The results of this study suggest that the two outcomes, use of standardized assessment tools and use of practice guidelines, are interdependent. In the first model (Time 1), managed care involvement, and provision of medical care services and use of medications had the strongest association with increased use of standardized assessment tools. Use of assessment tools and offering counseling and ancillary services was then associated with greater use of practice guidelines. In the second model (Time 2), again managed care involvement was associated with increased use of assessment tools, as was being a stand-alone drug treatment agency. The relationship between provision of counseling and ancillary services had a stronger association with assessment tools in this model, but the relationship between the use of assessment tools and a greater use of practice guidelines remained strong in the second model. Thus, programs that offer a larger range of services (i.e., medical, ancillary or wrap around, medication and counseling) and have strong relationships with or receive funding from managed care entities, appear to be more likely to adopt standardized assessments and practice guidelines or algorithms.
In terms of clinical staff and their impact on adoption, programs that were free-standing or non-hospital based were also more likely to have a greater number of staff in recovery and providers with fewer years of education. Managed care involvement was also associated with having more staff with higher education.
It is also interesting to note that the majority of agencies created their own assessments/intakes, or possibly used a mix of standardized assessments, suggesting a legacy of program tools that may not be updated frequently. This range of eclectic assessment practices may also reflect dissatisfaction with, or a lack of knowledge about validated assessment tools. Thus the models from this study provide data for state, county and community policy makers to consider when allocating resources, planning trainings, and implementing targeted systemic interventions.
Adoption of Evidence-Based Practices
Overall, research thus far suggests limited use of practice guidelines in community-based substance abuse treatment. To better understand why this implementation or such clinical decision guides is slow, organizational, environmental and staff factors must be examined. To date, services research suggests that the decision to adopt new practices is often influenced by the demands of the funding entities and institutional structures [14, 15]. Further, providers that are most likely to adopt new practices tend to have more formal education, come from a higher social status, and have a more favorable outlook regarding change . These early adopters are also more likely to take the initiative to seek out information and have a greater number of national professional contacts . Clinicians indicate that they are more open to new strategies when those practices clearly improve services and outcomes, are consistent with their philosophy of treatment, and are supported with training that includes observation, practice, supervision, and feedback prior to full adoption [31–33]. Organizational structures and practices (e.g., funding mechanisms, agency mission, and data management systems) also impact the use of innovations within substance abuse treatment programs. Thus, both individual treatment providers and agency level or organizational characteristics influence the implementation of new practices including practice guidelines and the use of assessment and screening tools [15, 18, 19]. The results of this study correspond with this implementation science and classical diffusion theory literature. Funding mechanisms (involvement with managed care), treatment philosophy (counseling and use of ancillary services), and staff characteristics were associated with greater implementation of practice guidelines and assessment tools. The ASAM patient placement guidelines and the ASI assessment tool were widely used and were considered best practices by the participating outpatient treatment programs.
Interestingly, although having greater involvement with managed care was associated with specific staff characteristics as well as with use of assessment tools and practice guidelines, in this project the level of counselor education in the programs and the number of staff in recovery was not directly associated with greater use of practice guidelines. Although other research has found that education is associated with more positive opinions about evidence-based practices  and greater acceptance of innovation , previous investigations were based on attitudes surveys with treatment center staff. The current data came from a survey of treatment center directors and not a direct assessment of counselors or staff. Methodological differences may explain the difference in findings. It may also be that use of assessment tools and practices guidelines present a greater level of controversy and challenge for providers. According to Miller, Zweben and Johnson , retraining providers with established habits and experiences is both challenging and time consuming. The use of standardized assessment tools and practice guidelines requires considerable training and supervision. Autonomy also has been associated with increased satisfaction and commitment to the organization , thus providers may value independence in decision-making and they may also not have access to the necessary training and supervision related to assessment and the use of guidelines. Finally, staff may have felt constrained by guidelines and may not have perceived any clinical or organizational benefit with their implementation in their treatment programs. Thus, the findings from this study may differ somewhat from previous studies that suggest counselor characteristics predict adoption of new practices due to methods (i.e., response of agency directors), the amount of training and supervision required to use standardized tools and guidelines, as well the constraint or lack of autonomy that counselors may resist when using use tools/guidelines.
Findings from the current study also suggest an important relationship between agency size, counseling and ancillary services, and use of practice guidelines and standardized assessment tools. Larger agencies offered more counseling and ancillary services which contributed to the use of practice guidelines and assessment/screening tools. Integration of primary care and mental health with substance abuse treatment offers increased access to a range of providers, medications, and independence from government funding which provides leverage for change and flexible, progressive decision-making for the organization [36, 37]. Large agencies that offer a wider range of services have the resources to purchase assessment instruments, provide training, conduct organizational change processes, and flex their resources in a fashion that is necessary for the implementation of standardized practice guidelines. Consistent with the literature, participation in ancillary service delivery appears to connect programs with other community providers, exposing managers and counselors to diagnostic tools and procedures outside their immediate organization . Applying this literature base to specific clinical practices, including use of standardized assessment tools and practice guidelines, provides greater understanding of these factors and their dynamic and interdependent impact on diffusion of innovations
This study had a number of limitations. The sample included six New England states, which may have limited generalizability to the rest of the United States. New England is a region with high managed care penetration and a high degree of public sector treatment agencies which may not be characteristic of the rest of the country. In addition, slight wording changes in the surveys between the two data collection points created some restrictions regarding the type of longitudinal analysis that could be used (i.e., models are not dependent time 1 to time 2). Had the models been able to be dependent, statistical tests would likely have been more sensitive. However, because not all clinics were represented in each sample, this was not possible and is a limitation of the study. Missing data imputation is not used with structural equation modeling, thus a listwise deletion technique results in parsimonious models based on cases with complete data available for analysis. Finally, although data were collected in 2001, dissemination of these findings is important because they are consistent with the literature confirming the difficulties changing practice patterns and disseminating research findings. The study results may also be used for comparison with future assessment that evaluates the use of practice guidelines.