Many studies establish the effectiveness of the Individual Placement and Support (IPS) model of supported employment and of similar models that follow the same evidence-based principles. Across 12 randomized controlled studies, nearly two-thirds of individuals with psychiatric disabilities who are assigned to supported employment attain competitive employment, compared to less than a third of those assigned to other vocational programs (Bond, 2004a). In just over 15 years, evidence-based supported employment has become the standard of care within psychiatric rehabilitation (New Freedom Commission, 2003; Bond, 2004b; Lehman et al., 2004).

It has also become clear, however, that outcome differences are considerable across supported employment programs, suggesting that more work is needed on standardization and quality improvement. While most IPS studies report competitive employment rates exceeding 60%, there are isolated examples of rates below 30% (Lehman et al., 2002; McGurk, Mueser, & Pascaris, in press). Some of these outliers are probably attributable to low fidelity. The Lehman study, for example, enrolled many clients who had little or no interest in employment, contrary to the principle of consumer choice.

As Becker and colleagues point out, their study is one of several that establish that fidelity to the evidence-based model, assessed by fidelity scales, explains a significant proportion of the regional variation in employment outcomes among supported employment programs (Becker, Xie, McHugo, Halliday, & Martinez, 2006). The three studies with the most convincing findings of a fidelity–employment outcome relationship explained 22% (McGrew & Griss, 2005), 26% (Becker et al., 2006), and 58% (Becker, Smith, Tanzman, Drake, & Tremblay, 2001), respectively, of the total variance in employment outcomes across multiple program sites. These findings suggest that training, supervision, and implementation monitoring should target the principles of fidelity that have emerged as critical components of evidence-based supported employment. Unpublished data from the National Evidence Based Practices Project (Mueser, Torrey Lynde, Singer, and Drake, 2003) show that high fidelity to evidence-based supported employment is attainable in virtually all programs, provided they have sufficient resources, supports, and training.

In addition to variance across programs, considerable variance in employment outcomes is often observed within supported employment programs. For example, within many IPS programs that we have studied, it is not uncommon to find that some employment specialists help less than 25% of their clients to obtain competitive employment, while others in the same program have success with greater than 75% of their clients. These differences appear to be related to clinical skills; some employment specialists simply are more proficient or competent at the relevant tasks. Fidelity scales typically address program structure and components rather than practitioner competence. Practitioners undoubtedly need to be working within the best structure and using evidence-based models, but trainers, supervisors, and employment specialists believe that practitioner skills are equally important.

Several different clinical skills have been described in the evidence-based model of supported employment (Becker & Bond, 2002; Becker & Drake, 2003). In general, these skills fit within a matrix of specific interactions (transactions with clients, with other staff, and with employers) and of stages of supported employment (engagement, assessment, finding a job that matches talents and interests, insuring success by addressing skills and supports, leaving a job appropriately, and finding another job). Employment specialists need skills in each of these areas.

Further work is clearly needed to clarify the critical skills of employment specialists, to develop ways to assess these skills, and to improve the skills through training and supervision. Some evidence suggests that supervision strategies, such as systematically monitoring employment outcomes and providing field supervision, are critical interventions for improving skills (Corrigan, Lickey, Campion & Rashid, 2000; Harkness & Hensley, 1991).

In addition to strategies undertaken at the individual program level, the paper by Becker and colleagues also suggests a vital role for an external agent in enhancing employment outcomes. In the U.S., this external agent is usually a state agency or a technical assistance center. The state can improve outcomes through a variety of mechanisms, including outcome and fidelity monitoring (Rapp et al., 2005). When financial incentives are attached to achieving benchmarks for fidelity and outcome, programs typically perform at a higher level.

Standardization of supported employment and other models of evidence-based mental health care can enhance implementation and outcomes. Fidelity assessment has clearly been an essential component of the process. A renewed focus on practitioner skills might enhance the field further.