Saturday, October 29, 2016

More Than 9 Questions About Sleep

I was just thinking about the PHQ-9 and it widespread use in managed care as a metric for depression.  The idea that 9 questions are all that is needed strike most psychiatrists as a gross oversimplification.  There has been plenty of debate over the years about the diagnostic criteria and waht should be included.  Kendler recently wrote an excellent paper on the fact that the diagnostic criteria as they stand in the DSM really indexes disorders rather than diagnosing them.  Some recent blog posts have looked at real patients and what is happening with them when they appear to have an elevated PHQ-9 score but are not depressed.  Past markers of psychiatric disorders like the dexamethasone suppression test had had to withstand more rigorous testing than the PHQ-9.  And lastly, the literature to support it seems to reflect the literature that justified managed care - a business concept with no basis in science or medicine.

And then I had the thought: "During an evaluation I ask more than 9 questions about sleep on the average."  By comparison the PHQ-9 has one question.  That question is:

3. Trouble falling or staying asleep, or sleeping too much?  
[Not at all] [Several Days] [More than half the days] [Nearly every day]

There are more complex sleep questionairres.  The Pittsburgh Sleep Quality Index (PSQI) and Functional Outcomes of Sleep (FOSQ, FOSQ-10) are good examples.  I thought I would tabulate my questions here.  Sleep disturbance can be a primary disorder independent of any psychiatric problem.  Sleep is also comorbid with many if not most psychiatric problems.  Most people do not recover from a priamry psychiatric disorder as long as their sleep is disturbed.  Sleep disorders can antedate the onset of mood and anxiety disorders by years and for that reason I think it is important to determine if the sleep disorder is primary rather than part of the mood or anxiety disorder.  That cannot be determined by a brief cross sectional look that considers all current symptoms as part of a mood disorder.  So during my standard evaluation I ask people the following questions about their sleep not necessarily in the following order:

When you were a kid in middle school or high school did you have trouble sleeping?
Did you have nightmares back then?
Did they occur early in the night or later in the early morning hours?
Did you sleepwalk?
Did anyone ever tell you that you had sleep terrors?
Did your sleep problems from childhood ever resolve - have you ever slept normally since then?

Do you work at night?
Do you do shift work where the work time changes?
Do you currently have sleep problems?
Do they occur when you try to fall asleep?
Tell me your experience of trying to fall asleep - what gets in the way?
Any idea how long it takes you to fall asleep?
Have you had sleepless night where you could not sleep at all?
Do you wake up off and on all night long?
Do you wake up early in the morning - like 4 or 5 AM and find that you can't fall asleep?
Is your sleep restorative - do you feel rested the next day?
Do you snore?
Have you ever had a sleep study?
Have you been diagnosed with a sleep disorder?
Do you take alcohol or any medication to help you fall asleep?
Do you take in many caffeinated beverages during the day as coffee, soda, tea, or energy drinks?

That is about 20 fairly basic questions about sleep.  It is a framework that requires elaboration.  Just the issue of sleep studies these days can lead to details about parasomnias, related surgeries, sleep disordered breathing diagnoses, restless leg syndrome, and all of the associated treatments.  For the pupose of this post that is about 18 questions or twice as many as the total on the PHQ-9. It should be apparent that severe sleep problems can lead to a score considered in the depressed range on the PHQ-9 by adding up the scores of questions 1, 3, 4, 7 and 8.  Critical distinctions need to be made between sleep problems, anxiety disorders, mood disorders, addictions, and also the excessive use of an adult attention deficit-hyperactivity disorder diagnosis.

This brief exploration should point out the problems with a screening versus a diagnostic evaluation.  If you are given a PHQ-9 or GAD-7 (for anxiety) in your primary care clinic that score alone is insufficient as a basis for a treatment plan for depression.  A person repeating the questions as confirmation does not constitute a diagnostic evaluation.  By itself it does not mean that an antidepressant prescription is indicated.

Ask that person how they know that you have depression and not a sleep disorder or something else.

George Dawson, MD, DFAPA

Waiting To Call An Ambulance Is Not Much Of Plan

I don't like to write about my own health problem - but it is a ready example and I already have another blog about it so here goes.  I also don't need to worry about violating my own confidentiality.  It involves a personal medical problem called paroxysmal atrial fibrillation.  I have consulted 5 different Cardiologists and the rhythm problem is not due to valvular or coronary artery disease.  It is probably due to excessive exercise - specifically exercise with sustained high heart rates.  After a period of frequent episodes, I started taking a generic brand of flecainide 4 1/2 years ago and have not had an episode since.  During that time I have had two episodes of influenza and 1 episode of acute bronchitis requiring prednisone therapy with no recurrence of atrial fibrillation.

Lately I have been seeing patients and about 40% of them have an upper respiratory virus and the various complications.  I knew it was just a matter of time.  Earlier this week I developed a cough, sneezing, facial burning, and a headache but no additional flu like symptoms.  It is not flu season here, but respiratory viruses abound.  Monday, Tuesday and Wednesday morning - I was awakened at 4 AM with an intense flurry of palpitations.  Taking the pulse showed a pattern of 4 or 5 regular beats followed by what seemed like a pause or dropped beat that I recognize as the early transition (I think) to atrial fibrillation.  In each case I drank a large glass of water, paced for a minute or two and I was back in sinus rhythm - the palpitations resolved.  Initial BP check was about 130/80 with a pulse of 88 rapidly back down to 110/70 with a pulse of 58.  The last readings are my typical baseline and I check them four times a day.  I know how much physical, mental, and emotional exertion affects those readings and I try to stay cool.

When it happened Wednesday, I decided to do the responsible thing and take the day off of work and see my primary care physician.  He did a physical exam, ordered labs and an ECG.  Everything was normal.  That is my ECG from that clinic visit at the top of this post.  It is normal sinus rhythm with a slight bradycardia (less than 60 bpm).  An interesting homage to artificial intelligence is that the ECG machine communicated with the electronic health record (EHR) and determined that there was no appreciable change between this ECG and one I had done 10 months ago.  The cardiologist is out of that loop.  The bottom line is that the tests were all negative and the plan was to see what happened and consider a Holter monitor if it persisted.  A Holter monitor in this clinic is a 48 hour recording of the ECG looking for discrete events that might suggest a cause of the rhythm disturbance.  It also allows the patient to mark any episodes of subjective disturbance on that record.

Yesterday morning I felt a little tachycardic at about 5 AM and got up and checked.  Heart rate was 66 and blood pressure was fine.   This AM, a flurry of palpitations wakes me up.  They are gone in two minutes after pacing and drinking water.  There was no chest pain or lightheadedness.  In fact when I had the initial episode about 10 years ago, I was speedskating and my heart monitor showed a rate of well over 200 bpm.  No chest pain or lightheadedness at that time and I drove to the hospital and told them I was in atrial fibrillation.

Today I respond to my primary care physician's note though the EHR and describe what happened.  I recall that he is not in, so I go back to the EHR,  agree that I can be billed if this is not a problem that I have been seen for in the last 7 days and attempt to cut and past my note to my primary care doctor into a separate email to his team.  The EHR cuts me off because it says that I can only use 255 characters.  It is the Twitter of EHRs.  I edit it down and send it - no response to my request for the Holter monitor.  I call the clinic and get on the phone with a triage nurse.  The conversation goes something like this (not a transcript):

Me: (Relating the history and Holter monitor request).
Triage RN:  "Well what is the emergency here?  It is Friday afternoon, there is no way that we are  going to get a Holter monitor today.  It will be Monday at the earliest. Your doctor can call it in then"
Me (a little steamed): "Maybe you could suggest criteria that I can use to call an ambulance."
Triage RN: "What?"
Me: "You know - when I wake up from a deep sleep with this arrhythmia at 4AM tomorrow morning, what criteria should I use to decide when to call an ambulance?"
Triage RN: "I did not know it was still happening."
Me: "It happened this morning.  That is why I e-mailed and called you.  That is why I stayed home from work."
Triage RN: "Well in that case I will run it by one of the attendings who is here and ask them about what should be done."

After another call back to get more of the usual information about cardiac symptoms, the Triage RN called again and connected me with the Holter Monitor tech.  I can apparently get in next Wednesday.  He told me the entire procedure would take 5 minutes so I would only have to miss a half day of work instead of a full day.  I did not pursue the obvious "Well why can't I just drive down there now and have it put on."  Everyone must be scheduled.  Schedules must be adhered to.

