternative forms of treatment, a comparison of additional costs and additional benefits (Luckey, 1987~. There has been almost no formal examination of the cost-effectiveness of alternative treatments (Pauly, 1988~. As discussed in Chapter 5, there is substantial uncertainty about whether interventions beyond the brief intervention level have any additional benefit when indiscriminately applied; there has been virtually no analysis of their additional cost. It is probably plausible to assume that brief interventions are cost-effective compared with no treatment. (The cost of a brief intervention is so low that a negative net cost is achieved through only a small cost offset. Moreover, even if it should have a positive net cost, that cost is likely to be low enough that the benefit provided by such interventions would usually be judged to be worth the cost.) Beyond brief interventions, however, there is simply no basis for determining the additional costs compared with the additional benefits because there is much uncertainty about benefits and virtually complete ignorance about costs. Thus, it is not known whether there are any additional cost offsets attached to treatment offered at levels beyond that of brief intervention. Also at issue is whether positive costs, if present, are exceeded by benefits. Holder and colleagues (1988) have indicated the sort of study needed to answer this question: one with standard measures of outcomes, standard means of treatment, and random assignment. Absent this type of analysis, one is forced to make informal comparisons by relying on the effectiveness literature and guessing what the comparative costs would be. The best (and most conclusive) example of such an analysis is based on the "more intensive-less intensives split. As noted earlier (see Chapter 5 and Appendix B), when forms of care that are more intensive than brief interventions are applied to undifferentiated populations of people with alcohol problems, there is usually no significantly greater effect than there would have been using less intense treatments. Under the reasonable assumption that more intensive care is more costly than brief interventions or outpatient care, one could conclude that more intensive care is less cost-effective than those less costly alternatives. This is not a tautology; if the improvement in effect with more intensive, more costly care had been positive and large, one might well have concluded that more costly care was more cost-effective. With zero effects and high costs,

OCR for page 455
460 BROADENING THE BASE OF TREATMENT FOR ALCOHOL PROBLEMS however, it is a foregone conclusion that it more costly care is not more cost-effective. More generally, given any set of treatment approaches that yield comparable outcomes-and this uniformity is the rule rather than the exception in effectiveness studies for alcohol treatment the least costly intervention in the set is always the one to be preferred on cost-effectiveness grounds. Holder and colleagues (198$) suggest some possible exceptions to this conclusion. They note that "it may be that certain types of programs/facility combinations better retain persons with chronic, severe alcohol problems in treatment or some form of institutional care (24-hour care) is necessary for patients with more severe physical disability resulting from chronic drinking" (p. 7~. They conclude, however, that at present "this is more speculation than science." Are there circumstances requiring qualifications to the conclusion that residentiaVinpatient care is less cost-effective than outpatient alternatives? Most of the argument, as usual, turns out to be about effectiveness, and not about cost. Many of the studies that show no differences are old and do not represent the type of treatment program currently being used in many residential programs. Others reveal that certain kinds of persons may do better in a residential setting, especially persons with less social stability (Kissin et al., 1971) or with a dysfunctional family or work relationships (McLellan et al., 1982~. There was one random assignment study where residential care proved more effective for socially disadvantaged persons (Wanberg, Horn, and Fairchild, 1974~. Against these exceptions, however, is mounted a group of studies that show that outpatient or day treatment settings offer effective treatment that is at least as good as-if not better than that offered in inpatient/residential settings. The finding that an inpatient program is sometimes more effective does not, of course, mean that it is also cost-effective. Put more bluntly, if residential programs for people with dysfunctional family relationships are a little better but a lot more costly than outpatient programs, they may still be undesirable because the benefits are not worth the cost, that is, are not large enough to "justify" the cost. To know