Congratulations to Salyers and Tsemberis (2007) for tackling the details of a difficult issue and making several positive and timely recommendations. Their suggestions should influence the current process of revising the assertive community treatment (ACT) fidelity scale. This article reminds us that, after nearly 30 years of ACT, many details of the model remain unspecified. Its recovery-orientation is paramount among them.

A recent study of ACT teams in Indiana found that the variance in the use of various coercive measures was enormous: some teams use a variety of forms of coercion routinely, others not at all (Moser 2007). Coercion is a flagrant violation of recovery values that can be specified, measured, and reduced. Other aspects of recovery-orientation are subtler, but they probably also vary widely across teams in the absence of clear standards.

Specifying values and quality in human service interactions is extremely difficult. Program manuals and fidelity scales generally emphasize structures and activities that are easily measurable, e.g., caseload size, number of meetings, and location of meetings, but they cannot address the attitudes of staff and the quality of relationships. Training and supervision are the traditional mechanisms for addressing the quality of interactions in human services, but assessing interactions requires direct observations or videotapes. As training, supervision, and workforce experience in the public mental health system erode under chronic financial duress, quality inevitably suffers (Woltmann et al. in press). Staff in the current U.S. public mental health system are under-trained, overworked, and under-supervised.

Salyers and Tsemberis point out that a third of the DACT fidelity items may interfere with recovery-oriented practice. For instance a client may prefer the privacy and trust of working with only one staff person but as a recipient of ACT must work with a team. Or a client might prefer to work with a psychiatrist of his or her choice, rather than the psychiatrist who leads the ACT team. These are challenging issues that make ACT difficult to reconcile with recovery principles and practices.

Assertive community treatment has been remarkably robust over 30 years. Started as an evagination of the hospital team into the community, ACT has been adapted by adding evidence-based approaches to community care: e.g., supported employment, family psychoeducation, and integrated dual diagnosis treatment. The team-based structure has permitted this flexibility for adaptation. Shifting to more of a recovery orientation can be seen as another helpful transition as the values and science of community-based care evolve.

Another perspective, however, is to question the long-term viability of the fundamental ACT model. Modifying ACT to incorporate recovery is just one of many current problems. ACT is also being challenged to address graduations and transitions, forensic issues, complicated medical problems, treatment of sex offenders, homelessness, people with borderline personality disorder, transition-age youth, different financing regulations, and so on. The proliferation of environments, clinical populations, and concerns raises a basic question, should we continue to modify ACT to address multifarious community mental health needs?

Two antithetical answers portray the chaotic, diverse, and inconsistent state of the U.S. public mental health system. On the one hand, in some state and county systems, basic principles of ACT, such as continuity of responsibility and working in the community to help people learn the skills they need to succeed in roles of their choice, have been incorporated by all community mental health teams as usual care. In these settings, as in European countries with more integrated systems of care (Wright et al. 2004), ACT services offer little or no statistical advantages in producing better outcomes (e.g., Essock et al. 2006) and almost certainly no advantages in quality of interactions. Furthermore, teams in these relatively well-funded regions become more specialized and more able to address the specific needs of consumers. For example, specialized teams serving people with borderline personality disorder, those with first episodes of psychosis, or elderly clients with heavy medical comorbidities are emerging in many places, and these teams only superficially resemble ACT.

On the other hand, many regions of the U.S. are so non-progressive and poorly funded that ACT still represents a service innovation that offers consumers with high needs their only chance at coherent non-institutional services of any type (NAMI 2006). These are states that spend at 10% the level of other states and that are still transferring people with mental illness between homeless shelters, hospitals, and jails and prisons at high rates because they have so few community-based services. In these areas, ACT is the only politically viable alternative to abuse and neglect. For consumers in these states, it is not the time to abandon ACT.

Is ACT compatible with recovery? It had better be. The mental health system needs both science and values, and for the time being, ACT faces the challenge of combining both in the chaos that constitutes the U.S. public mental health system.