|Integrated primary care is growing in attention and presence. Recently, the Institute for Healthcare Improvement predicted that “in the not-too-distant future, “non-integrated” primary care practices … will become rare or even extinct”. Even more recently, the American Psychological Association devoted an entire special issue to integrating psychology into primary care. Working in the field of collaborative care is like riding a train: it has taken time to build momentum but once it reaches a top speed the train can become a juggernaut. Although the view ahead of us is exhilarating and breathtaking, it is helpful to notice what kind of tracks we are using for this train.|| |
During my undergraduate training in psychology I enrolled in a class that required me to purchase a book entitled “What’s Behind the Research?” From this book I learned about the underlying (and sometimes hidden) philosophical assumptions of the behavioral sciences. For this blog post, I will borrow a page or two from that book to take a brief critical look at two models of integrated care. Understanding what railroad tracks we are using to move forward in this exciting field is important. This work involves helping human beings who are suffering. The way we view human suffering influences what kind of assessment and treatment we select and how we do collaborative care.
Care Management Model
One of the more prolific examples of this approach to integrated care is the IMPACT model, developed by psychiatrists from the University of Washington. IMPACT (Improving Mood-Promoting Access to Collaborative Treatment) is an evidence-based collaborative care model that has an incredible amount of supporting literature. IMPACT is a stepped-care approach that includes a care manager (usually a nurse) and a supervising psychiatrist who collaborate with a primary care doctor to treat one or two conditions (usually depression). Treatment typically includes psychoeducation (brochure, video), antidepressants, brief counseling and/or coaching, treatment monitoring, relapse prevention, and referral to specialty mental health as needed. This is class care management for chronic conditions. To learn more, click here.
|Are depression measures and treatments adequate enough to address the suffering of primary care patients? ||The track record of the IMPACT model is extensive. According to the data, this model leads to a 50% reduction in depression symptoms at 12 months, works in various settings (e.g., inner-city, Veteran’s Affairs, HMO, fee-for-service) and with a variety of populations (e.g., older adults, adolescents, minority populations, co-morbid patients). Moreover, the average cost of the program per patient is usually less than $600! The strengths of this care management approach are obvious: amazing cost-effectiveness, effective symptom reduction, potential for implementing the model on large scales, and a perfect fit for experimental research. Despite the impressive résumé, there are several limitations to this approach that are a direct result of the philosophical underpinnings of the model. |
Care management approaches like IMPACT are a product of the disease model which assumes that disease is an abnormal, organic condition with symptoms that require medicine. Successful treatment includes symptom reduction or elimination. Providers using IMPACT track depression symptoms using a screening tool (usually the PHQ-9) and use treatment protocol to reduce those symptoms. If PHQ-9 scores go down, then treatment was a success. From this vantage point, depression is a discrete disease state that is qualitatively different from other mental illnesses like anxiety. Unfortunately, the biological etiology and pathway of most mental health disorders is largely unclear and rates of co-morbid mental illnesses (e.g., depression and anxiety together) are very high. This raises the question: are depression measures and treatments adequate enough to address the suffering of primary care patients?
On the upside, the disease model and experimental research approach is a match made in heaven. Clinical trials that use an intervention carefully designed to result in symptom reduction will probably produce outcomes of symptom reduction. That’s what they are supposed to do! But does symptom reduction equate to patient happiness or improved quality of life? Can a patient have a low PHQ-9 score and still be dissatisfied with her life? Treatment success then is evaluated using measures chosen by the researchers and not the patient. Isn’t this a little presumptuous?
Another drawback to the disease model and chronic care model overall is the narrow scope of treatment. Care management models are very good at targeting specific mental health conditions like depression. In fact, many care managers are referred to as depression care specialists. But what does a physician do when a patient experiences anxiety? Does the care team create another treatment algorithm for that new condition? Do you hire an “anxiety care specialist”? What if the problem is relationship-based or a result of community and environmental factors? In sum, the care management model is an approach rooted in the disease model that is great for experimental research and large-scale implementation but maybe not so great for comprehensive patient care.
Another approach to integrated primary care is the consultant model which includes generalists (behavioral health consultants or BHCs who are licensed mental health providers) who provide treatment for a wide variety of concerns. BHC appointments are usually 15-30 minutes long and utilize brief interventions to reduce functional impairment. Like care managers, BHCs share an office with primary care physicians and work within the same clinic system. BHCs may see 10-15 patients a day with follow up appointments limited to 4 visits. The evidence base for the consultant model is not nearly as rigorous as the care management model. The major strength of this approach includes the wide scope of treating various conditions (e.g., depression, anxiety, insomnia, chronic pain, substance abuse, smoking cessation, anger).
Limitations include experimental research challenges, costs, and training. Since BHCs treat a wide array of problems and focus more on functional improvement than symptom reduction, it is very difficult to design randomized clinical trials for this model. The consultant model will never reach the same number of successful clinical trials as the care management model. It’s impossible because experimental research (which is reductionistic and mechanistic like the disease model) does not adequately capture what BHCs do.
I personally like the simplistic beauty of the care management model. You can train a large number of providers to treat a large number of patients using limited resources. This alone is a great improvement over the current system. I also like the breadth of the consultant model. You can treat mental illness as well as behavioral health problems (e.g., insomnia, substance use, poor concentration). Each integrated care approach has strengths and limitations but which one is the best for patients? I recommend using the pragmatic truth criterion to answer this question. According to this criterion, the “right” or “true” way is the one that facilitates a desired consequence. In other words, if a primary care clinic needs an integrated care model that can measure precise outcomes and treat a specific population of patients without increasing costs, then that clinic should use the care management model. If a clinic needs a model that can address a wide array of problems, then use the consultant model.
|What if you could combine both models? |
But what if there was a way to combine both to create a hybrid model that makes research and training a breeze and also meets the needs of a wide variety of patients? Let me know when you find it! In the meantime, let’s be mindful of the tracks carrying this train we call “collaborative care”. As the Cheshire cat told Alice, if you don’t care where you’re headed then it doesn’t matter which way you go.
|Matt Martin, PhD, LMFT, is Blog Editor for the Collaborative Family Healthcare Association. When he is not blogging or editing he teaches behavioral science to family medicine residents at the Duke/SR-AHEC residency program. Interested in writing for the blogs? Email Matt at firstname.lastname@example.org|