for sure, it is necessary to assign values to benefits, something health care analysts try to avoid if at all possible. Yet ducking the value judgment may not always be an option. The value judgement may need to incorporate considerations other than cost effectiveness. For example, for persons with severe and chronic alcohol problems, Holder and colleagues (1988), concluded that there is no treatment intervention with a full cost offset. Although residential or inpatient settings may be more costly and no more effective than outpatient approaches for people who have already decided to initiate treatment, they may attract more people into treatment than the "live in the community/face your friends alternative. If treatment does have a cost offset, spending more per person to attract more people to treatment may actually result in a lower net cost for a given population of people with alcohol problems. However, this "marketing" or "attractiveness" dimension which reflects the benefits individuals perceive will be forthcoming from a particular treatment-has not been investigated at all. More evidence may show that this ability to attract persons to treatment could be the strongest argument, if there is to be any strong argument to be made, for treatment in the residential setting. The intensity of treatment (as distinct from the treatment setting) again displays the No difference" results. There have been few direct comparisons of brief interventions (e.g., brief counseling) compared with intensive residential treatment. A recent study by Chick and coworkers (1988) showed better outcomes for extended treatment by the second follow-up year but did not provide estimates of the difference in the cost of the two approaches. Finally, even informal comparisons of treatment methods are simply not possible on cost-effectiveness grounds (Holder et al., 1988~. In summary, for a heterogeneous group of persons with alcohol problems, no one treatment has been shown to be more effective than any other, whether the treatment

OCR for page 455
COST-EFFECTIVENESS 461 variable used for study is the setting, intensity, modality, technique, or process. Thus, it cannot be ruled out that, even in some of the cases in which more intensive treatment is effective, the cost per unit of "effect" is disproportionately large. Better information is sorely needed. Matching and Cost-Effectiveness Matching individuals with alcohol problems to particular treatments seems to increase average treatment effectiveness. Matching is the strategy preferred by this committee. However, there is no literature that bears on the question of cost-effectiveness in matching programs because there is relatively little evidence on the question of the effectiveness of matching itself (Holder et al., 1988~. Posing the cost-effectiveness question requires first that one specify both the total level of resources to be applied to a population of persons with alcohol problems who have different needs and how those resources are distributed among different individuals. One might imagine that in theory there is some distribution, given a resource constraint, that "maximizes effectiveness" but what is it? Without such knowledge, the allocation of resources based on clinical judgment about appropriate matching could lead to lower aggregate "effectiveness" than if there were no matching. A related issue concerns the cost-effectiveness of various matching procedures. The "best" treatment for a particular individual, given his or her condition, may not be the most cost-effective one. As the dismal comment of the previous paragraph suggests, it may even be best to do nothing, or to do very little, rather than match treatment to someone for whom the most effective treatment has nevertheless low effectiveness and high cost. Still a third cost-effectiveness question concerns the cost of matching itself. Careful assessments use up real resources, and the benefits from better matching may not be worth the cost of determining who needs which treatment. One suspects this may be particularly the case with less severe problems, where it may be less costly to give everyone a standard brief intervention and perform assessments much later and only for those people who seem to be responding poorly. Conclusions here all depend on information that has not even been a subject of speculation, much less an object of fact: the cost of assessment; the consequences (in both cost and outcome) of mismatches, whether false positives or false negatives; and the cost in real-world settings of managing clinically ideal matching programs. If overlaid on the problem is the notion that, from a policy perspective, all one can do is encourage matching or try to structure insurance and financing policy to permit matching, the situation becomes even more complex. How effective, and how cost effective, would it be to