Sunday, July 2, 2017

Collaborative Care Just Gets Worse.....

I am a long time opponent to the expansion of the collaborative care model and have explained why in earlier posts on this blog.   At the Minnesota Psychiatric Society (MPS) conference last week, I learned that the collaborative care model had expanded to more than just the treatment of anxiety and depression.  The presenter discussed an expanded model to treating bipolar disorders based on questionnaires based screening for that disorder.  The overriding rationale for this model is that psychiatrists can't possibly see all of the patients with mental illness, therefore a more  hands off approach to care was acceptable.  The presenters were very explicit about the model not involving direct patient care in the primary care clinic.  The concern is the psychiatrist would start to to develop their own practice in the clinic and within several months their schedule would be full and they would have no capacity to see anyone else.  I can say from my experience that a primary care examination room is the wrong setting to do psychiatric consultation.  At the minimum a psychiatrist needs a service where they can take detailed notes.  Scribes are apparently on the rise these days.  I would be be very concerned about the training necessary for a scribe to record the details that I consider to be important and remain in the background during the interview.  I am a purist and believe that another person in the room produces a different interview.

The argument about expanding the collaborative care model fails at the level of the total number of psychiatrists and the total number of people needing care by psychiatrists.  Being medically trained I have always defined those people as having the most severe forms of mental illnesses.  That is the essence of having a defined number of physicians for any population and it works very well for other specialties.  The ones I have written about here include ophthalmology and orthopedic surgery.  Despite having fewer physicians available, both of these specialties cover a much larger spectrum of eye, bone, and joint disease and trauma.  They are seeing a larger number of patients and in many cases performing lengthy operative procedures on these patients.

The collaborative care model has rapidly evolved in the hands of the APA from the Diamond Project of about a decade ago.  The original Diamond Project involved collecting PHQ-9 scores in primary care setting and having case managers remain in touch with patients for supportive counseling and to review the progress of patients based on those scores with psychiatrist.  The psychiatrist recommended medication changes in order to improve treatment of the depression and improved PHQ-9 scores.  The state of Minnesota took this one step further and decided to implement widespread reporting of PHQ-9 scores from all primary care clinics as part of an accountability initiative called Minnesota Community Measurement.   Lacking any scientific or statistical merit did not slow down the politics of the least accountable (politicians) holding the most accountable (physicians ) - even more accountable.  At least one group of experts has come out against the idea of depression screening, because using the current models it eventually equates to more antidepressant exposure.  That has not slowed down health plans in the state of Minnesota or national organizations that essentially represent health plans. So far, I am unaware of any reporting of PHQ-9 changes.  I sent the project an e-mail about 5 years ago pointing out that their statistical approach was meaningless on a longitudinal basis - so it will be interesting to see what they eventually report.      

The course presented was Applying the Integrated Care Approach: Practical Skills for the Consulting Psychiatrist.  It was presented as an official American Psychiatric Association backed course and part of the Transforming Clinical Practice Initiative.  Since I have never heard of this initiative before I just assumed it was another in a series of top down decisions by an organization that I thought was supposed to support its members.  I would include the very unfavorably rated Maintenance of Certification initiative to be another in that series.

I will proceed to the end product to illustrate the general feel of this course for experienced psychiatrists.  Every psychiatrist has had on-call experience.  During those times it is common to be operating in a decision-making environment where there is either inadequate or partially adequate information to make a decision.  An example is being on call and admitting patients by some combination of phone calls or internet network connections or both.  A new patient comes in at 10 PM, it is impossible for the psychiatrist to get up and drive to the hospital to do a comprehensive admission evaluation on each patient, so temporary orders are given over the phone, until the staff psychiatrist can see the patient and refine the process in the morning.  In the uncomplicated process, this is an easy task.  The healthy patient comes in taking fluoxetine 20 mg.  The medication is continued until the next day.  But things can get much more complicated in a hurry.  What happens when you are asked to write the on-call orders for a bulimic patient with depression on bupropion who may be in alcohol and benzodiazepine withdrawal?  Or the patient who has been on escitalopram, using methamphetamine, and is complaining of some symptoms of serotonin syndrome?  What happens when a sixty year old patient comes in taking 10 different medications for hypertension,  diabetes mellitus, and atrial fibrillation?  Medications need to be modified or held and significant additional plans need to be implemented.  These are the kinds of calls that you will be making in the APAs integrated care model.  The only difference is that they will be strictly regarding psychiatric medications, but they will be all of the medications and more than just antidepressants and anxiolytics.  You must be prepared to treat bipolar disorder by proxy on partial information and assume the primary care physician has the skill set to take it from there.

