In the external consultation approach, external facilitation is required because the practices lack the knowledge, skills, and experience to carry out their own change process. This approach is modeled after the one Dietrich and coworkers45-46 demonstrated to be effective for systematically improving clinical preventive services. It involves 4 components: feedback of comparative performance data to clinicians from a patient survey or chart audit; use of academic detailing47 to encourage trying new ways for a limited time period without requiring long-term commitment up front; provision of a menu of system tool examples for selection and implementation (eg, both electronic and paper-based systematic follow-up tools; brief questionnaires for depression screening; a chart stamp to create a place to enter screening scores and timing of follow-up; evaluations of referral resources; and organized patient education pamphlets); and consultation visits and phone calls to each practice to encourage and facilitate use of the tools. The New England project has been using this approach with 4 of the smaller and more isolated practices that participated in our surveys.
In contrast, the internal change process uses a variation on the techniques popularized by Langley and colleagues,48,50 Berwick,49 and the Institute for Healthcare Improvement Breakthrough Series.51 Since traditional CQI methods have had mixed effects and are widely perceived to require too much personnel and calendar time, the Breakthrough Series emphasizes 2 new ideas that allow a multidisciplinary improvement team to build a new system piece by piece: change concepts—new ideas for ways to do things that have been found to be useful,48 and rapid-cycle tests—using small-scale tests of individual parts of a larger change concept, measuring whether it worked, then modifying the approach on the basis of what was learned.49
The team begins by answering 3 questions:48 What are we trying to accomplish? How will we know that a change is an improvement? and What changes can we make that will result in improvement?
The first question (“What are we trying to accomplish?”) forces clinic leadership to be very clear about the focus and aim for the project. The second question (“How will we know that a change is an improvement?”) requires the team to collect a minimal amount of data to permit them to understand the problem and measure the effects of any changes made. Finally, the third question (“What changes can we make that will result in improvement?”) requires them to suggest change concepts that can form the basis for specific improvements to be tested in a series of rapid-cycle tests. Some examples of change concept-based improvements applicable to systematizing the care of patients with depression are: a clinic’s registered nurse who calls back patients with depression to provide information, support, and coordination of care after the visit with a primary care clinician; a system to initiate and schedule these call-backs with little or no clinician effort; a system to provide these patients with information about depression and about the various resources available to them to facilitate their self-care; predefined systematic care options, ideally including ways to help depressed patients who do not want or need medications or active counseling; and systematic facilitation of referrals and communication between primary care providers and off-site mental health resources.
The internal change process approach is being used by the DIAMOND (Depression Is A MANageable Disorder) Project in Minnesota. This setting differs from the practices in New England because it is a large (25-clinic) medical group located in a large metropolitan area with the experience and resources to lead its own change process. Despite the apparent differences, focus groups of physicians and nurses in the Minnesota practices perceived similar problems with access, communication, and collaboration with mental health providers and had similar overuse of medications, underuse of other treatments, and lack of systematic follow-up.52 Thus, there was a similar need for effective systems to support desirable care.
Deciding which of the 2 approaches will be most useful to clinicians in other settings depends on the specific situation and experiences of the site. Real improvement in clinical systems also requires satisfaction with the conditions identified by Shortell and colleagues53 in their recent critique of the current state of the quality improvement field: the areas slated for change need to be of real importance to the organization and should be addressed with clearly formulated interventions; the organization must be ready for change with capable leadership, trusting relationships with its physicians, and adequate information systems; and the external environment must not be hostile in its regulations, payment approach, and competitive factors.
Conclusions
There is considerable room for improvement in the primary care of patients with depression, and we believe that most of the problems associated with current care patterns would benefit from a more systematic approach.