Improving the filter between primary and secondary care for mental disorders

Almost 25 years ago, Goldberg and Huxley described patterns of recognition and treatment for mental disorders using the concept of filters separating different levels of care (1). We might view David Goldberg's introductory paper in this Forum as a guide to improving the functioning of the filter or interface between mental health specialty services and primary medical care. This guide contains several principles for the optimal functioning of that filter.

First, the filter or border zone separating primary and secondary care should be broad - a gradual transition rather than a sharp boundary. In fact, the optimal border region might cover as much territory as the two nations it separates, the regions devoted exclusively to primary or secondary care. David Goldberg describes several models for primary care and specialty services to share responsibility for both common and more severe mental disorders.

Second, the filter should be sensitive to clinical need. The primary factor determining level of care should be severity of symptoms and degree of impairment. Because severity of illness varies considerably over time, regular monitoring of clinical condition is necessary for appropriate triage.

Third, the filter should be freely permeable in both directions. Given that clinical need varies over time, level of care should vary according to need - with relatively low barriers to transitions upward or downward.

Fourth, the filter should be relatively insensitive to non-clinical factors. Unfortunately, access to higher levels may be overly influenced by non-clinical factors such as insurance coverage, ability to pay, race, or social class. A well-functioning filter would ignore these factors and might, in some cases, actively work to circumvent barriers to appropriate care.

Fifth, the filter should be sensitive to local resources and constraints. Optimal criteria for specialty consultation or referral will vary widely depending on the availability of specialty services. In other words, the marginal benefit of specialty involvement may often be positive, but limited specialty resources must be reserved for situations in which they produce the greatest good.

Sixth, the filter should be sensitive to the particular strengths of primary and secondary care. In general, primary care has the advantages of easier access, long-term continuity of care, and coordination across multiple health conditions. Advantages of specialty care include greater expertise, focus, and (in some cases) efficiency due to a narrower scope of practice. The relative importance of these factors will vary between patients. For example, the advantages of primary care management might be greatest for a patient with comorbid chronic medical illness and a longstanding relationship with the primary care team.

Others have used the term 'stepped care' to refer to organized care adjusted according to severity of illness and response to initial treatment (2,3). David Goldberg's paper describes several promising stepped-care models for a range of health care environments.