Psychiatric Annals

Treatment of Alcohol and Drug Addiction 

Treatment Outcomes for Abstinence-based Programs

Norman G Hoffmann, PhD; Norman S Miller, MD

Abstract

Effective treatment for alcohol and drug dependence (addiction) has been available for many years in the private and public sectors hi the United States. H is documentad that thousands of patients have achieved and maintained recovery as measured in long-term abstinence and enhanced quality of life through the benefits of structured abstinence-based treatment programs. The treatment outcome studies described here involve large samples at multiple sites. The results regarding treatment effectiveness are analyzed with reliable statistical techniques and are generalizable to comparable treatment populations.

Unfortunately, many physicians (including psychiatrists) do not generally refer their patients to abstinence-based treatment for alcohol and drug addiction. Perhaps the most telling illustration is that less than 12%. of almost 20,000 alcoholics (drug addicts) are referred by physicians to monitored treatment programs of cohorts presented in this article.1 The membership of alcoholics in AA is over a million in the US alone, and the prevalence rate of alcoholics and drug addicts in medical and psychiatric practices range from 25% to 80%. 2 The conclusion is that many psychiatrists are not inclined toward developing skills in the specific treatment of addictive disorders, and have little knowledge of effective treatments for them.

The purpose of this study is to provide the psychiatrist with reliable data on outcome of abstinencebased models of treatment. It is hoped that a careful examination of these data will motivate the psychiatrist to learn more about this form of treatment.

A survey found that the abstinence-based method is used by over 95% of the treatment facilities for alcohol and drug addiction in the US (unpublished data, Roman PP). The abstinence-based method is derived from an integration of professionally delivered psychotherapy services and recovery principles of Alcoholics Anonymous, namely, that alcoholism (drug addiction) is an independent disorder that requires specific treatment. Therefore, the term often used to describe this method is the "12 step" model after the 12 steps of AA.

The 12-step model of treatment can be provided in outpatient and inpatient settings, and typically is structured in form and content. Both group and individual therapies are scheduled regularly. Structured treatment consists of a primary or intensive phase in which group and individual sessions are given daily for weeks either as an inpatient or outpatient, followed by a less intensive and tapered "aftercare period" of weekly attendance for a period of months to a year or two in an outpatient setting. Regular prescribed "doses" of educational group and individual therapies are given on an individualized basis depending on the severity and associated psychiatric and medical comorbiditics of the patients.3

METHODS

CATOR, Comprehensive Assessment and Treatment Outcome Research, is the largest independent evaluation service for the chemical dependency field in the US. CATOR, as a division of New Standards, Inc., is not government supported, is not part of a governmental agency, and is not owned by a treatment provider. CATOR functions as a clinical auditor in evaluating the efficacy of programs in achieving recovery from alcohol and other drug addictions and documenting correlates of such recovery.

The programs monitored tend to be variations of the Minnesota Model,3 but CATOR has evaluated psychiatrically based programs and aversion therapy programs as well. The common element to all of these programs is that the ideal goal is to assist the patient to remain abstinent from all moodaltering drugs of abuse. To be admitted to these programs, one must have a diagnosable addictive disorder. Thus, problem drinkers and drug misusers should not be admitted to these programs nor included in this patient registry system.

The patient data presented here are from the version of the general registry system…

Effective treatment for alcohol and drug dependence (addiction) has been available for many years in the private and public sectors hi the United States. H is documentad that thousands of patients have achieved and maintained recovery as measured in long-term abstinence and enhanced quality of life through the benefits of structured abstinence-based treatment programs. The treatment outcome studies described here involve large samples at multiple sites. The results regarding treatment effectiveness are analyzed with reliable statistical techniques and are generalizable to comparable treatment populations.

Unfortunately, many physicians (including psychiatrists) do not generally refer their patients to abstinence-based treatment for alcohol and drug addiction. Perhaps the most telling illustration is that less than 12%. of almost 20,000 alcoholics (drug addicts) are referred by physicians to monitored treatment programs of cohorts presented in this article.1 The membership of alcoholics in AA is over a million in the US alone, and the prevalence rate of alcoholics and drug addicts in medical and psychiatric practices range from 25% to 80%. 2 The conclusion is that many psychiatrists are not inclined toward developing skills in the specific treatment of addictive disorders, and have little knowledge of effective treatments for them.