So that is where it stands tonight.  All of the bullshit that passes in the press for medical news does not apply.  There is no IBM Watson computer out there that knows more than I know about this condition or how to treat it.  There is no personalized medicine.  I have not encountered a single cardiologist interested in the genetics of atrial fibrillation or why I might have it.  Most  physicians assume I have neglected hypertension or have done something wrong with regard to my self care and therefore deserve it.  I still encounter physicians who doubt that I have never smoked a single cigarette in my life - even though it is true.  Hard to believe that somebody could bring this on by excessive exercise.  Isn't exercise supposed to be good for you?

I am probably being overly dramatic.  This is most likely a benign atrial arrhythmia.  On the other hand - why am I so certain if my physician wants another Holter?  I did a Holter and a longer event monitor 5 years ago.  I run a heart rate of 130 bpm during 4 hours of exercise per week and have tolerated a sustained heart rate of 140 bpm from a medication side effect - calmly pacing and taking incremental amounts of beta blockers to slow it down.  The final instructions from the triage nurse were to get to a hospital if a sustained heart rate of 120 bpm or greater and call an ambulance if chest pain.

Personalized medicine in the early 21st century is in many ways inferior to medicine the way it was practiced in the 20th century.  In those days, there may have been an interested physician who said: "Spend a night on telemetry and we will see if we can capture the beats and figure out what to do about it."  I saw those people being admitted when I was a medical student and an intern.  That was before you had to be dying to get into a hospital and the admission rules were dictated by case managers.  In those days personalized meant a long term personal relationship with a real physician who  could make things happen.

Now like me - those people are sitting at home waiting for something to happen and guessing about when they should call an ambulance.

George Dawson, MD, DFAPA


Don't try any of this at home.  This is not medical advice.  Only your personal physicians and consultants can give you that advice.

Friday, October 28, 2016

NEJM Three Part Series - Severe Medical Illness In the Context of Severe Mental Illness

Lisa Rosenbaum, MD has written a three part series in the New England Journal of Medicine that should be read by non-psychiatric physicians, family members, patients, and policy makers.  It is focused squarely on providing medical care to patients with severe mental illness and all that involves.  It is really impossible for psychiatrists to avoid the issue.  At some point in your career you find yourself in a situation with a patient who has a severe medical problems and refuses to address it because of the way that their decision making process has been impacted by mental illness.  If you are an inpatient or ACT team psychiatrist - it is usually up to you to come up with a plan to address that problem,  Several of the scenarios she describes across this series are directly from those settings.  Like any other specialty, psychiatrists will all migrate toward a certain niche.  For many reasons being that person who has to confront mentally ill patients about the fact that they are seriously ill or dying is not a position that is in great demand.  But mostly it is because inpatient and ACT team positions are rationed and none of the payers or administrators want those psychiatrists to do what they are capable of.

This series is part of an overall increase in psychiatric topics that are discussed in the NEJM.  So far this year there have been three Case Records of the Massachusetts General Hospital and a review of the basic science of addiction and a number of other articles on addiction and mental health policy.  Although I have not attempted to quantify it, there does seem to be a general increase in the coverage of psychiatric topics that include reviews of diagnostic groups, basic science mechanisms, clinical trials, and articles of general interest from the field.  The search function of the web site allows for grouping according to specialty and there are generally 20-30 psychiatry references per year over the past 5 years.  Lisa Rosenbaum is a cardiologist and is a national correspondent for the NEJM and in that field has probably seen a number of cases of people with severe mental illnesses and cardiac problems.

In the first article in the series she discusses the issues of informed consent and coercion in treating patients with severe mental illness.  It is well known that this population is undertreated from the perspective of primary treatment that is focused on the mental illness.  She uses an estimate of 9.8 million people with severe mental illness and only 60% of those people receiving treatment.  She cites the systemic problems of a lack of bed capacity 11.7/100,000 currently compared with 337 per 100,000 in 1955.  The contrast is more striking if per capita health expenditures for health and the number of psychiatric care beds per 100,000 population are compared.  For OECD data per capita expenditures for healthcare rank the USA at the very top by far and the per capita number of psychiatric care beds near the bottom compared with other countries.  The article discusses how deinstitutionalization was supposed to be linked to more community resources and not less.  The underfunded systems is portrayed as somewhat of a mysterious problem rather than system-wide rationing of psychiatric services.  The author in this case ties the underfunding to problems that it can't fix namely severely ill patients not seeking treatment  and antipsychotic drug side effects.  The high discontinuation rates in the CATIE trial are cited as proof of this problem.  I think that underfunding due to rationing is the problem.  With adequate resources comprehensive care is possible when the focus is comprehensive care more than medication.

The philosophy behind ignoring psychosis due to a psychiatric disorder compared with other physical illnesses is discussed.  The author points out that given the severe disability of these illnesses they cannot be ignored and that unlike other illnesses there are no major campaigns focused on prevention or treatment.  From there she transitions into the issue of intervention against the patient's wishes.  She discusses Stone's theory of paternalistic intervention if the affected person is likely to be grateful after they have been stabilized.  The flaw with that theory is that in many cases only a degree of psychiatric stabilization can be achieved.  In many cases it is likely that the person will have continued problems with insight and medical decision making and will continue to disagree with any suggested treatment.  She discusses the flaws with "dangerousness" as a component but does not take that discussion to its logical conclusion.  Dangerousness is of course not a psychiatric concept.  It is used by the courts as a basis for the initial stages of civil commitment and by managed care companies to decide if they will pay for psychiatric hospitalizations.  Any inpatient psychiatrist has found themselves talking with an insurance company reviewer - usually many states away who wants to know "where's the dangerousness?" and who is quick to deny payment for the treatment of a severely disabled person on that basis.  Dangerousness is probably the single word in the English language responsible for shutting down psychiatric care bed capacity and driving skilled psychiatrists away from the treatment of severe psychiatric disorders.

Rosenbaum makes the mistake of overemphasizing the importance of stigma.  She accepts as a given that stigma causes "countless harms" when the real harm is caused by systemic discrimination by the government and insurance companies.  The dangerousness concept as a justification for no treatment has done far more harm than the stigma of mental illness.  She also connects the attempt to counter stigma with illness minimization behaviors such as treating a person with a severe mental illness and impaired decision-making capacity as a competent decision maker.  Stigma is of course tied in with the recovery movement and policy that flows from the Substance Abuse and Mental Health Services Administration (SAMHSA).  Overmedicalization and disregard for civil liberties is cited as a need for this recovery based model.  That cannot be reconciled with the fact that for at least 40 years, all states have had civil commitment and guardianship/conservatorship laws in place that are in place to assure the civil liberties of anyone where there is a suggestion of coerced treatment.  When people "die with their rights on" they are generally dismissed by courts and end up dead as the result of untreated mental illness.  Only psychiatrists are generally bound by these laws and not other physicians and that part is left out.

The other significant group left out of course are the antipsychiatrists and their considerable allies whose raison d'etre is basically to bash psychiatrists and make them look bad.  That group frequently uses the term overmedicalization and pretends that there has not been decades of legislation to protect the civil rights of the mentally ill.  She is patronising in suggesting that "most psychiatrists" recognize "peer support, structured activities, psychotherapy, employment assistance, and case management" might be useful -  ignoring essentially a century of research by psychiatrists on these elements of treatment.  It also ignores the truth that every psychiatrist knows and that is - people recover and people with severe illnesses recover.  I don't need a recovery movement to tell me that.  I have people walking up to me on the street who I treated decades ago who tell me that they are married, they have children, and they are working.  Despite that severe shortcoming, I give her credit for pointing out how recovery proponents "twist data to advance their agenda".  There is currently a lot of twisted data out there.

She ends the first discussion with commentary on whether a living will for treatment of psychiatric disorders, innovations like a "one-time autonomy violation for forced treatment", or civil commitment as usual is the best approach to forced treatment of severe mental illness.  That assumes that some innovation is necessary and that it will be universally applied.  Both assumptions are incorrect.  It is possible these days to look at the editorial pages of any major medical journal and of course the non-medical media and read endless suggestions about how "the system" of medical care should be changed.  The reality is that there is a loose system of medical care and for psychiatric care the system is non existent except in the minds of antipsychiatrists.  As far as I can tell the only thing wrong with mental health law is that it is subjectively interpreted and that interpretation is clearly affected by the attitudes of attorneys and judges and the resources of the county that is financially responsible for treatment.  That skews undertreatment more severely to the counties with fewer resources.  There is nothing wrong with the law as it is written.  Civil rights are protected and the decision about rights is made by the legal system and not psychiatrists or physicians.

