Can addiction-related self-help/mutual aid groups lower demand for professional substance abuse treatment?

Professional substance abuse treatment in the United States grew extensively through the 1970s and 1980s, becoming literally a billion dollar annual enterprise. In recent years, however, the professional treatment network has contracted due to the arrival of managed health care and to recent public sector disinvestment. Clinicians and health policy planners now face the difficult challenge of attempting to care for addicted individuals within increasingly tight fiscal constraints. But there is a potential bright spot in the current gloomy addiction care picture–the possibility that self-help/mutual aid organizations can help substance abusers recover, while at the same time lowering demand for scarce formal health care resources.

The self-help/mutual aid group movement covers a vast range of health, social, and political concerns, but the largest and probably best-established component of the movement focuses on addiction. Alcoholics Anonymous (AA), Moderation Management, Narcotics Anonymous (NA), and Women for Sobriety (WFS) are among the many self-help organizations that help addicted individuals live productive, substance-free lives. These organizations differ from professional services in many ways–peer control, an emphasis on personal experience rather than credentials and training, informality, mutual rather than one-way helping–but in this context their most important defining quality is that they are essentially free of charge. Many of them “pass the hat” for small, voluntary contributions, but this is a trivial cost compared with those of professional addiction services.

The Impact of Self-Help Groups on the Demand for Treatment
In a national, representative survey of US adults, Kessler and colleagues (Social Policy, Spring 1997) found that individuals who attend self-help groups were more likely than non-attenders to have been treated by a professional for substance abuse problems. Perhaps surprisingly, this positive correlation neither detracts from nor supports the hypothesis that self-help groups lower demand for professional addiction treatment. As this reasoning might not be intuitively clear, consider an example that is a bit more down-to-earth.

If we surveyed all Americans about the restaurants they had patronized in the past year, we would find that the frequency of eating at Wendy’s, Burger King, and McDonald’s were all positively correlated (i.e., people who ate at one of the chains some of the time were more likely to also eat at the other two chains). Yet we know for a fact that these chains are in competition with each other, and when one of them opens new stores the others lose business. This apparent contradiction between our cross-sectional correlation and what we know to be true comes about because underlying common factors influence eating at all three chains: a tendency to eat out rather than cook and a preference (by taste or economic necessity) for fast food restaurants over other types of restaurants. Hence, the three restaurant chains can share customers yet be in competition at the same time.

Similarly, the positive cross-sectional correlation between self-help group involvement and utilization of professional addiction treatment services is likely attributable to common factors, namely having a substance abuse problem and being comfortable seeking help from others outside of one’s immediate family and social circle. These common factors produce the positive correlation noted in Kessler’s survey but, as with the example of the fast-food chains, they cannot tell us one way or the other whether self-help group participation lowers demand for addiction-related health care. The only way to answer that question is to conduct studies that examine patterns of health care utilization and self-help groups over time.

For example, a study of 227 alcohol-abusing workers randomly assigned participants either to an inpatient treatment program followed by AA meetings or to community-based AA meetings only. The research team monitored substance abuse and job-related outcomes (e.g., absenteeism), as well as alcohol-related health care costs over the two-year period. At the conclusion of the study, the average participant assigned to AA only had consumed $8,840 in alcohol-related health care resources versus $10,040 for the inpatient treatment participants. This amounts to a savings of $1,200/person or 10 percent. However, although the two groups fared equally well on job-related outcomes, the AA group had more relapses over the course of the study. Hence, potential health care cost savings from self-help have to be considered in light of whether outcomes will be compromised, producing additional suffering and health harms.

Following up on this research, a colleague and I studied 201 alcohol-abusing individuals who sought help either from AA or a professional outpatient treatment service provider. Over the ensuing three years, both groups of individuals improved substantially, reducing, for example, their daily alcohol consumption by 70 percent on average. Unlike in the study of inpatient treatment, AA attendees’ outcomes were just as positive as those of individuals who had received professional services. Importantly, those in AA consumed 45 percent less alcohol-related health care resources over the three years of the study (about $1,800 less/person). This study indicates that self-help group involvement can lower demand for professional treatment without compromising health outcomes. My colleagues and I are now attempting to replicate this finding in a different sample of patients, and initial results confirm the findings of the first study.