replace blanket 28-day program insurance coverage with coverage for managed care? Would case managers really be able to avoid enough long inpatient stays to pay for easier access to outpatient care? Or would one see the more common real-world setting in which outpatient treatment expands when coverage is extended but the long inpatient stays fail to contract? Conclusions and Recommendations Although there are some critical findings from cost-effectiveness research that have helped the committee to formulate its recommendations on financing treatment for alcohol problems, it is painfully obvious that much of what we need to know is not known. Defensible measures of the cost-effectiveness of treatment beyond brief interventions for particular populations are simply not available; consequently, the data on cost-effectiveness are still insufficient for unambiguous policy guidance. Accordingly, the committee sees it

OCR for page 455
462 BROADENING THE BASE OF TREATMENT FOR ALCOHOL PROBLEMS as appropriate to recommend an intensive program of research to determine the costs of alternative treatments for persons with alcohol problems relative to the benefits of such treatments. The committee recommends that thefederal government sponsor an expanded program of research and ar~aisis directed at discovering the costs, effectiveness, and responsiveness to insurance coverage of the various treatment strategies now in use and the matching strategy, which the committee favors, that is now being introduced. The program should include studies of the impact of all types of third party funding on the utilization of different forms of treatment for alcohol problems. Agencies which should be involved in developing this program of research would appropriately include the National Center for Health Services Research, the Health Care Financing Administration, and the National Institute on Alcohol Abuse and Alcoholism. None of these agencies currently has such an effort under way. The recent NIAAA (1988) program announcement outlining areas of interest and requesting applications for research on financing issues and the costs of various treatment services and settings represents an initial step toward developing such a program. Much more, however, needs to be done (Wallen, 1988: IOM, 1989~. In particular, studies should be undertaken to determine the question of whether it is necessary to provide discrete coverage for brief interventions in order to bring more persons into treatment and to compare alternative detoxification and primary rehabilitation strategies. As recommended in Chapter 18, there should also be an expansion of the federal government's services research effort to establish the cost-effectiveness of alternative strategies and models for treating alcohol problems. Studies of treatment effectiveness should not be undertaken without a consideration of the comparative cost-effectiveness questions (Holder et al, 1988~. Thus, treatment outcome studies should routinely include mode of payment and cost data in order to begin to define the relative cost-effectiveness of various approaches to treatment of alcohol problems. REFERENCES Alander, R., and T. Campbell. 1975. An evaluative study of an alcohol and drug recovery program, a case study of the Oldsmobile experience. Human Resources Management 14:14-19. Becker, F. W., and B. K Sanders. 1984. The Illinois Medicare/Medicaid Alcoholism Services Demonstration: Medicaid Cost Trends and Utilization Patterns-Managerial Report. Prepared for the Illinois Department of Alcohol and Substance Abuse. Center for Policy Studies and Program Evaluation, Sangamon State University. Springfield, Ill. Brock, C.P., and T. G. Boyajy. 1978. Alcoholism within Prepaid Group Practice HMO's. Prepared for the National Institute on Alcohol Abuse and Alcoholism. Washington, D.C.: Group Health Association of America. Calkins, R., E. Kemp, J. Lock, J. Ramsey, and M. Cohen. 1986. Enhanced Evaluation of the Michigan Medicare/Medicaid Alcoholism Services Demonstration Project-Medicaid Costs and Utilization. Prepared for the Michigan Department of Substance Abuse Services. Lansing, Mich.: Michigan Department of Substance Abuse Services. Chick, J., B. Ritson, J. Connaughton, A. Stewart, and J. Chick. 1988. Advice versus treatment for alcoholism: A controlled trial. British Journal of Addictions 83:159-170. Gregory, D., R. Jones, and R. Rundell. 1982. Feasibility of an alcoholism health insurance benefit. Pp. 195-202 in Currents in Alcoholism: Recent Advances in Research and Treatment. vol. 7, M. Galanter, ed. New York: Grune and Stratton. Hardwood, H. J., D. M. Napolitano, P. L. Kristiansen, and J. J. Collins. 1984. Economic Costs to Society of Alcohol and Drug Abuse and Mental Illness: 1980. Prepared for the Alcohol, Drug Abuse, and Mental Health Administration. Research Triangle Park, N. C.: Research Triangle Institute.

