Friday, January 1, 2016

New England Journal of Medicine Discovers Assertive Community Treatment

I have been a reader and subscriber to the New England Journal of Medicine (NEJM) every year since medical school.  One of the first courses they taught us in those days was Biochemistry and being an undergrad chem major I had a natural affiliation with many of the biochem professors.  The format in those days was lectures focused on the major topics and seminars to take a more detailed look at the experimental and theoretical aspects of the field.  They were fairly intensive discussions and critiques of research papers selected by the professors.  The department head was the mastermind behind this technique and one days he discussed his rationale for it.  He hoped that every medical student coming through that course would continue to read current research.  He strongly recommended subscribing to and reading the NEJM not just in Medical School but for years to come.  In my case it worked.

One of the sections that you don't hear too much about is the clinicopathological exercise that comes out each week.  It is basically a publication of formal case records of Massachusetts General Hospital and the associated findings and discussions.  These case reports are interesting for a couple of reasons - they show patterns of illness that clinicians can familiarize themselves with and they show at least some of the diagnostic thinking of experts.  During the time I have been reading them, they also discuss psychiatric comorbidity of physical illness and medical etiologies of psychiatric symptoms.  At one point I was a member of an informatics group and was very interested in studying this section of the NEJM from a psychiatric perspective.  At that time it seemed that I was the only psychiatrist with that interest.  With modern technology a study like this is more possible than ever.  For example, searching the case records feature of the NEJM from December 1989 to December 2015 yields a total of 31 cases of psychosis.  The etiologies of these cases range from purely medical etiologies, to delirium associated with the medical condition to pure psychiatric disorders with no specific medical etiology.  I have never seen this referred to as a teaching source for psychiatric residents admitting patients to acute care hospitals or consultation liaison services, but I could see it serving that function.  Instead of the usual lectures on medical psychiatry that typically contain PowerPoint slides of the "240 medical etiologies of psychosis" - a discussion of common mechanisms noted in these cases might be more instructive and be a better source for acquiring pattern matching capacity to broaden diagnostic capabilities.  It also put the DSM approach to psychiatry in proper perspective.  Knowing the lists and definitions of psychosis is nowhere enough to be a psychiatrist in a medical setting.  A seminar including this material can make these points and teach valuable skills.

That brings me to the case this week A Homeless Woman with Headache, Hypertension, and Psychosis.  Two of the authors are psychiatrists and the third is an internist.  The authors describe a 40 year old homeless woman with a diagnosis of schizophrenia and severe hypertension and how they established care over a number of years using the Assertive Community Treatment (ACT) model of care.  The patient's history was remarkable for a 12 year history of psychosis characterized primarily by paranoid and grandiose delusions.  She was homeless sleeping in public buildings for about 4 years and that seemed to be due to the thought that she needed to stay outside to watch over people.  She had a brief episode of treatment with olanzapine during a hospitalization about 5 years prior to the initiation of care by the authors, but did not follow up with the medication or outpatient treatment.  She was also briefly treated with hydrochlorothiazide 4 years earlier with no follow up care or medication.   She was admitted for treatment of a severe headache and a blood pressure of 212 systolic.  At the time of the admission physical BP were noted to be 208/118 and 240/130 with a pulse of 95 bpm.  She had bilateral pitting edema to the knees and bilateral stasis dermatitis.  She had auditory hallucinations consisting of voice of God and Satan and grandiose delusions.  Lab data showed a microcytic anemia.  She had standard labs to rule out myocardial infarction and vitamin deficiency states.  Blood pressure was acutely stabilized and she was discharged on lisinopril, thiamine, multivitamin, omeprazole, and ferrous sulfate.  The final diagnoses include schizophrenia, cognitive impairment associated with schizophrenia, hypertension, and homelessness.