The second article in this series (2) is focused on the mortality gap between persons with mental illness and those without.  Unlike recent authors Rosenbaum points out that the problem was first identified by a psychiatrist in 1932, long before there were any effective treatments or treatments that allegedly cause medical complications.  She almost avoids the accusation of some recent extremists that psychiatrists and psychiatric treatment are responsible for this increased mortality rate but does bring that point up.  Nobody can argue that a lot of this mortality is based on pessimism in dealing with the mentally ill.  I can still recall the obsessing about the decision go to smoke-free inpatient psychiatric units.  One of the arguments is that it would lead to much more aggression among patients and toward staff.  The  other argument was that smoking was "all they had" and it should not be taken away from "them."  Both arguments were advanced by trained mental health staff and are extremely demeaning.  It is not surprising that patients with mental illness have lower rates of guideline recommended care or disease altering interventions.  I have contacted many primary care physicians who were surprised to hear that their patient was still alive.  I have also found myself in the position of treating people with sustained very high blood pressures or people who were actively bleeding because the patient was refusing treatment.  The only advice I got from consultants was to call them back when something bad happened and in some cases it did.

The medical rather than psychiatric assessment of capacity to consent is discussed.  A study is presented that shows  that on a typical medical service 40% of patients lacked capacity to consent and the treatment team was aware of only 25% of those cases.  This implicit consent on medical and surgical services is widespread.  It is generally questioned only in the case of high risk procedures or if a patient illogically refuses routine care.  In hospitals with psychiatric consultation-liaison services - they are typically consulted with the question.   Rosenbaum suggests that in contested situations forced treatment results in brutal seclusion, restraint and forced medication.  That is an unlikely outcome in these situations.  Most people no matter how delusional agree to care if a judge orders it or they realize that just leaving is not an option.  The suggestion is made that more time and training for capacity to consent evaluations be considered.  That is not going to happen.  There is a reason that people avoid going into psychiatry.  Talking to patients and making these assessments is one of them.  And there is no way that hospitalists are ever going to have the time it takes to do this job in addition to their medical evaluations.  It is just another psychiatric resource rationed out of the system.

Rosenbaum concludes with a valentine to integrated care models.  She has access to some of the few internists who are adept are treating people with severe mental illness, including one internist who specializes in it.  The recent MGH Case Reports have discussed similar models.  The main model that was invented in 1974 by Len Stein, MD, Mary Anne Test, MSW and  others is never mentioned.  Since that time there are ACT teams that coordinate medical care for people with severe mental illnesses and help them stay healthy outside of hospitals.  As a psychiatrist trained in these community psychiatry interventions there was nothing about patient care in these three articles that was news to me.  There are plenty of psychiatrists trained in these interventions but very few settings to implement them.  That is because the rationed care approach has split psychiatric care off from managed care organizations and placed it under the purview of the state - usually as a way to justify shutting down state hospital bed capacity.   In states where the resource becomes state-run managed care poor outcomes can be expected to follow.  Since ACT treatment was originally focused on deinstitutionalization and quality of life it is easy to dismiss based on typical cost effectiveness arguments.  I can still remember Len Stein showing us the reason why ACT was invented.  It was a single slide showing a gymnasium sized room full of cots that were placed edge to edge.  The patients were all men wearing the same uniforms off to one side.  His question to the residents was whether it was better to help all of those men live independently or let them live in those state hospital conditions?  Unlike today - incarceration and homelessness were not acceptable options.      

The final article in the series suggests that inaction on the part of physicians in addressing the problem of medical illness in people with severe mental illness is a result of learned helplessness.  That refers to a well known animal model of depression where test animals are subjected to some insurmountable stressor and generally give up and stop trying.  That model alone provides an interesting disconnect between psychiatrists and the rest of medicine.  Since the days of German asylums, psychiatrists have wanted to talk to their severely ill patients - even before there were any effective treatments.  That drive is one of the most  compelling reasons for medical students to go into psychiatry in the first place.  I don't think that properly trained psychiatrists view these problems as hopeless situations, even though some of the inpatient units they work on are clearly repositories for problems that other physicians consider hopeless.

I have no problem at all talking with anyone who has a severe psychiatric problem, understanding them, and discussing treatment options for both their psychiatric and medical problems.  I have no problem understanding that the only reason I am the one offering help is a matter of genetics and good luck.

Any psychiatrist that I know can do that.      

George Dawson, MD, DFAPA


1: Rosenbaum L. Liberty versus Need - Our Struggle to Care for People with Serious Mental Illness. N Engl J Med. 2016 Oct 13;375(15):1490-1495. PubMed PMID: 27732817.

2:  Rosenbaum L.  Closing the Mortality Gap - Mental Illness and Medical Care.  N Engl J Med. 2016 Oct 20; 375 (16): 1585- 1589.

3:  Rosenbaum L. Unlearning Our Helplessness - Coexisting Serious Mental and Medical Illness. N Engl J Med. 2016 Oct 27;375(17):1690-1694. PubMed PMID:27783917.

Monday, October 24, 2016

Stigma and Addiction

The basic position that I take on this blog is that stigma is an overblown concept.  Certainly no professional should ever be in the position of treating a person with a mental illness or addiction in any way that conforms to stereotypes.  I have been in many situations where that occurred during my training.  In those days a lot of alcoholics were admitted to medicine services because they needed detoxification by people who knew what they were doing .  They also needed close monitoring by nursing staff.  That did not mean that they were treated like all of the other medical patients.  There was usually a sense of hopelessness on the part of the house staff who could see several of these men admitted repeatedly during a 3 month rotation.  Men with hepatic encephalopathy, recurrent pancreatitis, alcoholic hepatitis, and upper GI bleeding from varices.  During one of the rotations, I encountered the term "incorrigible alcoholic" right there in the PGY3 note countersigning my intern note.  I had never seen a term in a medical chart like that before.  I had to look it up to make sure I knew what it meant and sure enough the first definition was bad beyond reform.

These reactions extended far beyond alcohol use problems.  Young addicts using various street and prescriptions drugs would present confused and aggressive.  At times paranoia and aggressive behavior were also prominent problems.  Nursing staff and house staff were frequently injured and in these emotionally charged encounters, the word "dirtball" was frequently uttered.  It was clear that at least some professionals viewed the confusion or aggressive behavior as volitional on the part of the patient and classified them as people who were intentionally trying to injure the staff.  The only way that you can make it in psychiatry is if you realize that aggressive behavior is a component of the illness.  It needs to be contained, but it does not need to be seen as a conscious "choice" of the patient involved.  Neither does their hygiene, cognitive problems, general lack of self care, inability to follow through with discharge instructions or stay away from drugs or alcohol.  That is not "excusing" them because you don't think they have a legitimate illness or can't prove that their behavior  is biologically based.  It is treating them like a human being and recognizing that you might be bringing too much emotion into the equation yourself.  There is nobody who needs a doctor with a cool head more than an addict or an alcoholic.

Those experiences led me to pay close attention to an opinion piece in JAMA about stigma and addiction.  One of the authors was from the White House Office of National Dug Control Policy.  The other was from Harvard T.H. Chan School of Public Health.  I looked even closer when it became apparent that their arguments were focused on the stigma arguments that were used for mental illness.  The authors use mental illness and the early days of the HIV epidemic as examples of how the language used to describe these patients implied moral deficiencies and led to discrimination.  They go on to cite studies of how differences in words can affect how treatment decision made by professionals can be similarly biased depending  on how loaded the stigmatizing term is.  They describe how the fear of stigma keeps people out of treatment.  Finally they outline the government's approach to changing the language about addiction and how that will help.  The White House Office of National Drug Control Policy is releasing Changing the Language of Addiction for guidance on these issues.  Common examples include changing "substance abuser" or person with a drug "habit" to a "person with a substance use disorder."  Near the end of the essay they acknowledge language changes are not enough.

Their initiative will not have any impact for the same the same reason that the anti-stigma campaigns for mental illness don't have any impact and here is why:

1.  The real bias occurs at the level of the insurance industry -  Coverage for addictive disorders varies widely and the only unifying theme seems to be rationing of treatment resources.  That rationing has been going on for 30 years and has led to inadequate treatment capacity.  The best time to provide treatment is right at the point that the affected person needs help.  Setting them up for an appointment 2 - 4 weeks later does not make any sense and can be dangerous if they are using dangerous levels of addictive compounds.  It makes absolutely no sense at all to deny care to a person who is using dangerous levels of addictive drugs simply because they have not yet tried outpatient treatment.