OCR for page 455
COST-EFFECTIVENESS 463 Hardwood, H. J. 1988. The burden of alcoholism on business: A simulation of alternative policies. Prepared for the IOM Committee to Study Treatment and Rehabilitation Services for Alcoholics and Alcohol Abusers. Hayami, D. E., and D. K Freeborn. 1981. Effect of coverage on use of an HMO alcoholism treatment program, outcome, and medical care utilization. American Journal of Public Health 71:1133-1143. Holder, H. D., 1987. Alcoholism treatment and potential health care cost saving. Medical Care 25:52-71. Holder, H. D., and J. B. Hallan. 1976. A Study of Health Insurance Coverage for Alcoholism for California State Employees: Two Year Experience Summary Report. Prepared for the National Institute on Alcohol Abuse and Alcoholism. Raleigh, N.C.: H-2, Inc. Holder, H. D., and J. B. Hallan. 1986. Impact of alcoholism on total health care costs: A six year study. Advances in Alcohol and Substance Abuse 6~1~:1-15. Holder, H. D., and J. O. Those. 1986. Alcoholism treatment and total health care utilization and costs: A four year longitudinal analysis of federal employees. Journal of the American Medical Association 256:1456-1460. Holder, H.D., R. Longabaugh, and W.R. Miller. 1988. Cost and effectiveness of alcoholism treatment using best available information. Prepared for the IOM Committee to Study Treatment and Rehabilitation Services for Alcoholics and Alcohol Abusers. Institute of Medicine. 1989. Prevention and Treatment of Alcohol Problems: Research Opportunities. Washington, D.C.: National Academy Press. JWK International Corporation. 1976. Benefit-Cost Analysis of Alcoholism Treatment Centers. Prepared for the National Institute on Alcohol Abuse and Alcoholism (NLAAA). Rockville, Md.: NL\AA. Jones, K, and T. Vischi. 1979. Impact of alcohol, drug abuse and mental health care utilization. Medical Care 17 (12, Suppl.~:1~2. Kissin, B., A. Plautz, and W. H. Su. 1971. Social and psychological factors in the treatment of chronic alcoholism. Journal of Psychiatric Research 8:13-17. Luckey, J. W. 1987. Justifying alcohol treatment on the basis of cost savings: The offset literature. Alcohol Health and Research World 12:8-15. Magruder-Habib, K, J. Luckey, V. Mikow, P. Barrow, and H. Felts. 1985. Effects of Alcoholism Treatment on Health Services Utilization Patterns. Technical Report IIIR-82~26. Washington, D.C.: Veterans Administration. McLellan, A. T., L. Luborsky, L. O'Brien, C. Woody, and K S. Druley. 1982. Is treatment for substance abuse effective? Journal of the American Medical Association 247:1423-1428. Manning, W. G., K B. Wells, N. Duan, J. Newhouse, and J. Ware. 1986. How cost sharing affects the use of ambulatory mental health services. Journal of the American Medical Association 256:1930-1986. National Institute on Alcohol Abuse and Alcoholism (NIAAA). 1988. Program Announcement: Research on Economic and Socioeconomic Issues in the Prevention, Treatment, and Epidemiology of Alcohol Abuse and Alcoholism. Rockville, Md.: NIAAA. Pauly, M. V. 1988. Delivery and financing of alcoholism treatment. Presented at the National Institute on Alcohol Abuse and Alcoholism/National Center for Health Services Research "Workshop on Health Economics of Prevention and Treatment of Alcohol Problems," Rockville, Md., September. Saxe, L, D. Dougherty, K Esty, and M. Fine. 1983. The Effectiveness and Costs of Alcoholism Treatment. Washington, D.C.: U.S. Congress, Office of Technology Assessment. Sherman, R. M., S. Reiff, and A. B. Forsythe. 1979. Utilization of medical services by alcoholics participating in an outpatient treatment program. Alcoholism: Clinical and Experimental Research 3:115-120. Wanberg, K W., J. L. Horn, and D. Fairchild. 1974. Hospital versus community treatment of alcoholism problems. International Journal of Mental Health 3:160-176.