Welcome to my blog, a place to explore and learn about the experience of running a psychiatric practice. I post about things that I find useful to know or think about. So, enjoy, and let me know what you think.

Tuesday, September 9, 2014

A Duck!

Not talking about emotional support ducks, here. Nor am I referencing Shrink Rap's mascot, cute as it may be. I'm talking about the collaborative care model. I enthusiastically checked out "new and interesting clinical updates from UpToDate", the psychiatry section. And this is one of the things I learned:

Collaborative care for patients with depression and medical illness (May 2014)

Collaborative care that integrates mental health specialists and case managers into primary care practices is an effective way to treat depression and may also improve general medical outcomes. In a meta-analysis of seven randomized trials that included 1895 patients with comorbid depression and diabetes, both depression scores and glycemic control demonstrated greater improvement with collaborative care than with usual care.

Having written about this exact topic a little while back, I wasn't sure where they were drawing their conclusion from. So I went to the link and found this:

 Many studies demonstrate that collaborative care improves depression outcomes:
A meta-analysis of 37 randomized trials (n >12,000 depressed patients) found a significant but clinically small effect favoring collaborative care over standard primary care, which persisted for up to five years; heterogeneity across studies was moderate.
A subsequent meta-analysis of 79 randomized trials (n >24,000 depressed patients) that compared collaborative care with usual primary care (eg, pharmacotherapy alone) found that after six months, response (reduction of baseline symptoms ≥50 percent) occurred more often with collaborative care (relative risk 1.3, 95% CI 1.2-1.4); heterogeneity across studies was large. The benefits of collaborative care persisted for up to 24 months.
In meta-analyses of randomized trials that compared collaborative care with usual care in depressed patients with chronic conditions (eg, arthritis, asthma, cancer, coronary heart disease, diabetes, and HIV), improvement of depressive symptoms, psychosocial functioning, and mental and physical quality of life were greater with collaborative care that usual care. However, other analyses found that diabetic control and all cause mortality were each comparable for the two treatment groups.

I'm not sure how this demonstrates improved glycemic control.

Another study that I know of, by Katon, et al, and published in NEJM in 2012, looked at collaborative care for patients with depression and co-morbid diabetes and/or coronary heart disease. This study found significant blah blah blah. It also changed its primary outcome measures during the course of the study:

The initial primary outcome was the percentage of patients achieving disease control at 12 months on all three disease-control measures...This outcome was changed in August 2009 (before study data became available for analysis in November 2009)...

I assume they're telling us the outcome measure was changed before study data became available so we won't think there was any bias. They didn't know the outcome yet, right? Well then, why change the outcome measure?

NEJM. Oh, yeah.

This is an excerpt from the description of the study intervention, just so we can establish what goes on in a collaborative care model.

The intervention combined support for self-care with pharmacotherapy to control depression, hyperglycemia, hypertension, and hyperlipidemia. Patients worked collaboratively with nurses and primary care physicians to establish individualized clinical and self-care goals. In structured visits in each patient's primary care clinic every 2 to 3 weeks, nurses monitored the patient's progress with respect to management of depression (according to the PHQ-9 score), control of medical disease, and self-care activities. Treatment protocols guided adjustments of commonly used medicines in patients who did not achieve specific goals...First-line agents included...citalopram or sustained-release bupropion for depression. Nurses followed patients proactively to provide support for medication adherence.

Using motivational and encouraging coaching, nurses helped patients solve problems and set goals for improved medication adherence and self-care...Patients received self-care materials, including The Depression Helpbooka video compact disk on depression care, a booklet and other materials on chronic disease management, and self-monitoring devices...

Nurses received weekly supervision with a psychiatrist, primary care physician, and psychologist to review new cases and patient progress. An electronic registry was used to track PHQ-9 scores and glycolated hemoglobin, LDL cholesterol, and blood-pressure levels. The supervising physicians recommended initial choices and changes in medications tailored to the patient's history and clinical response. The nurse communicated recommended medication changes to the primary care physician responsible for medication management.

This is a link to the supplementary appendix. It includes, among other things, the power point they used to teach the nurses the study intervention. It's too upsetting to describe. Take a look, if you have a strong constitution. 