2.  Clinicians don't resist evidence based treatment  - there is nobody around to deliver it -  It is well known that psychiatric and addiction services are understaffed and have been for decades and the situation will probably get worse.  The number of addiction psychiatrists and addictionologists is even lower on population based metrics.  Policy makers seem to have the idea that primary care physicians will start actively treating addiction because treatment is currently described as being contained in a medication.  A recent study showed the underutilization of buprenorphine by these physicians.  They expressed in that same survey that they wished that they had someone who they could refer their patient to.  It is very difficult to go from prescribing opioids for a pain diagnosis to diagnosing and treating addiction in the  same setting.  It is also very difficult to provide treatment without adequate cross coverage.  There need to be adequate numbers of clinicians in any primary care clinic who are interested and competent to treat addictions.  In the case of buprenorphine maintenance. they need to be licensed to prescribe it.  Even then they need referrals sources to physicians who specialize in treating addictions and have some access to more resources.

3.  Community factors are prominent -  Insurance companies still discriminate against anyone with a substance use disorder.  I had a recent conversation with a person who needed some form of treatment. but was concerned about what would happen once the medical records got out to a long term insurance carrier.  Previous experience suggests that company takes 5 years to reconsider any application from a person with an alcohol or drug use disorder.  He  declined any form of treatment that would become part of the medical record that could be accessed by the insurance carrier.

4.  The Mental Health Parity and Addiction Equity Act of 2008 is a bust - time to stop pretending that it means much -    This is the highly acclaimed parity act started by Senators Paul Wellstone and Pete Domenici.  Discrimination and unfair treatment are widespread and contrary to what was expected there has been no boom in treatment for addictions.  Addiction and mental illnesses are still subjected to the same rationing policies and lack of infrastructure as they were before this act.

5.  It all starts and ends with the government -  This essay has that familiar ring to it:  "We are from the government and we are here to help you."  Let's not forget who started the system of discrimination against people with mental illness and addiction in the first place - the government at all levels.  The government invented the managed care industry as its surrogate in the first place.  If they were really interested in solving the problem - they could use the same top down approach that they used to create it in the first place.  They could provide medical detoxification in hospitals and coordinate the development of those guidelines.  They could provide access to Addiction Psychiatrists and  Addiction medicine practitioners.  They could open up bed capacity for residential and sober house care.  They could fund clinics where medication assisted treatments for opioid use and alcohol use are conducted.  They could fund addiction centers of excellence.  They could fund research on treatment for court ordered offenders and whether it is effective.  This is all evidence-based care, but the article suggests that primary care physicians who are currently overworked by government mismanagement are going to suddenly see hundreds or thousands of new patients with addictions.  Suggesting that this is a language based problem put the blames directly on clinicians.  It is clear to me that there are no psychiatrists blaming people for mental illness or addiction.  Who are these people and how extensive is the problem?  The idea that everyone needs to be reformed or reeducated is a familiar tactic used by politicians and policy makers.  It was the rationale for managed care rationing in the first place.

6.  Prevention is a priority - The prevention of drug use is the surest way to prevent increasing number of people from experiencing morbidity and mortality due to drug and alcohol use.  Prevention of drug use at this point in time is historically difficult as the country swings into another era of permissive attitudes toward drug use.  Individuals not abusing their first opioids will have a much greater impact on the prevalence of addiction than all of treatments after an addiction has started.

All of these factors are what clinicians like me see as everyday interference with helping patients who have a substance use disorder.  Semantics may help some.  Training and recruiting physicians who know that it is only luck that separates them from people with addictions and mental illnesses will help more.  Ending insurance company dominance over clinicians will help the most.

In the end - words don't keep people with addictions from lifesaving treatment.

The government and health insurance companies do.             


George Dawson, MD, DFAPA


1: Botticelli MP, Koh HK. Changing the Language of Addiction. JAMA. 2016 Oct 4;316(13):1361-1362. doi: 10.1001/jama.2016.11874. PubMed PMID: 27701667. (free full text).

Sunday, October 23, 2016

The Largest Psychiatric Hospitals in the USA

In about 2012, I read an article that described the largest psychiatric hospitals in the USA as

1.  LA County Jail
2.  Riker's Island
3.  Cook County Jail

In the past month there is good evidence that in Minnesota, one of the most liberal states in the country - the situation is no different.  A recent study estimated that 25-30% of the jail population had a mental illness and 11% were on maintenance antipsychotic medications.  The  Hennepin County Sheriff Rich Stanek is quoted:

“What we’re seeing is crisis levels of mental illness among our inmates. This is solid evidence that our jails continue to serve as the largest mental health facilities in the state.” (1)

That same article refers to a Legislative Auditor's report describing the problem as widespread throughout Minnesota.  There is an alarming statistic that since the year 2000 there have been 770 suicide attempts and 50 suicides in Minnesota jails.  The article does outline some helpful measures.  Inmates will be screened by psychologists and psychiatric nurses.  The screening by psychologists is focused on low level offenders who can possible be released earlier,  although the offenses of the mentally ill were not substantially different from the non-mentally ill population.   The mentally ill inmates had a higher recidivism rate and were 30% more  likely to have 10 or more bookings at the jail. In a separate opinion piece (2), Sheriff Stanek reports that jail personnel are all going through 32 hours of crisis intervention training (CIT) to learn about the specific problems that mental illness causes and how to interact with people experiencing those symptoms.  The Barbara Schneider Foundation - a non-profit organization dedicated to end the criminalization of the mentally ill through positive training and education -  provides the training.  In his opinion piece the following quote is instructive:

"The urgent need for this training is a direct consequence of federal action requiring states to close our state psychiatric hospitals with no immediately viable community alternative. Our county jails should never have become the largest mental health facilities in the country."

I have previously written many times about the abysmal system of care that is available for people with severe mental illnesses.  In my experience, the people rationed out of the system include many of those who end up homeless and in jail.  If you have severe problems with mental illness that affect your decision-making, your social behavior, and your ability to assess the impact of your decisions you are at much higher risk  of an adverse interaction with the police or incarceration.  If incarcerated you are less like to have the resources to make bail, obtain and cooperate with an attorney, or follow the conditions of release. 

 These impairments combined with severely rationed resources accounts for the explosion of mentally ill in jails and this is not an acute problem.  It has been progressively worse over the past 20 years.  Stanek also call on the Governor and the legislators to provide finding for adequate placements (and hopefully supervision) and funding for CIT.  So far legislators have passed a law that allows Sheriffs to transfer mentally ill offenders to the limited beds at Anoka Metro Regional Treatment Center (AMRTC).  That has resulted in an increasing backlog of admissions of committed patients and increasing violence at the hospital.  Nobody in any of this controversy has spoken to the needs of the antisocial or career criminal with severe mental illness.  The issue of addiction and how that creates mental illness, criminal offenses, and leads to recidivism and worsening mental illness is also not addressed. 

In these articles and most, the families are left out of the equation.  The families I have seen are typically parents who have been dealing with the severe mental illness of their children for years.  They are shocked to find out that their children have been incarcerated instead of being hospitalized.  They are shocked that their children are not receiving any care for their mental illness while incarcerated especially that their medications have been acutely discontinued.  I have talked with many of these patients who were on methadone or buprenorphine for opioid addiction who had these medications acutely discontinued and went into opioid withdrawal until that resolved.  Psychiatrists everywhere have heard the pleas of these parents and their request to assist them in getting their child out of jail.  There is generally noting that can be done.     

 An associated issue is how government systems are managed at all levels.  In recent times, the idea that government systems can be managed like for-profit businesses that much show a profit for shareholders is all the rage.  It was one of the reasons that scientific and humane treatment of mental illness could be rationed out of existence.  Suddenly there was no longer a concern that a patient with mental illness was disruptive and might end up in jail or homeless if that behavior continued.  Now they had to be "dangerous" or the hospital asked them to leave.  When there were still too many demands on rationed beds they had to be "imminently dangerous" or they were asked to leave.  Sheriff Stanek and sheriffs across the country are dealing with the fallout of this managerial practice.  If people can't be treated in a cost effective manner (defined as getting them out in a defined number of days that are paid for)  they are not admitted.  Many of these patients are brought to hospitals by the police and not admitted because the hospital knows they will be taken to jail.       

The other problem of course is that jails are not really hospitals.  In today's political climate - even saying that out loud can set a dangerous precedent.  Even though Sheriff Stanek is doing what he can and he has a mandate by the Minnesota legislature allowing him to hospitalize patients on a priority basis at a state hospital, minimal to non-existent standards of care in jails do not make them hospitals.  The clear evidence from the editorial is the sparse medical coverage (1 RN very 12 hours for over a hundred mentally ill inmates), lack of adequate medical training (provided in this case by a not-for-profit foundation), and a lack of discharge resources for continuity of care.    

There are no psychiatric hospitals that can function or legally operate with that level of care.

George Dawson, MD, DFAPA


1.  Chris Serres.  Mental illness in Hennepin County jail far higher than previous estimates, new study finds.  Star Tribune September 22, 2016.

