Improving quality step by step: the conclusion to a two-part series about improving the quality of behavioral health services.

Editor's note: In part one of this series in the August issue, Dr. Dougherty outlined the problems facing behavioral healthcare as it attempts to improve service quality. In the conclusion to this series, Dr. Dougherty outlines an approach that holds significant promise--learning collaboratives.

Learning collaboratives have been used extensively to promote quality improvement in healthcare, particularly over the past five years. There are several types of collaboratives, but generally they involve the use of: (1) cross-discipline and interorganizational teams; (2) work on a specific problem; (3) leadership by evidence; (4) faculty experts; and (5) project management coaches to modify and/or improve specific practices. (1)

Collaboratives have been formed in large multidivisional organizations, among different providers, across purchaser or health plans, and at the community level. They provide a richer experience than traditional quality improvement approaches, in which the focus often is limited to a unit or organizational division. Collaboratives have aimed to improve chronic illness care, the treatment of depression, general community health, asthma care, and a number of other health concerns. (2-4) The Institute for Healthcare Improvement is one of the leading proponents of this approach. (5) Examples of mental health and substance abuse-related collaboratives include the:

* Robert Wood Johnson Foundation Depression and Primary Care Initiative;

* National Initiative for Children's Healthcare Quality ADHD and primary care effort;

* Center for Health Care Strategies' Best Clinical and Administrative Practices Program for Health Plans Service to Children with Serious Emotional Disturbance;

* California Institute for Mental Health's California Learning Collaborative; and

* Massachusetts Department of Mental Health's Readmission Collaborative

One of the earliest collaboratives in mental health was undertaken by the New York City Department of Health and Mental Hygiene, Mount Sinai School of Medicine, and the New York State Office of Mental Health. (6) This collaborative sought to improve access to services (first intake appointments) for children seeking care for mental health conditions.

[FIGURE OMITTED]

Over the past three to four years, the Robert Wood Johnson Foundation initiated a large national effort to improve access to and retention in addiction treatment. This work, led by the Network for the Improvement of Addiction Treatment (NIATx) at the University of Wisconsin, has been extensively supported by the federal Center for Substance Abuse Treatment. In behavioral health, NIATx is without peer in advancing quality improvement practices among providers.

Collaboratives start where your organization is now. No major change is first required. They seek to improve performance through a series of focused short-term interventions and improvement methods nested within a long-term goal. Groups of staff members work together to define specific project aims, measures, and interventions. Efforts are action-oriented. Barriers are identified quickly, and data become available to support larger systemic changes. Shewhart's "rapid cycle" use of the Plan-Do-Study-Act (PDSA) method has shown extensive promise and evidence of improvement in addiction treatment through the work of NIATx, (7-9) and has resulted in documented improvements in other healthcare settings as well.

Quality improvement collaboratives are data-based change processes. With training, improvement tools, and a new language of change, they can encourage the growth of a true culture of quality, in which continuous efforts to improve practice and outcomes are the norm. Through small initial projects and measurement of their results, managers and staff learn the techniques, test the concepts, modify them as needed, and ultimately take the project to scale. They also can use the tools for change in other areas and on other projects.

Fundamental to most quality improvement approaches is sequential use of evidence-based and generalizable PDSA cycles to improve organizational processes. This approach has been tested in a variety of settings during the past five decades. (10-13) PDSA cycles also have advantages when it comes to translating research to practice because the incremental nature of PDSA efforts allows for the systematic implementation of research findings and adaptation to local conditions. (14) The PDSA method's efficacy may be due to imbuing systematic improvement with the scientific method (data and evidence) and sequentially implementing cycles of applied change and learning. (15-18) The PDSA cycle also is fundamentally consistent with behavioral learning and general systems theory. In fact, one of the most important cultural changes is that by using the PDSA method, managers and organizations develop a new way to talk about group learning and systems change.

Implementing Learning Collaboratives

Learning collaboratives can range in size from 3 to as many as 20 organizations or units, with staff teams from each. The collaborative consists of a series of learning sessions spread over a project lifecycle. This might range from three 2-day meetings over 12 to 16 months (similar to the programs run by the Institute for Healthcare Improvement and the Center for Health Care Strategies) to eight to ten 1- or 2-day meetings over the same period (similar to the California Learning Collaborative and the Massachusetts Readmission Collaborative). Between the sessions, the groups implement the action steps they have planned, collect data, implement PDSA cycles, and sustain the quality improvement cycle internally. Support can and should be provided in group or individual calls or meetings between the learning sessions. New Web-based technologies can enable group meetings on the Internet and shared Web sites to maintain and increase collaboration.

Problems to address may be selected in advance or jointly defined with others in and outside of the organization. Leaving the choice of the improvement topic up to participants requires that they take extra time to choose it. It is preferable that all the groups work on projects in a similar area. Having different projects reduces the opportunities for participants to learn from one another. The opportunity to engage with and learn from peers in other settings over an extended period is particularly rewarding. All projects should focus on areas with an evidence base in the literature and readily available data.

Several factors have been shown to be important to collaboratives' success. (19) Projects need to be practical and relevant to the organization. Senior managers must be involved and active supporters. Objectives must be relevant, reasonable, and measurable. Finally, it is critical to explicitly plan for diffusion and spread within the organization.

Transformation: A National Quality Improvement Effort?

Transformation of the mental health system is the national goal outlined by the President's New Freedom Commission on Mental Health and a priority for the Substance Abuse and Mental Health Services Administration (SAMHSA). Subsequently, many in the field have called for significant system reform. For the vast majority of us, however, the only way we will cross the "quality chasm" is through incremental efforts at improvement. In fact, many argue that incremental changes, taken together and building on each other, have the power to truly transform practice.

So how can we maintain and expand the use of collaboratives? Adrian Bishop and I have argued for an "Apollo Program" in mental health transformation. (20) It took eight years for the Apollo managers to move from President Kennedy's vision to the moon. If our country can do this, we should be able to achieve some level of system transformation for mental health services. Yet important questions remain: What will be our measure of success? How will we get there? How will we pay for it?

Clearly, training and leadership are needed. We need a consistent message and voice in the field that speaks to the importance of quality improvement. We may have that already from several corners: first, and perhaps most importantly, from NIATx; from the ongoing efforts of the National Council for Community Behavioral Healthcare, the National Association of State Mental Health Program Directors, and others to improve access to treatment and quality; and perhaps also in the work by the federal Center for Mental Health Services in leading the transformation states.

Learning collaboratives are ways to simultaneously solve business problems, provide training to staff, and build a foundation for longer-term change. The figure shows how stakeholders can learn from each other. Sustaining this, at a state and national level, may require an entity that functions as a "meta-collaborative." The field needs a national organization or, at a minimum, a national set of conferences to coordinate and facilitate the various efforts under way and to collect and disseminate the results. Optimally, a variety of funding sources would support this effort, and it would include national, state, and provider organizations. The effort would begin by establishing a dialogue with the state and local collaboratives, and by looking for opportunities to spread the approach and increase learning. By encouraging multiple seeds of quality to grow and by nurturing them, we will find that, one tree at a time, a "forest" will grow. We all need to lend a hand, contribute resources, and participate in the work.