The purpose of this study is to provide the psychiatrist with reliable data on outcome of abstinencebased models of treatment. It is hoped that a careful examination of these data will motivate the psychiatrist to learn more about this form of treatment.

A survey found that the abstinence-based method is used by over 95% of the treatment facilities for alcohol and drug addiction in the US (unpublished data, Roman PP). The abstinence-based method is derived from an integration of professionally delivered psychotherapy services and recovery principles of Alcoholics Anonymous, namely, that alcoholism (drug addiction) is an independent disorder that requires specific treatment. Therefore, the term often used to describe this method is the "12 step" model after the 12 steps of AA.

The 12-step model of treatment can be provided in outpatient and inpatient settings, and typically is structured in form and content. Both group and individual therapies are scheduled regularly. Structured treatment consists of a primary or intensive phase in which group and individual sessions are given daily for weeks either as an inpatient or outpatient, followed by a less intensive and tapered "aftercare period" of weekly attendance for a period of months to a year or two in an outpatient setting. Regular prescribed "doses" of educational group and individual therapies are given on an individualized basis depending on the severity and associated psychiatric and medical comorbiditics of the patients.3

METHODS

CATOR, Comprehensive Assessment and Treatment Outcome Research, is the largest independent evaluation service for the chemical dependency field in the US. CATOR, as a division of New Standards, Inc., is not government supported, is not part of a governmental agency, and is not owned by a treatment provider. CATOR functions as a clinical auditor in evaluating the efficacy of programs in achieving recovery from alcohol and other drug addictions and documenting correlates of such recovery.

The programs monitored tend to be variations of the Minnesota Model,3 but CATOR has evaluated psychiatrically based programs and aversion therapy programs as well. The common element to all of these programs is that the ideal goal is to assist the patient to remain abstinent from all moodaltering drugs of abuse. To be admitted to these programs, one must have a diagnosable addictive disorder. Thus, problem drinkers and drug misusers should not be admitted to these programs nor included in this patient registry system.

The patient data presented here are from the version of the general registry system revised in 1987 and includes records from 38 inpatient and 19 outpatient programs. The total patient population eligible for follow-up in this version of the registry consists of 8,087 inpatients and 1,663 outpatients. Although an increasing number of programs employ combinations of inpatient and outpatient services for a given case, this report is limited to cases receiving only one of these treatments.

Most research falls into one of three levels of inquiry: survey, evaluation, or experimental. Survey research involves much of epidemiology where one attempts to assess the prevalence of certain phenomena and their relationship to various factors. Evaluation research is a bit more focused in that it involves naturalistic research without modifying the existing practices and procedures, but attempts to provide baseline standards for existing practices in routine clinical settings. Experimental research employs the classical research designs that involve the manipulation of the environment or clinical practices and testing whether one condition is superior to the other. In a sense, as one moves from survey to evaluation to experimental research, one focuses on a narrower and narrower perspective with increasing external control over the target of study.

Effectiveness of a particular treatment can be studied by any of these methods - survey, evaluation, or experimental. These methods are best viewed as separate approaches that yield different emphases of information regarding a form of treatment. Each has advantages and disadvantages when compared to the others, and can be viewed as complementary toward assessing treatment effectiveness.

Evaluation research, as a naturalistic research activity, seeks to explore relationships without experimental manipulation. This frequently results in a misunderstanding of the nature of the work by those schooled in traditional experimental procedures. In some cases, quasi-experimental designs must be employed,4,5 and in others, one must control for the lack of randomized assignment of subjects by using statistical approaches to parcel out differences if different treatments or procedures are being compared. Naturalistic research also typically requires much larger sample sizes to compensat* for the rigorous external control possible in experimental work. However, experimental designs frequently suffer from sample bias because of natural selection of treatment types by subjects who will not submit to randomization. Moreover, the generalizability to general practice is limited because of the narrow focus in a particular time required of the experimental design. The differences between evaluation and experimental research reflect differences in approach and purpose rather than in scientific rigor.