2.  Rich Stanek.  Commentary:  Addressing the mental health crisis in our jails.  Star Tribune October 14, 2016.


1.  The photo is Hennepin County Jail from Wikimedia Commons.  The source information is by Micah (Transferred from en.wikipedia by SreeBot) [Public domain], via Wikimedia Commons.  The page URL is:

Saturday, October 22, 2016

Coffee Shop Neuroscience

I went into my favorite coffee shop the other day for my usual mocha.  They typically post a trivia question of the day that gives you a 10 cent discount on the coffee.  After a conference with a recent focus on neuroscience it was interesting to see a question about the number of thoughts per day.  My wife ventured a guess.  I asked the barista for the source and all that she could tell me was: "We get it off the Internet like most of our questions."  No footnote or reference available.  The source was not difficult to find.  It was a typical Internet site that has never impressed me as a knowledge source, but it did have a link to the original paper.  It turns out to be a neuroscience site - The Laboratory of Neuroimaging (LONI) at UCLA.  Read the fine print at the bottom of this page for qualifiers on what counts as a thought.  I looked for any papers on this estimate on the web site as well as Medline and did not come up with anything.

Irrespective of the methodology the question poses interesting questions for clinical psychiatrists if they are comfortable outside the confines of the DSM.  How much attention is being paid the the baseline conscious state of the patient and why might that be important?  What is their stream of consciousness every day?  Is it disrupted by mental illness or addiction?  To what extent is that stream of consciousness broken up into daydreams, memories, and fantasies?  To what extent is it impacted by a process that is not even suggested by the DSM?  The best example that I can think of is boredom.  Being easily bored can be a diagnostic criterion, but it seems to be an uneasy mental state on its own.  People who are bored get driven to do things to alleviate boredom and sometimes those activities are very risky.  Are the thoughts mentioned in the coffee shop question memories, fantasies, or daydreams? Why the large number?  I am not aware of brief frequent thoughts.  My stream of conscious thought is comprised of more coherent stories or images.

I did a Grand Rounds on fantasy and daydreaming about 15 years ago.  There has never been much quantitative work on fantasy.  There were some new research approaches to daydreaming being used at the time and I incorporated some of those references into the presentation.  I don't recall the exact number of daydreams per day but they were considerably less than 100.  The only research approximating the numbers of thoughts per day may be the research on the exact number of spoken words per day.  This research has generally been a comparison between men and women and a test of which sex speaks the most words.  Those numbers across different cultures and sampling periods range from 12,867 to 24,051 words per day (5).  Standard deviation were large and the authors conclude that on the average both sexes spoke about 16,000 words per day.  To me speech and language is unconsciously processed thought, but even counting all of those words does not get us to the level of the coffee shop question.  Do the authors believe that they have a way to capture tens of thousands of unconsciously processed thoughts?  I am very interested in hearing how they arrive at these figures.

The research in this area has since moved into the area of the wandering mind.  Wandering mind is defined as a cognitive focus on information unrelated to the immediate sensory input or task on hand.   It would include daydreams, fantasies, and the typical stream of consciousness that every person experiences at times throughout the day.  The critical research questions include when is mindwandering adaptive as is the case of generative fantasies and when it is maladaptive?  Smallwood, et al (7) have written an excellent brief review of how mindwandering can negatively impact medical decision making and the cognitive performance of physicians.  They point out that fatigue, depression, and circadian rhythm disturbances can lead to increased mindwadering with negative consequences by decoupling attention to the external environment from the necessary memories, patterns and access to decision-making that are the physician's cognitive set.      
The question also involves neuroscience.  Is there a physical representation of this process in the brain and what is it?  Neuroscience tells us that the brain has a Default Mode Network (DMN).  It was initially noted to be a network of brain regions that remain active during functional brain imaging studies in the absence of an external task.  These studies typically involve an active task for the research subject and the resulting brain image is analyzed as a response to that stimulus.  It was determined that this DMN comprises the system that allows for internal dialogue and stimulus-independent thought.  The physical representation includes a primary system comprised of the anterior medial prefrontal cortex (aMPFC) and the medial posterior cingulate cortex (PCC) communicating with two subsystems.  The medial temporal lobe subsystem is comprised by the retrosplenial cortex (RSP), parahippocampal cortex (PHC), hippocampal formation (HF), the ventral medial prefrontal cortex (vmPFC), and the posterior inferior parietal lobule (pIPL).  The dorsal medical prefrontal cortex subsystem is comprised of the temporal pole (TempP), the lateral temporal cortex (LTC), the temporoparietal junction (TPJ), and the dorsal medial prefrontal cortex) (1).  Some groups differ on the physical representation of this system and some groups use Brodmann area designations.  I drew a slightly different model based on Sporn's text (6) with some obvious distortion due to the lack of a three dimensional representation (the rTC should be folded over to the right temporal area away for the medial view):

Subsequent research has shown that the DMN may be implicated in several psychiatric disorders (2). Several functions for the DMN have been proposed that cut across a number of disorders including mind-wandering when no specific external stimulus is present, memory consolidation, to possibly maintain a baseline level of arousal, to divide attention across tasks and for continuity across time (3).  Disruptions and functional disconnects to the DMN have been studied for a number of psychiatric disorders.  If the conceptualization is correct it is useful to think about the implications of functional or anatomic disconnects or hyperconnects to other systems.

How is all of this relevant to psychiatry?  Consider the case of two patients with severe depression.  They both have insomnia, anhedonia and decreased appetite.  They both have typical depressogenic thought patterns including abundant self criticism, hopelessness, and suicidal thoughts without intent to harm themselves.  The only difference is that one of these patients has intense rumination about a job loss.  This patient was downsized along with 50 other people.  The job loss was a straight business decision rather than any performance deficiency.  The patient without rumination is treated with standard methods and recovers.  The patient with intense rumination does not.  The depression and rumination persists despite multiple antidepressant trials.  The degree of disability persists.  There is not much guidance about how to treat this person from a biological standpoint.  It comes down to empirical drug trials and additional treatment for what has been considered anxiety, psychosis, or possible obsessive compulsive disorder.  I have seen these patients recover with detoxification from drugs or alcohol, treatment with antipsychotics, treatment with electroconvulsive therapy, but not treatment with benzodiazepines.  Will the cognitive neuroscience that incorporates models of the wandering mind and default mode network offer fast and more effective treatments?  I think that we may already be seeing that.  Mayberg's classic 2005 article (8) explicitly testing the network hypotheses about treatment resistant depression with deep brain stimulation was a start.  That literature has greatly expanded since that point.

There has been an explosion of network based theories of both psychopathology and normal conscious thought in the past decade.  These models are increasingly relevant as psychiatry is dragged out of a receptor and reuptake protein based discipline, where the practitioners may have a vague idea of where those receptors are located and what they really might be doing.  It was a necessary second step, but only neuroscience will get us to better models and models that we can apply to the treatment of unique individuals.  Psychiatrists have a critical decision to make at this point.  Are we going to remain stuck in a diagnostic and treatment paradigm that clearly applies to a minority of the people with mental illness or are we going to embrace the science that will both allow us to treat everyone better and give us a more complete understanding of human consciousness?

Learning about the Default Mode Network is a good starting point.  

George Dawson, MD, DFAPA


1;  Barron, DS, Yarnell S.  Default Mode Network: the basics for psychiatrists.  Fundamentals of neuroscience in psychiatry.  National Neuroscience Curriculum Initiative.

2:  Mohan A, Roberto AJ, Mohan A, Lorenzo A, Jones K, Carney MJ, Liogier-Weyback
L, Hwang S, Lapidus KA. The Significance of the Default Mode Network (DMN) in
Neurological and Neuropsychiatric Disorders: A Review. Yale J Biol Med. 2016 Mar 
24;89(1):49-57. eCollection 2016 Mar. Review. PubMed PMID: 27505016; PubMed
Central PMCID: PMC4797836

3: Mason MF, Norton MI, Van Horn JD, Wegner DM, Grafton ST, Macrae CN. Wandering minds: the default network and stimulus-independent thought. Science. 2007 Jan 19;315(5810):393-5. PubMed PMID: 17234951; PubMed Central PMCID: PMC1821121.

4: Stafford JM, Jarrett BR, Miranda-Dominguez O, Mills BD, Cain N, Mihalas S,Lahvis GP, Lattal KM, Mitchell SH, David SV, Fryer JD, Nigg JT, Fair DA. Large-scale topology and the default mode network in the mouse connectome. Proc Natl Acad Sci U S A. 2014 Dec 30;111(52):18745-50. doi: 10.1073/pnas.1404346111. Epub 2014 Dec 15. PubMed PMID: 25512496

5: Mehl MR, Vazire S, Ramírez-Esparza N, Slatcher RB, Pennebaker JW. Are women really more talkative than men? Science. 2007 Jul 6;317(5834):82. PubMed PMID:17615349.