 The screening instrument for bipolar disorder is the CIDI-3 developed by the World Health Organization for lay screening of large populations.  I had absolutely no luck in locating CIDI-3 anywhere on the Internet or the WHO website.  I was able to locate this Harvard site containing containing what appear to be numerous sections of the Comprehensive International Diagnostic Interview (CIDI).  To anyone familiar with structured interviews (DIS, SCID, SADS, etc) it is a the same technology.  The CIDI-3 screen described in the PowerPoint for the course had two stem questions - one for euphoria and one for irritability.  Neither of them matched my stem questions due to a lack of duration criteria and no rule outs for medical or substance use problems.  It is also not clear about how a consulting psychiatrist is going to learn about the pattern of illness from these screens.  The it seems that the precedent set by the PHQ-9 and GAD-7, that a positive screening equals diagnosis - also applies in this case.      

As I thought about all of the work that is involved in the quality treatment of bipolar disorder, I asked myself about whether all of that work and all of the necessary information transfer to the patient and family can be accomplished in a primary care setting.  There is also the idea that a medication cures the problem.  Although bipolar disorder is undoubtedly one of the most biologically based psychiatric disorders, it takes plenty of skill in managing side effects, associated symptoms (especially anxiety and sleep), and additional supportive psychotherapy.  There is also the issue of assessing suicide potential and generally functional capacity including risk for aggression but most importantly the ability to care for oneself.  In psychiatric practice - each of those dimensions amounts to an additional primary care visit.  All things considered, I don't see bipolar disorder or any type being assessed and managed well in primary care settings with a psychiatrist phoning it in.  The lecturer in this case had ample justifications - but to me that is all a reaction to excessive and continued rationing of psychiatric services.

And speaking of rationing - the money was discussed.  First - the psychiatrist in these consultations does not submit any billing.  The primary care clinic submits a collaborative care billing code and then they reimburse the psychiatrist.  At no point in my career as a physician employee have I ever seen an exchange like this occur where an administrative fee was not tacked on - just for the purpose of cutting the check I guess.  Second - there is all sorts of hype about how these arrangements save money in primary care settings.  Since managed care stole the field of medicine 30 years ago - there are ad nauseum articles written about cost-effectiveness.  To me it is just another buzz word for managed care.  There is no reason to expect that treating severe psychiatric disorders should be any more cost-effective than treating severe non-psychiatric medical disorders - in fact, one often leads to the other.  The lecturer in this case was very honest about that.  He pointed out the two studies that claimed costs savings and bluntly said that he doubted that would apply to clinical situations.

All things  considered, collaborative care continues to leave a bitter taste  in my mouth.  It translates to second class care for psychiatric patients based on managed care rhetoric.  The argument can be made that these are not psychiatric patients - but primary care patients who would never see a psychiatrist.  I don't know  if that is really a legitimate argument or not because it comes down to legal and political convention rather than professionalism.  In that case it depends what faction ultimately "wins."  The APA has clearly adopted it and it openly promoting it.  At the end of this course, there was the doubly ironic offer to enroll in an online collaborative care course that would result in both CME credits and also MOC credits for maintenance of certification.

I don't know how covering call suddenly becomes psychiatric innovation.

George Dawson, MD, DFAPA


1:  John Kern.  Applying the Integrated Care Approach:  Practical Skills for the Consulting Psychiatrist.  Presented at the 2017 MPS Spring Scientific Meeting; Thursday June 15, 2017 at 1:00-5:00 PM.


Above image is from National Severe Storms Lab (NSSL) web site and reproduced here per the NOAA intellectual property notice.


  1. These people do not seem to know the meaning of the word "screening." It means lots and lots of false positives - by design.

    Using a screening test to diagnose bipolar disorder is just another way to expand the diagnosis to people who do not even come close to having it (preventing that from happening is why the duration criteria are in the DSM in the first place), and to expand the market for potentially toxic antipsychotic medications so that people who do not need them will be buying them and taking them - particularly people who suffer from the diagnosis I specialize in, borderline personality disorder. Mark Zimmerman has already done studies showing people with that diagnosis being misdiagnosed with bipolar disorder at an astonishingly high rate.