One of the perennial questions in evaluation research is the validity of self-report data. This has engendered much controversy, but the mounting evidence is that selfreport data can be reliable if the proper care is taken in formulating the data-gathering instruments and strategics.6-10 Our own work has shown reasonable agreement between patients' and significant others' reports. A recent analysis of 625 cases revealed that when patients reported abstinence, the family member contacted agreed in 88% of the cases.7 This is a reasonable level of agreement for estimating overall outcome.

Table

TABLE 1Demographics

TABLE 1

Demographics

Another controversial issue is how to deal with cases that are lost to follow-up during the posttreatment period. Many researchers have simply declared these to be relapsed cases. However, our work suggests that this may be an overly pessimistic assumption for the type of middle-class individual served by these programs. An alternative is to develop a range estimate based on a confidence interval concept to provide a rate-ofrecovery range in which the entire population would be expected to fall. The extremes of the projected recovery range are defined at the top by the observed outcomes of contacted cases and at the bottom by the estimate based on the assumption of relapse for all noncontacted cases. A narrower projected recovery range can be defined by estimating the upper bound of the outcome of the range by using prognostic indicators of the noncontacted cases or partial data cases to estimate the outcome of those cases who were not contacted at all. The lower bound of this narrowed recovery range can be defined by assuming that the noncontacted cases have a sobriety rate lower than the projection.

For example, if one wishes to estimate the outcome of all cases at six months after treatment from a group of patients followed for a longer interval, the following procedure might be employed. The cases at six months, but lost at a later interval, have been found in our previous work to have a recovery rate lower than those who continue to be followed. The recovery rate of these single contact cases might be used to estimate the outcome of those not contacted at all, and would yield a more conservative estimate of the upper bound of the projected recovery range than the overall observed recovery rate at six months. One could then assume that the recovery rate of the noncontacted cases might be only half that of these single contact cases. This could be used as a lower bound for the projected recovery range.

A more sophisticated strategy would be to develop a prognostic index based on contacted cases. Such an index would take into account differential probabilities of recovery based on client characteristics. Noncontacted cases could then be assigned a recovery probability to estimate the overall projected recovery rate for all noncontacted cases. This estimate would be the upper bound of the narrowed projected recovery range. One might reduce the projected recovery of noncontacted cases by a specified amount to allow for any unrecognized factors that might result in lower recovery rates for these cases. This adjusted estimate would then be the lower bound of the narrowed projected recovery range.

The entire population of 8,087 inpaticnts and 1,663 outpatients will be the focus of the general discussion of the population descriptions and outcome results, but smaller subsets will be used to address the issues of the effects of the continuum of care and impact of treatment. These latter subsets are dictated by focusing on contacted cases and also selecting appropriate cases, such as limiting the vocational functioning results to those who are actually working both before and after treatment. This report is not a comparison of inpatient and outpatient treatment, but both programs are presented to illustrate that different people may require different treatments, and to also illustrate how some factors appear to be more generic in nature.

RESULTS

Demographically, the inpatient and outpatient populations are relatively similar except for age and work status. As is seen in TabJe 1, the outpatients are much more likely to be in their 20s with fewer cases over 50 or under 20. Also more of the outpatients are employed full time. The inpatients include more minorities and a slightly lower educational attainment.

In contrast, vast differences are noted on the clinical characteristics of the two groups (Table 2). Although a comparable number of patients in both treatment groups are alcoholics, almost three times as many of the inpatients are dependent on prescription drugs and opiates. Cocaine dependence is much higher in the inpatient group; marijuana and stimulant dependence are also higher. While almost half the inpatients are dependent on illicit drugs, only slightly over onethird of the outpatients are dependent on these drugs. Table 2 indicates that inpatients are dependent on more drugs and are much more likely to have ingested such drugs just prior to admission. Therefore, when considering the outcome findings, one must remember that these are not comparable populations.