6: Olaf Sporns.  Networks of the Brain.  MIT Press.  Cambridge, Massachusetts, 2011.

7: Smallwood J, Mrazek MD, Schooler JW. Medicine for the wandering mind: mindwandering in medical practice. Med Educ. 2011 Nov;45(11):1072-80. doi: 10.1111/j.1365-2923.2011.04074.x. PubMed PMID: 21988623. (link to free full text).

8: Mayberg HS, Lozano AM, Voon V, McNeely HE, Seminowicz D, Hamani C, Schwalb JM,Kennedy SH. Deep brain stimulation for treatment-resistant depression. Neuron. 2005 Mar 3;45(5):651-60. PubMed PMID: 15748841.

9: Christoff K, Irving ZC, Fox KC, Spreng RN, Andrews-Hanna JR. Mind-wandering asspontaneous thought: a dynamic framework. Nat Rev Neurosci. 2016 Nov;17(11):718-731. doi: 10.1038/nrn.2016.113. PubMed PMID: 27654862

Supplementary 1:

Olaf Sporn's book Networks of the Brain is an excellent resource to study this topic and to try to catch up on a decade of research:

Supplementary 2:

Default Mode Network links that are relevant for psychiatrists (unedited):

"default mode network" dementia

"default mode network" addiction

"default mode network" "bipolar disorder"

"default mode network" schizophrenia"

"default mode network" ADHD 

"default mode network" depression

"default mode network" anxiety

"default mode network" mind-wandering

"default mode network" day dreaming

Supplementary 3:

The answer to the coffee shop question is a.

Sunday, October 16, 2016

The Balanced Rhetoric Against Neuroscience

The New York Times editorial pages continue to be a place where anti-neuroscience rhetoric can be expressed primarily as decreased funding or more accurately portion of the available NIMH funding.  Maybe there has been some pro-neuroscience opinion expressed there and if there was I have missed it.  I recently posted an exciting development in neuroscience teaching for psychiatrists and psychiatric residents.  In that post I reference an opinion piece by Richard Friedman, MD a psychiatrist (1).  Dr. Friedman makes several arguments for psychotherapy as if it is unrelated to neuroscience and based on that premise concludes that there is no substitute for psychotherapy, that people are more than a brain in a jar, and that anyone benefiting from psychotherapy seems to prove  that.  I found that to be an incredible statement considering that (according to Koch in above graphic):  "The brain is the single most complex object in the universe." There is also the fact that with 7.4 billion people on earth - there are 7.4 billion unique conscious states - the vast majority of which are not accurately described by any DSM or psychodynamic diagnosis/formulation.  All the time that Dr. Friedman is mounting this critique he also discusses the importance of clinical research and suggests shifting the funding balance away from neuroscience.

In the recent case John C. Markowitz a professor of clinical psychiatry at Columbia has a more subtle form of the argument.  In this case and the previous opinion piece the authors both endorse the importance of neuroscience to a point.  In this case the argument is - yes neuroscience is important but let's reestablish balance between neuroscience and clinical studies such as looking at the efficacy of psychotherapies.  Breaking it down, Dr. Markowitz makes the following points:

 1.  Under the directorship of Thomas Insel, the NIMH clinical research budget was "strangled" and the resources were diverted to neuroscience research.  The author acknowledges both the need for neuroscience research and the primitive stage of psychiatric diagnostics based on clusters of signs and symptoms.  This was really the basis for Insel's RDoC initiative looking at more reliable markers of psychiatric syndromes.  Any practicing psychiatrist who has seen all of the iterations of the DSM realizes that we are as far as we can go with this manual.  That includes from the standpoint of validity but also in terms of the clinical examination by psychiatrists.  As long as we are all contained by this manual, the clinical method of psychiatry will remain stuck somewhere in the 1940s.  That should be extremely disconcerting to the profession and future psychiatrists.

DSM technology is extremely limiting in terms of the usual clinical trials.  The NIMH sponsored Star*D study is a decade and a half old at this point.  It has defined the response rates for both antidepressant therapies and provided a discussion point for psychotherapy trials of depression.  Clinical trials of antidepressants provide an equally varied result.  Any practicing clinician knows that these studies are all seriously flawed out of the gate by using DSM diagnoses and also an intent-to-treat analysis that does not resemble clinical practice.  The variation in diagnoses from depression to anxiety to depression plus anxiety as seen in clinical practice should point to the fact, that patient selection into clinical trials currently results in very heterogenous patient populations in terms of both therapeutic effects and medication tolerability.  We can continue to spend large sums of money on these trials of mixtures of patient populations and post modest positive results or we can attempt to identify patients who will respond specifically and not experience side effects from a particular therapy.  That is the real promise of neuroscience based research.

2.  The patients who need help are poorly served by current neuroscience research.  The helpful psychotherapies listed by the author like interpersonal psychotherapy (IPT), cognitive behavioral therapy (CBT), and other psychotherapies have been around for decades.  I happen to have copies of Interpersonal Psychotherapy by Klerman, Weissman, Rounsaville, and Chevron and Cognitive Therapy of Depression by Beck, Rush, Shaw, and Emery.  The publication date of the former is 1984 and the latter is 1979.  Both therapies have been out there for over 30 years.  At this point both have been studied hundreds of times.   Looking at clinical trials on Medline yields 1711 for CBT and 261 for IPT.  Not only that but some of the clinical trials that were successful (like IPT for cocaine use) have never made  it into clinical practice.  In fact, in most places getting a therapist who actually practices any of the specific research proven psychotherapies is impossible.  The problem does not seem to be a lack of psychotherapy research but a lack of access to practitioners who use research proven psychotherapies.  Mental health treatment is the most highly rationed treatment resource and additional studies that continue to prove that existing psychotherapies work seems superfluous at this point.  Any current studies are often compared to existing therapies and with the DSM problem contributing to diagnostic heterogeneity.  Any new trials should only be funded for serious conditions where the therapy might be useful.  There is no reason to expect that a new therapy applied using the current diagnostic system or clinical trials technology will lead to any enhanced treatment effects.

3.  Existing treatments are not "good enough".  The author attributes this "good enough" statement to Insel himself.  I understand the point he is trying to make.  The author points to continued suffering, treatment failures and suicides as evidence that more is needed now.  The problem is that there is no assurance that clinical research will add any more at this time.  Certainly a focus on suicide as a stand alone problem (not suggested at all by DSM) and on serious disorders with no treatment like adult anorexia nervosa is warranted.  But even then we are left with a clinical trials technology that consistently produces modest results at best.  More multimillion dollar trials of psychotherapy that we already know is somewhat effective when patients have no chance of ever receiving it against a backdrop of "is this really depression or anxiety" seems like a waste of time and money to me.  It seems like a much better idea to develop a neuroscience method to determine who needs psychotherapy and who might benefit from medications.  But even then, the only treatments that will be readily available will be the medications and even then less than half of the affected patients will get access to treatment.  Good luck trying to find a psychotherapist and an insurance company willing to cover the cost of the number of sessions used in the psychotherapy research. Research proven therapies are only as good as the number of practitioners using them and access to those practitioners.

4.  The placing all of your eggs in one basket argument.  This is basically saying that if the ratio of clinical to neuroscience funding is 10% to 90% the risk is missing something big in the clinical research and not getting any useful results from neuroscience.  Given the history that I have provided, there needs to be a clear advance on the clinical side in order to fund large trials.  It does not make any sense to continue to  fund more of the same  or slight modifications of treatment for common disorders.  Our eggs have been all in one basket and I would call that treatment as usual.  In the 30 years that I have been in practice, there is nothing that I would call a major breakthrough.  Clinical research results come and go.  Effective psychiatrists are effective psychiatrists not based on breakthroughs but how they approach clinical practice.  Even that mode of treatment is threatened by widespread support for "collaborative care" that is being justified using the same kind of research that justified managed care in the first place.  In the end there has been nothing more destructive in terms of access to care for mental disorders than managed care.

In many ways these ongoing arguments resemble the arguments of the biological psychiatrists and psychotherapy psychiatrists that I trained under in the 1980s.  Many programs were split under this artificial division with the residents left to identify with biological or psychotherapy faculty.  It is interesting to note that this division occurred at a time when Kandel wrote a paper on how psychotherapy is neuroscience in action (3).  That may have been missed because the biologically based psychiatrists at the time were really focused on pharmacology and neuroendocrinology rather than a comprehensive neuroscience.  Neuroscience and the old diagnostic technology and clinical methods seem to be the current points of division.