The authors provide a good discussion of diagnosis of primary and secondary psychotic disorders and provide some guidance on timely medical testing for metabolic, intoxicant, and neurological abnormalities.  Delirium is identified as more of a medical emergency and necessitating more scrutiny.  The idea that delirium can be mistaken for psychosis is a valuable point that is often missed during emergency assessment especially if the patient has a pre-existing psychiatric diagnosis on their medical record.  The authors sum up screening tests that are necessary for all patient with psychosis and the tests that  are reserved for specific clinical concerns like encephalitis, seizures, structural brain disease, and inflammatory conditions.  They also suggest screening for treatable conditions and inflammatory conditions.

There is a good section on the follow up care that this patient received.  She was seen in a clinic for the homeless, where problems were gradually noted and worked on with her full cooperation.  This is not the typical approach in medicine where it is assumed that the patient will tolerate a complete history and physical exam and then cooperate with any suggested medical testing and treatment.  In this case, the practical problems of foot care were addressed.  She was eventually seen in 60 visits over two years.  By visit 19 she described concerns about cognitive symptoms and by visit 33 she was accepting treatment for psychosis with olanzapine.  She eventually allowed a more complete treatment of here associated physical symptoms including an MRI scan of the brain and treatment for migraine headaches.  The authors point out that tolerating medical and psychiatric uncertainty is a critical skill in treating people who need to habituate to medical systems of care.  A more direct approach is alienating.  It does tend to create anxiety in physicians about what is being missed and not addressed in a timely manner.  There is always a trade off in engaging people for long term care in more stable social settings and pushing to maximize diagnosis and treatment in a way that they might not be able to tolerate.  The ACT model stresses the former.           

There are some very relevant ACT concepts illustrated in this article.  First and foremost the rate at which medical interventions are prescribed depends almost entirely on the patient's ability to accept them.  This is at odds with the timeliness of medical interventions that most physicians are taught.  I say "almost entirely" in this case because the authors were very fortunate that the patient cooperated with treatment of extreme hypertension.  One of the common hospital consultations for psychiatric is a person with a mental illness and life-threatening illness who is not able to recognize it.  Even on the subacute side of care there are many tragedies due to patient with mental illness not being able to make decisions that could have saved their life.

I think that there are also some very practical applications for psychiatry on an outpatient basis.  Most patients with severe mental illnesses are never going to see a primary care provider 60 times before starting treatment.  It only happens in a subsidized setting with physicians who are highly motivated to see a certain approach work.  The care model described in the paper is certainly not the collaborative care model that some authors, the American Psychiatric Association (APA), and the managed care industry keeps talking about.  There is also the obvious point that people don't go into primary care because they like talking with people who have severe mental illnesses.  Psychiatrists need to see these people either in ACT teams or community mental health centers.  It won't work in a standard managed care clinic seeing a patient who is this ill - 2- 4 times a year for 10 - 15 minutes. ACT psychiatrists need to know about primary care providers who work better with the chronically mentally ill or people with addictions and make the appropriate referrals.  All psychiatrists should be focused on blood pressure measurements and work on getting reliable data.  Funding for psychiatric treatment often precludes ancillary staff present in all other medical settings to make these determinations.  Existing collaborative care models in primary care clinics can get blood pressure measurements on the chart but restrict patient access to psychiatrists.  

This Case Report is a good example of what can happen with a real collaborative care model that focuses on the needs of a person with severe chronic mental illness.  It is a model of care that I learned 30 years ago from one of the originators and it is more relevant today than ever.  It is also a model of care that is currently rationed and provided in the states where it is available to a small minority of patients.  It is not the method of collaborative care that you hear about from the APA, the managed care industry, or government officials.  It should be widely available to all psychiatric patients with complex problems.

George Dawson, MD, DFAPA


1: Shtasel DL, Freudenreich O, Baggett TP. CASE RECORDS of the MASSACHUSETTS GENERAL HOSPITAL. Case 40-2015. A 40-Year-Old Homeless Woman with Headache, Hypertension, and Psychosis. N Engl J Med. 2015 Dec 24;373(26):2563-70. doi: 10.1056/NEJMcpc1405204. PubMed PMID: 26699172.