About the same time I got the UpToDate email, I read an article in the July 4th edition of Psychiatric News, The First and Last Word in Psychiatry. (It's the APA). It's by Jurgen Unutzer, one of the pioneers of integrated care, and it's entitled, Case Study: Heartbreak and Lessons Learned. In it, Dr. Unutzer laments the abandonment of a collaborative care model by a clinic, the Federally Qualified Health Center (FQHC). It was a 3 year program that worked spectacularly in its first two years. But then there were staff changes, and poor financial planning, and the program was abandoned. "Most importantly, the organization never fully integrated the program into its overall mission, vision, and clinical operations." The poor, lost, misguided souls.

Reading this made me think of the duck.

"Why is collaborative care like a duck?" you ask. Not for the same reason a raven is like a writing desk. 

It goes like this:

Patients seen in primary care clinics often have poor access to mental health care. 
We know that most prescriptions for antidepressants are written by primary care doctors. 
Therefore, all patients seen in a primary care setting should be administered a PHQ-9.
The best treatment for those who qualify as "depressed" is a care manager.
Psychiatrists are trained to treat patients with depression, and should therefore have no contact with the patients in a collaborative care program.
The care manager will administer sequential PHQ-9s via phone contact, to determine if the patient gets better on the medication that is recommended by the psychiatrist and prescribed by the primary care provider. 
When these patients say they are more satisfied with their lives because of the collaborative care model, this means that their diabetes has improved. 
Therefore, all psychiatric care should be practiced via the collaborative care model.

I suspect the reasoning is based on the following logical sequence, from the witch scene in Monty Python and the Holy Grail:

Terry Jones: There are ways of telling whether she is a witch.
Rabble: Are there? What are they? Tell us!
TJ: Tell me, what do you do with witches?
R: Burn them!
TJ: And what do you burn apart from witches?
R: More witches!
R: Wood!
TJ: So, why do witches burn?
(Confused pause)
R: Cuz they're made of...wood?
TJ: Good!
TJ: So, how do we tell whether she is made of wood?
R: Build a bridge out of her!
TJ: Ah, but can you not also make bridges out of stone?
R: Oh, yeah.
TJ: Does wood sink in water?
R: No, it floats. It floats!
R: Throw her into the pond!
TJ: What also floats in water?

Very small rocks!
Lead! Lead!

Graham Chapman: A duck!
TJ: Exactly. So, logically:
R: If she weighs the same as a duck, she's made of wood!
TJ: And therefore?
(confused pause)
R: A witch!


  1. These are very bad studies, and I think that the PHQ-9 might be the most depressing thing ever invented. I can't help thinking, though, that there ought to be a better way to integrate care. if only because so many people with really serious mental health problems get such bad medical care.

  2. Great idea for a new study! Does the PHQ-9 cause depression in clinicians?

    One hopeful type of integrated care is the CoMeBeh project, at the University of Iowa, in which primary care is provided by physicians rotating through the psychiatry clinic, rather than the other way around. See: http://thepracticalpsychosomaticist.com/2013/11/20/comebeh-follow-up-further-on/

  3. I find reading about collaborative care to be boring and I won't do it (I don't meant this blog). I do, however, use the PhQ-9 and the GAD-7 in my outpt practice which is mostly psychotherapy. Why? I worked for a non-profit HMO (not Kaiser but it could have been) and got used to it. We HAD to use it unless a pt absolutely refused and few did. I learned things about my pts by using it and I could follow their progress and so could they-- I'm too tired from a long day to fully explain why I still use it when i don't have to any longer (I'm my own boss now) but I still learn a lot by using it.

  4. I think that the PHQ-9 might cause depression in patients myself.

    I once went to a patient education program at Man's Greatest Hospital where they showed a video and did a little survey asking for feedback on how well a physician communicated with the patient. I think that the doctor was reviewing a PHQ-9 and offering psychological and medication options. The doctor was doing everything that they were trying to evaluate her on, but I still wrote that she didn't, because she seemed so lifeless. Getting all of the individual interventions right did not make the whole person effective.

    And I always overthink those questions and screw them up. I took some sort of psychological test once when I was pretty depressed where they asked "Do you feel that you are losing your mind?" I was depressed and having a hard time concentrating and reading and thinking, so I answered truthfully that I felt like I was losing my mind. This frightened the postdoc doing the assessment until I explained what I meant on follow-up. Luckily, this was semi-structured testing.

    In my utopia, there would be either psychiatrists or psychiatrists integreated into pediatrician's offices. Every child would have an annual well-child psychological exam so that evaluating developmental milestones wasn't left to general practice pediatricians.

    I think it might catch serious problems earlier while allowing for watchful waiting and normalize the experience of seeing a mental health professional. Wishful thinking, I know.