A lot of the criticism is directed at Insel.  I have heard him talk about the initiatives and the rationale sounded clear to me.  I think that rationale is very similar to what I have discussed so far, but for clinical psychiatrists it is also the realization that as long as we live in an approximate world - we will get approximate results.  The inertia to stay in that place is always puzzling to me.

But - it is time to move out of the 1950s.

Clinical psychiatry the way it is currently researched and practiced holds no promise for understanding the most complex known object in the universe.  Neuroscience is one of the big ways out of that predicament.

George Dawson, MD, DFAPA      


1:  Friedman RA. Psychiatry's Identity Crisis. New York Times July 17, 2015. p SR5.

2:  Markowitz JC.  There’s Such a Thing as Too Much Neuroscience.  New York Times October 14, 2016. p A21.

3:  Kandel ER. Psychotherapy and the single synapse. The impact of psychiatric thought on neurobiologic research. N Engl J Med. 1979 Nov 8;301(19):1028-37. PubMed PMID: 40128.


Friday, October 14, 2016

National Neuroscience Curriculum Initiative - a brighter future for psychiatry

As any reader of this blog probably knows, I am a big proponent of neuroscience education for psychiatrists and always have been.  I have suggested in the past that it would take broad collaboration.  I posted some examples of an NIMH initiative on neuroscience  education.  I teach neuroscience (also known as neurobiology) myself and that has led me to be acutely aware of the lack of educational resources on the field.  That background is what has made this the happiest day at a CME conference that I have ever spent.  I am currently at the University of Wisconsin 4th Annual Update and Advances In Psychiatry - a conference that has really been in place for the past 41 years.  After watching a comprehensive update on eating disorders I settled in to listen to Melissa Arbuckle, MD, PhD; Professor of Clinical Psychiatry, Director of Residency Training; Department of Psychiatry, Columbia University Medical Center.  The title of her presentation was Discussing the Neuroscience of Mental Illness with Your Patients.  I turned to the section in the syllabus and there was one page with a case report on the front and a crude drawing of the brain (two views) on the back.  Being a traditional conference guy who likes a ton of technical information, no audience participation, and no role playing - I was prepared to be disappointed.

I was not prepared for what would transpire in the next 90 minutes.  I have posted here many times why I thought that every psychiatrist should know neuroscience and ways to do it, specifically the need for widespread collaboration due to a lack pf neuroscience manpower in most departments.  Dr.  Arbuckle started out explaining what the National Neuroscience Curriculum Initiative was.  It was started by a collaboration of like minded residency directors  to come up with a program to teach neuroscience to psychiatry residents.  She showed the explosion in neuroscience papers in psychiatry just over the past decade.  She referred to an article in JAMA Psychiatry (1) with her collaborators on why neuroscience needs to be integrated into psychiatry right now.  She discussed the New York Times editorial that showed up three weeks later with the criticism of their viewpoint (2).  Although she did not mention it, like a lot of articles it as written from the perspective of a psychiatric identity crisis.  Whenever I see that term it seems like the authors are firmly behind the curve and don't seem to understand what neuroscience encompasses.  Dr. Arbuckle said that the article was critical of the criticism that the brain was the basis of human behavior.  Quoting the article:

"Indeed an article in May in one of the most respected journals in our field JAMA Psychiatry echoed this view: "The diseases we treat are diseases of the brain."........ Even if this premise were true - and many would consider it reductionist and simplistic - an undertaking as ambitious as unraveling the function of the brain would likely take many years."  The author is a psychiatrist and goes on to say that he is all for neuroscience and even talks about some recent research techniques he (implicitly) just doesn't think psychiatrists should study it?  He also seems to conflate psychotherapy as being independent of neuroscience when in fact we have known just the opposite since since Kandel's 1979 seminal lecture Psychotherapy And The Single Synapse.  

I am equally incredulous when people seem to argue about the importance of neuroscience in psychiatry.  I find reductionism and reductionist approaches to be perfectly understandable and acceptable.  It is an interesting form of rhetoric to use these terms pejoratively.  Most people go into medicine because they want to know how things work.  If they did not enter with that goal, it soon becomes apparent that knowing mechanisms whether they are theoretical or not is an important aspect of studying medicine.  Some of the first mechanisms I studied in medical school involved cholera and diphtheria toxin. How is it possible to determine these mechanisms and recent significant epigenetic mechanisms without taking a reductionist approach?

The exercise that Dr. Arbuckle introduced to the audience was the diagnosis and treatment of complex cases.  The case vignette involved a young woman with borderline personality disorder.  The task for the audience was to pair up and role play discussing the relevant neuroscience concepts of treatment with the patient using the brain diagram as an aid.  Eliciting responses from 300 people in a room slows things down.  After the audience was done, she showed a film of an expert presenting this information to a patient and what presentation materials might be available.  It went very well and it presented the rationale for dialectical behavior therapy and not a medication.  It was a clear example of a neuroscience based discussion that provides a rationale for psychotherapy.  There are numerous materials on the web site (9 modules, 56 sessions, 40 authors) and wide scale participation is encouraged.

The information up to that point was quite exciting.  Dr Arbuckle had plenty of enthusiasm in her closing remarks.  In those remarks she pointed out the goal of "getting the entire field up to speed" in neuroscience.  She pointed out that everything on the site is free but at some point they may ask physicians to pay for CME.  She said that she realizes that this is literally "changing the world and that is what we are going to do."        

This was the most exciting commentary from a psychiatrist about teaching the entire field and the future of psychiatry that I have ever heard.  I have never been this impressed by any development in the field during my career.  And I am a psychiatrist who is as pro-neuroscience as anybody.  How is it that I am just hearing about this initiative right now and only because I am attending a conference?  That is why I am posting my experience here and a link to the NNCI web site and materials.

Dr. Arbuckle and her collaborators are one of the few bright spots for the future of psychiatry.

But they are very bright.

George Dawson, MD, DFAPA


1:  Ross DA, Travis MJ, Arbuckle MR. The future of psychiatry as clinical neuroscience: why not now? JAMA Psychiatry. 2015 May;72(5):413-4. doi: 10.1001/jamapsychiatry.2014.3199. PubMed PMID: 25760896.

2: Friedman RA.  Psychiatry's Identity Crisis.  New York Times July 17, 2015. p SR5.


The graphics at the top are two slides from one of my lectures.  I like to present data on the unique aspects of every individual brain and why that can happen.  The slides are not from the NNCI program and I am not affiliated with that program.  Click on each slide to enlarge.

Wednesday, October 12, 2016

Public Service Announcement - Moderation Policy

This is just a reminder and to establish an anchor point about moderation on this blog.  If you attempted to post a response on the blog and it did not show up it is most likely due to one of these reasons.  It is always possible that I made a mistake or that your response ended up in my spam file.  In fact, several posters clued me in about that happening and I did find and post their responses.  It should be pretty clear that you do not have to agree with me but there are some limits that are fairly basic to a small barely read blog like this one.  These limits follow:

1.  No trolls - I am not fascinated by trolls or the troll culture.  One of the reasons that I started this blog was that I used to post in what I thought was a professional forum that was dominated by a troll.  There was no moderation and eventually all of the interesting posters left and went elsewhere.  So spare me the First Amendment arguments - that kind of rhetoric just doesn't fly here.  My reasonable test is what would happen to a troll in any medical staff meeting?  Consider this a similar atmosphere.

2.  No advertising - This is a strictly non-commercial enterprise.  I type it up on my own and there is no sponsor or source of revenue.  As I have pointed out to people who want to advertise here, I have many copyright permissions at this point that are all contingent upon the non-commercial status of this blog so at no point is it possible for me to accept advertising.  The Creative Commons license allows anyone to repost the contents of this blog with the proper attribution.  If you like the content - just repost it or link to it.  No advertising also includes posts with hyperlinks to product advertising and that kind of link results in a rejected post.  It also includes posts with no content and apparent compliments that are hyperlinked to advertising through the name of the poster.

3.  No anti-psychiatry rhetoric - There can be discussions of anti-psychiatry from a philosophical standpoint, but if the main point of the post is to bash psychiatry or psychiatrists this is not the place and you probably already know that.  There are many anti-psychiatry blogs that you can flock to.  Some of them are even contain active posts by psychiatrists.  Psychiatry attracts a lot of haters and my original analyses of the problem in 2012 still stands as well as a clear anti-psychiatry bias in the media.  At least part of that media bias is that they are just being provocative to draw a crowd and of course they have no responsibility to people with clear problems.