As was mentioned earlier, the assumption that all patients who were not contacted have relapsed may be unwarranted and overly pessimistic. We have made an effort to refine the estimation of outcome for all cases based on the information available and specified assumptions about the remainder of the sample. In this case. Table 3 presents six-month and one-year post-treatment estimates for both inpatients and outpatients. The highest estimate of the projected recovery range is based only on the outcomes of contacted cases, and the lowest estimate is based on the assumption that all noncontacted cases have relapsed. For the sixmonth period, the upper bound of the narrowed projected recovery range is estimated from those cases contacted at six months but not at 12 months after discharge. For the one-year estimate, cases known to have relapsed at either interval and missed at another were counted as relapsed. Other cases with no or partial contact were estimated from observed expectancies. The lower bound of the narrowed projection range was based on the assumption that all cases without complete data had a recovery probability of 30%.

Table

TABLE 2Clinical Characteristics

TABLE 2

Clinical Characteristics

Table

TABLE 3Projections of Sobriety Rates

TABLE 3

Projections of Sobriety Rates

Table

TABLE 41-year Abstinence by Continuum of Care and Self-help Support

TABLE 4

1-year Abstinence by Continuum of Care and Self-help Support

Table

TABLE 5Stress and Abstinence During the First Year for Treatment Completers

TABLE 5

Stress and Abstinence During the First Year for Treatment Completers

The narrowed projected recovery range for both inpatients and outpatients at six months falls within an 8% band, that is, the upper and lower estimates are within 87r of each other. The oneyear estimate reveals a somewhat broader band due in part to the larger number of cases with missing data. The case made here is not the argument of which estimate is most defensible, but that outcome estimates must be based on explicitly defined assumptions. For some populations, the assumption of relapse for all noncontacted cases may be justified. For this population, we believe that a higher estimate is warranted.

We have previously published findings that individuals who attend Alcoholics Anonymous (AA) after treatment are more likely to be sober than nonattenders.11 This finding, and the observation of those patients who remain active in continuing care after the acute care portion of treatment, has been extended to a two-year follow-up of inpatients and outpatients7,12 The current findings presented in Table 4 support these earlier reports. Both inpatients and outpatients who attend either AA or the aftercare provided by the treatment program are more likely to remain abstinent than nonattenders.

More detailed analyses of the interplay between AA and aftercare reveals that each has an additive effect in continuing recovery. Approximately 7(Y!/< of those who attended AA regularly but did not go to aftercare remained sober, and conversely, a comparable proportion of patients who attended at least four months of aftercare were sober even if they did not attend AA. However, up to 90% of those who attended both AA on a weekly basis and went to afte-rcare for the entire year maintained their abstinence. Clearly addictions need to be addressed as chronic, not acute, conditions.

Table 5 presents evidence of the association between a variety of Stressors and relapse. Consistently, the relapsed patients at the end of the first year reported the most stress during both the first and second six months after treatment. At one year, patients identify such Stressors as contributing to relapse risk and making recovery difficult. One might hypothesize that the greater the level of stress, the greater the probability of relapse. Aftercare services may be more appropriate to address some of these Stressors, while AA might be as good or better at addressing issues of craving or other Stressors. This could help account for the apparent additive effects of AA and aftercare.

Cost-offset issues have been an interest of CATOR for the past decade. The evidence clearly indicates that treatment has the ability to offset much of its costs.13 The data on medical care utilization from Table 6 are compatible with earlier work.14,15 Interestingly, the inpatient utilization rates are consistently higher than the outpatients' both before and after treatment. This again addresses the differences between the two treatment groups. Nevertheless, both show significant decreases in posttreatment medical care utilization for expensive hospital services. Clinic visits show little change before and after treatment, suggesting no shift of cost from hospital to ambulatory services.

Even more striking changes are noted in the improvements in vocational functioning as indicated by a decrease in work problems, absenteeism, and working while under the influence. Table 7 shows two- to fivefold decreases in job problems for both the inpatients and outpatients employed before and after treatment. Dramatic change is seen between absenteeism in the months prior to admission and at each of the two follow-ups.

Moving traffic violations and other arrests show dramatic declines as well (Table 8). In addition, motor vehicle accidents are significantly reduced in the year after treatment. These events also involve cost issues to individuals and society. Accidents increase insurance rates, and arrests result in law enforcement and judicial expenditures.