2:  New England Journal of Medicine Case Records of MGH x psychosis (on Medline).  Shows 101 references as opposed to 31 on NEJM search engine and 10 on basic Medline search.

3:  Marx AJ, Test MA, Stein LI. Extrohospital management of severe mental illness.Feasibility and effects of social functioning. Arch Gen Psychiatry. 1973 Oct;29(4):505-11. PubMed PMID: 4748311.

4:  Stein LI, Test MA, Marx AJ. Alternative to the hospital: a controlled study.Am J Psychiatry. 1975 May;132(5):517-22. PubMed PMID: 164129.

5:  Test MA, Stein LI. Alternative to mental hospital treatment. III. Social cost.  Arch Gen Psychiatry. 1980 Apr;37(4):409-12. PubMed PMID: 7362426.

6:  Stein LI, Test MA. Alternative to mental hospital treatment. I. Conceptualmodel, treatment program, and clinical evaluation. Arch Gen Psychiatry. 1980 Apr;37(4):392-7. PubMed PMID: 7362425.

7:  Weisbrod BA, Test MA, Stein LI. Alternative to mental hospital treatment. II.   Economic benefit-cost analysis. Arch Gen Psychiatry. 1980 Apr;37(4):400-5. PubMed PMID: 6767462.


Photo at the top of this post is by Jonathan McIntosh (Own work) [CC BY 2.5 (], via Wikimedia Commons.  Original photo at


  1. Great post George. I am on a sub-internship right now. One of the residents predicted that in some settings he would be acting as a primary care provider for low functioning patients similar to this one because of compliance issues. This was after I had asked him about whether he worried about atrophy of his general medical knowledge. Do you think he is being realistic? While I recognize Psychiatrists like yourself take pride in staying up to date, others run to refer or consult when ever minor problems arise.

    1. He might be acting as a primary care provider. There are a number of models that I have practiced in along the way including:

      1. As the primary care provider.
      2. With a primary care provider in close consultation.
      3. With multiple specialty and subspecialty consultants and no primary care.

      No psychiatrists should let their medical skills atrophy. What your learn in medical school on the diagnostic side holds up very well as long as you pay attention to the non-psychiatrists you work with. A lot of what we do depends on getting medical history that primary care physician's can't get and knowing how to synthesize it. I don't think you need to know state of the art care (and let's face it that changes a lot) - but you do need how to identify the problems, suggest evaluation, and in may case suggest specific consultants. You also need to advocate for the patient. If they are seeing a doctor and their medical problem is not being solved - one of the roles I have taken on is how to get that problem resolved.

      I caution every early career psychiatrist to NOT tell any prospective employer that you have excellent medical skills and take care of medical problems because you will end up doing twice the work and only get paid for your job. And I doubt that the productivity expectations will decrease. Insist on adequate medical coverage and same day consultation for medical problems that you identify. Also insist on the availability of timely labs testing, ECGs and imaging studies.

    2. I forgot to add a bit about medication reconciliation. One of the main reasons why hospital based psychiatrists need to stay current is that they are usually responsible for making sure that ALL of the patient's medications are restarted in the hospital at the correct doses. The new jargon for this courtesy of the EHR and the Joint Commission is medication reconciliation. Depending on the patient this can be a huge time drain if you are treating medically complex patients on 15-20 medications for various conditions and nobody has taken a look at how all of the medication interact. I have spent an additional 60-90 minutes in some cases just trying to find out what the patient was taking and the correct doses.

      Any early career psychiatrist looking at a new inpatient job needs to ask about how that reconciliation is handled or it will be a major source of frustration. The administrator's solution is for you to just stay longer to complete all of the work.

    3. Thanks for the tip. I've seen problems with medication reconciliation (and inadequate formularies) in my brief time so far.