These studies, considered together with the research literature showing that self-help groups for psychiatric problems seem to lower demand for professional mental health services, coalesce in support of the proposition that self-help groups can lower the demand for addiction-related professional services. This finding becomes more believable and more understandable when we recognize that both self-help organizations and professional treatment services are networks. Most helpers are more comfortable referring an addicted individual to someone in their own network than to an outside network. For example, when people go to their first AA meeting, they are likely to hear about other AA meetings and potential AA sponsors, whereas when they go to a professional, they are likely to hear about other professionals and the services that they can provide.

Other research projects currently underway will provide more definitive evidence on whether and to what extent addiction-related self-help groups can lower demand for formal health care while promoting positive outcomes. In the meantime, it is worth speculating on what it would mean for substance abuse treatment policy if future studies replicate the findings reviewed above.

Rethinking the Relation between Professional Services and Self-Help Groups
Addiction treatment professionals and self-help groups currently collaborate and intersect in many ways. Many “recovering” individuals, for example, work as counselors in treatment programs and many professionals refer their patients to self-help groups. At the same time, many if not most professionals think of this relationship in professionally-centered terms. Professionals often conceptualize self-help groups as an “adjunct” or “aftercare” intervention that can play a supportive role to the “real” intervention provided by professionals.

This perspective is partially attributable to professional socialization and training that, across fields, inculcate a bias that professionals are better at a wide range of tasks than are nonprofessionals. The now collapsing system of fee-for-service health care, however, also played a role by giving professionals a financial disincentive to “lose” patients to self-help groups. Under the fee-for-service system, a patient who elected to attend Women for Sobriety instead of a four-week inpatient treatment program represented a loss of several thousand dollars to the professional program. Hence, even if such an intensive intervention were unnecessary, the incentives were clearly to discourage self-help group involvement except as an adjunct or aftercare component to the costly professional program.

In contrast, under most managed care arrangements, a professional or health care plan is financially better off when an addicted individual who does not really need hospitalization or another professional treatment seeks out a free self-help group instead. From a cost-conscious point of view, self-help groups should be the first option evaluated when an addicted individual makes initial contact with professional services (e.g., in a primary care appointment or a clinical assessment at a substance abuse agency or employee assistance program).

No one should be forced to attend a self-help group, but the risk to an addicted person of trying out a few meetings is sufficiently slight (and the payoffs in terms of outcome and health care cost savings sufficiently large) that a recommendation to visit a few groups should be standard practice among health care professionals. In this framework, self-help groups would move from being the “junior partner” of professionals to being “the first line of defense.” In contrast, the role of addiction treatment professionals would be redefined from treating as many individuals as possible (including individuals who could recover without treatment) to treating that subset of individuals who do not benefit from self-help groups. Such patients might include addicted individuals who also have severe comorbid psychiatric and medical problems that require close monitoring, as well as patients who have tried self-help groups and found them unhelpful.

Were this arrangement consistently adopted, scarce professional services would be better targeted to individuals with more severe problems. Further, health care costs would likely drop because many individuals who initially sought professional treatment would ultimately rely more on self-help groups than on professional services. From the point of view of wise and socially just resource allocation, this seems an ideal arrangement. Nevertheless, some professional organizations may resist such a change.

Without besmirching any particular helping professional (I am one, after all), it is safe to say that professions pursue their own guild interests, even when they are not in line with the public interest. Although the re-conceptualization of the role of addiction professionals that I have outlined would serve the public interest, it might put professionals in a difficult position. For example, the more avaricious managed care companies may overreach from the evidence and refer all addicted patients to self-help groups, regardless of need or preference, and then stop reimbursing professional treatment services. Faced with this legitimate fear, professionals may resist becoming secondary to self-help groups because they understandably do not want to be put out of business.

Bringing professionals on board will require both education and advocacy. The education component is necessary for those professionals who still conceptualize their work in terms appropriate for the old fee-for-service model (i.e., that they are competing to provide more services to more patients). These professionals would benefit from learning the workings and underlying logic of newer capitated managed care models, in which everyone “wins” when an addicted patient who recovers with the aid of a self-help group relies less on professional services. In terms of advocacy, either through the pending federal HMO legislative process or as part of individual contracts negotiated through employers, addiction professionals and payers will need to work together to ensure that the benefits and cost savings of self-help groups are not used as a rationale to eliminate professional services. If addiction professionals have assurance that a greater emphasis on self-help groups in the US system of formal and informal care for addiction will not result in the elimination of professionals, they will be more likely to implement the model proposed here, and that should result in better allocation of our addiction-related health care resources.

By Keith Humphreys

Leave a Reply