Table

TABLE 6Medical Care Utilization 1 Year Before and After Treatment

TABLE 6

Medical Care Utilization 1 Year Before and After Treatment

Table

TABLE 7Vocational Issues for Employed Patients 1 Year Before and After Treatment

TABLE 7

Vocational Issues for Employed Patients 1 Year Before and After Treatment

Table

TABLE 8Other Comparisons for 1 Year Before and Alter Treatment

TABLE 8

Other Comparisons for 1 Year Before and Alter Treatment

SUMMARY

Evaluation studies can be used to estimate general outcome of addictions treatment. Moreover, evaluation studies demonstrate the effectiveness of treatment for alcohol and drug disorders in large populations from multiple sites. The generalizability from such studies with large sample sizes is high. These studies suggest clinical innovations to improve treatment efficacy, and demonstrate cost-offsets over time in changing clinical practice. Such naturalistic studies can help focus experimental research on those areas with the highest potentials. However, experimental studies will continue to have the potential for sample bias and limited generalizability. Also, experimental studies often do not reflect actual clinical practice and usually cannot extend over time.

REFERENCES

1. Ch.vppel J. Long-term recovery from alcoholism. Psycliiatr Clin North Am. In press.

2. Schuckit MA. The disease alcoholism. Postgrad Med. 1978; 64:78-84.

3. Laundergmi JC. Easy Does It: Alcoholism Treatment Outcomes, Hazelden and the Minnesota Model. Center City, Minn: Hazelden Foundation; 1982.

4. Campbell DT, Stanley JC. Experimental and Quasi-Expeimental Designs for Research. Chicago, 111: Rand McNally College Publishing Company; 1986.

5. Cook TD, Campbell DT. Quasi-Experimentation: Design iy Analysis Issues for Field Settings. Boston, MUSS: Houghton Mifflin Company; 1974.

6. Babor TH Stephens RS, Marlatf GA. Verbal report methods in clinical research on alcoholism: response bias and its minimization. J Stud Alcohol. 1986; 48:410-424.

7. Hoffmann NG, Harrison PA. CATOR Report: Treatment Ontcnmc - Adult Inpat ien tu Tice Years Lnter. Sl. Paul, Minn: Ramsey Clinic; 1488.

8. O'Farrell TJ, Cutter USG, Bayog RD, Dentch G, Fortgang J. Correspondence between one-year retrospective reports of pretrcatmenl drinking by alcoholics and their wives, Bt'lntriawi Assessment. 1()84; 0:263-274.

9. Sobell LC, Sobcll MB, Outpatient alcoholics give valid self- reports. J Nerv Ment Dis. 1975; 161:32-42.

10. Verinis [S. Agreement between alcoholics and relatives when reporting followup status. Int J Addict. I4S3; 18:891-844.

11. Hoffmann NG, Harrison PA, Belille CA. Alcoholics Anonymous after treatment: attendance and abstinence, lut I AtitHct. 1984; 18:311-318.

12. Harrison PA, Hoffmann NG. CATOR Report: Adult Outpatient Treatment - Pcrsjiecthvoii Admission ami Outcome. St. Paul, Minn: Ramsoy Clinic; 1988.

13. Holder HD. Alcoholism treatment and potential health cnre cost savings. Medical Cure. 1987; 25:52-71.

14. Hoffmann NG, Harrison PA, Belille CA Multidimensional impact of treatment for substance abuse. Adv Alcohol Subst Abuse. 1984; 3:83-94.

15. Rode (DeHarO SS, Hoffmann NG, FuIkerson JA. Medical care utilization before and after alcohol treatment for the elderly. Ttte Sautiiwestent: The journal of Aging in the Southwest. 1990:6:140-148.

TABLE 1

Demographics

TABLE 2

Clinical Characteristics

TABLE 3

Projections of Sobriety Rates

TABLE 4

1-year Abstinence by Continuum of Care and Self-help Support

TABLE 5

Stress and Abstinence During the First Year for Treatment Completers

TABLE 6

Medical Care Utilization 1 Year Before and After Treatment

TABLE 7

Vocational Issues for Employed Patients 1 Year Before and After Treatment

TABLE 8

Other Comparisons for 1 Year Before and Alter Treatment

10.3928/0048-5713-19920801-05

Sign up to receive

Journal E-contents