    I know you don't think PHarma is part of this but I think that much is becoming clearer every day. People are being bought off - I suspect at the highest level of the APA - and have been pushing this horrid and downright evil model.

    1. You are correct about my opinion about Pharma. Antipsychotics that work well in these situations are all generic with the exception of newer atypicals and of those lurasidone is more difficult to prescribe than a lot of people think. I think the bigger exposure will be lamotrigine which is broadly prescribed for anyone with suggested irritability or even anxiety.

      But more importantly - I have witnessed the process first hand as a product of managed care companies and their ability to get what they want codified by the government.

      Pharmaceutical companies would never advocate for the long term cross sectional collection of rating scale scores not necessarily completed by the same patient. They have statisticians who know better.

      The APA doesn't know how to deal with the "cost effectiveness" rhetoric that is unevenly applied to psychiatric services and sees this as an effort to provide cost effective care.

  2. A while after I published this post about Pharma disease mongering by the lead author, who denied to me that he was pushing antipsychotics for bipolar disorder, Sachs was a co-author of the article in JAPA showing that Lurasidone was "effective" in bipolar depression. This is not an isolated example by any means.

    1. Thanks for that link.

      One of the exercises that came up in the above conference was the collabo care psychiatrist coming up with a plan to treat a patient with either agitated depression or bipolar 2. The agitation was to the point that her ability to function was affected. There was really insufficient data to make a clear diagnosis, but the task was to advise the primary care MD and not start seeing these patients by yourself because the party line is that the psychiatrists schedule would fill up in 2 months and at that point there would be no way to see new acute referrals.

      At any rate most of the audience went with lamotrigine. There was only one other acute care psychiatrists there and she and I went with quetiapine. I think that there is a bias against antipsychotics, when in fact they are most likely to get the job done in acute settings and have always been used for maintenance therapy.

      When I first started practicing, it was apparent to me that during the deinstitutionalization era, a significant number of the chlorpromazine discharges were people with bipolar disorder. In many cases they were stabilized with monotherapy. The people I saw tolerated it well, did not get tardive dyskinesia, and many returned to their professional careers and ended up retiring from that career. They took chlorpromazine for decades and it restored their professional and family life.

      I am certainly not advocating a return to chlorpromazine. It is one of several medications that I have never initiated in my life due to toxicity. My point is that there is no doubt that antipsychotics work for bipolar disorder, only half (or less) of bipolar patients can be treated with mood stabilizers alone, and every medication we use today is probably safer than chlorpromazine.

      To me there are no inherently bad medications especially if the problem is severe. One of the main jobs of a psychiatrist is to determine if the patient can tolerate a medication and determine what medication needs to be used based on urgency. Collaborative care also takes this psychiatric function out of the loop and suddenly medication used is being dictated by other considerations.

    2. Yes, antipsychotics are good for mania, but I only used them for acute stabilization - or if the patient fails a trial of or cannot tolerate, lithium, valproate, or carbemazepine. Tardive is a major risk, although of course many patients do not get it. Metabolic syndrome is an even bigger risk with the atypicals. (If a patient is psychotic, I have NO problem prescribing anti-psychotics).

      Antidepressants are highly effective for bipolar depressions as long as a patient is on an anti-manic drug to prevent switching. Antipsychotics alone are far, far, far less effective, except for augmenting an antidepressant if that is necessary. Lamotrigine is worthless in bipolar - those doctors use it for "bipolar II" - a truly non-existant disorder IMO, although there are mild bipolar I's who respond to lithium - because what they are treating is in fact the affective instability associated with cluster B personality disorders. Lamotrigine does help that some, although it is not nearly as effective as the combination of an SSRI and a long acting Benzodiazepine (MAOI's plus the benzo also works).

      There are almost no studies about the later combination because the drug companies know it works and they want you to prescribe things that are more expensive. I've been treating borderline personality disorder since the 70's, and I know this combo is very helpful. And BPD's do not abuse them any more than anyone else does (most don't unless they have a pre-existing substance abuse problem that is not just due to self medication for panic disorder) if you are seeing them in regular psychotherapy, which I do.