4.  No personal attacks - ad hominem is the poorest form of criticism.  If the argument cannot be addressed, attacking the person is not a substitute - at least here.  There are many other blogs and media sites where that is acceptable.  The standard I use for physicians is the expected behavior in professional settings with colleagues.  If a post does not meet that standard it is on shaky ground.

5.  No medical advice - this blog is not the place to seek personal medical advice.  Nothing here should be construed as medical advice or a discussion of an actual patient.  My opinion stated here in many places is that only your personal physician or physicians knows enough about you to make suitable recommendations.

6.  No interminable arguments - I don't have to have the last word on anything and frequently just stop posting.  I will not post repeated posts making the same argument or with very slight modifications.  One of the most interminable arguments has to do with the link between psychiatrists and pharmaceutical companies.  I have debunked that argument on this blog many times.  I would refer you to blogs where they have interminable arguments about how bad psychiatric medications are and how corrupt psychiatrists are for their connections to Big Pharma.  Neither argument is accurate but this is the place to read about how medications are actually used.

7.  Stay on topic - It is a reasonable requirement to address the topic.  Newspaper web sites are good examples of what can happen when this simple rule is not followed.  A corollary is to address the topic in a timely manner.  I have had people notice posts from 3 years ago and write a response.  Blogger does not allow me to terminate discussions like some web sites.  My only option is to not post untimely responses.  Six months is a reasonable time frame.

That is what I have so far.  I reserve the right to add more as they come to me.  The goal is to maintain coherence and rational discussion.  Any inspection of the blog posts shows that there is not a lot of discussion and stimulating discussion is not one of my primary objectives. I appreciate the people who have taken the time to read this blog and post their comments - many of which make excellent points and do stimulate further discussion.

My goal all along has been to produce opinion and analysis consistent with the way that real psychiatrists train, work, and practice.  I do not see myself as unique in any way.  I routinely have contact with excellent psychiatrists who I have trained with and who are colleagues.  They approach problems in psychiatry and think about those problems in very similar if not identical ways that I do.    I see this blog as a creative outlet as well as one of the few places on the Internet where this kind of content is available.

George Dawson, MD, DFAPA


The photo:  John Schneider.  RCA 40A Ribbon Microphone. August 19, 2007.
Attribution-NonCommercial 2.0 Generic (CC BY-NC 2.0)

Sunday, October 9, 2016

Big Data - What Is It Good For?

Big Data and Data Science have been buzzwords in science and industry for over a decade.  A Medline search shows over a thousand current references to Big Data in healthcare.  A good starting point is consider what is meant by Big Data and then discuss the implications.  A quick scan of the references shows that they vary greatly in technical complexity.  A standard definition from Google is: "extremely large data sets that may be analyzed computationally to reveal patterns, trends, and associations, especially relating to human behavior and interactions."  These techniques developed because of the widespread availability of digitized data and the ease with which sets of behavioral choices (in the form of mouse clicks) can be collected on web sites.  In many cases specific data collection paradigms can be used to elicit the information, but there is also a wealth of static data out there as well.  In health care, any electronic health record is a massive source of static data. Financial, real estate, educational records and records of all of these transactions are also a significant source.

Most  Americans logging in to set up a Social Security account online (  in the past couple of years would be surprised at what it takes to complete the job.  After the preliminary information there is a set of 5 security questions.  Four of those questions are about your detailed personal credit history - home mortgage information and credit card history.  When Social Security was initially set up in 1936 there was widespread concern that Social Security Numbers would become national identifiers.  At one point Congress had to assure the electorate that the number would not be used for that purpose.  Since then the SSN has been used for multiple identification purposes including credit reporting.  At this point it seems that we have come full circle.  Congress invented the SSN and told people it would not be an identifier.  They mandated its use as an identifier. Congress authorized and basically invented the credit reporting system in the United States.  The federal government currently uses the credit reporting system to quiz taxpayers wanting to set up a Social Security account online.  In the meantime, large amounts of financial, legal, and health care data are being collected about you under your SSN in data systems everywhere.  At this point the full amount of that data and the reasons why it is being collected for any person in the US is unknown because it is all collected without your knowledge or your consent.  It is impossible to "opt out" from this data collection.  The federal government does have an initiative to remove SSNs from health records, but there are so many other identifiers out there right now, this effort is too little and too late.

Additional sources of data include your online foot print including sites that you may have visited and what you seem to be interested in.  A visit to Amazon for example and a quick look at an expensive digital camera may result in that same camera with a link to Amazon in the margin of very other web page you see for the next two weeks.  Expensive digital cameras of a different brand than the one you originally looked at may start showing up.  You may notice product ads showing up in your Facebook feed that you mentioned casually to your friends during a conversation there.  The conversation could be as generic as bicycle seats and suddenly you are seeing a flurry of ads for bicycle seats.  Any number of web sites encourage to sign in with other accounts and then share your account information with them.  All of this data provides companies with what they need to fuel their predictive algorithms to sell you a product.  It provides the major advertisers in this space like Google and Facebook with a huge revenue source because based on the scale and personalization of these ads - they are effective.  Big Data seems to be very good for business.  But is there a downside?        

That brings me to a current resource on the nefarious uses of Big Data written by an expert in the field.  Cathy O'Neil is a PhD in mathematics.  Her PhD work was in algebraic number theory.  She started work as an academician but subsequently worked for a hedge fund, work as a data scientist for several firms and currently heads the Lede Program in Data Journalism at Columbia University.  I am familiar with her work through her blog MathBabe.  Her newly released book Weapons of Math Destruction takes a look at the dark side of Big Data specifically how data collection and biased algorithms can be good for administrators, politicians and business but bad for anyone who falls under the influence of those agencies and their work.  In the introduction she leads of with the example of teacher assessments.  I was familiar with a scattergram that she had posted on her web site showing that year to year teacher assessment scores were essentially uncorrelated or random.  In the book she describes the human toll in this case a teacher fired because of this defective algorithm.  In another example later in the book, an experienced teacher scored a 6 out of 100 on a "value-added" teacher evaluation.  Only tenure kept him from getting fired.  The scoring algorithm was opaque and nobody could tell him what had happened.  The next year he scored the 96 out of 100.  But the algorithm was so flawed he knew that score was no more legitimate than the last one.  With the politicized environment surrounding teaching the proponents of teacher "accountability" like this variation since it fits their ideas about the system retaining incompetent teachers that need to be weeded out.  In fact, the algorithm is defective and like many is based on erroneous assumptions.

I personally know that physicians are subjected to the same processes as teachers, but so far it is less technologically advanced.  O'Neill points out that there is nothing magical about algorithms.  That they frequently incorporate the biases of the people who design and contract for them.  Opacity and a lack of correction by feedback is another feature.  I worked for the same employer for a number of years when physician "accountability" measures were put in place.  The "algorithm" for salary went something like this RVU Productivity + Outside Billing + Citizenship = Pay.  RVUs were the total number of patients seen according the the biased government and managed care billing schemes.  Outside billing was any consulting work done outside of the clinical work that was billed through the department.  Citizenship included teaching and administrative duties as well as any Grand Rounds or CME lectures that were done.  In other words apart from the subjectively based billing scheme all of the inputs are almost totally subjective and influenced by all kinds of pseudoaccountability measures along the way.  For example, in parallel with the teacher ranked on the algorithm, I was told one year that I had achieved the top rank in terms of documentation in a group of about 25 physicians.  The next year - making no changes at all in terms of that documentation - I was dead last.  My conclusion, like the teacher in the example was that the rating scheme was completely bogus and with that kind of a scheme who cares about the results?

The number of based algorithms applied to physicians has eerie parallels to those mentioned in WMD.  Here are a few that I picked out on the first read:

1.  The algorithm is based on faulty data - the teacher evaluation algorithms were based on a faulty interpretation of data in the Nation at Risk report.  The report concluded that teachers were responsible for declining SAT scores between 1963 and 1980.  When Sandia Labs reanalyzed the data 7 years later they found that an great expansion in the number of people taking the test was responsible for decreased average score but subgroup analysis by income group showed improved scores for each group (p. 136).  The only reason that teachers are still being blamed is political convention.  I posted here several years ago that the top ranked students in the world in Finland are taught by teachers who are assumed to be professionals and who are not critiqued on test results.

The parallel in medicine was the entire reason that medicine is currently managed by the government and the healthcare industry.  It was based on criticism in the 1980s that doctors were lining their pockets by performing unnecessary procedures and that work quality was poor.  That should sound familiar because that criticism has been carried forward despite a major study that showed it was completely wrong.  The massive Peer Review Standards Organizations (PRSO) in each state in the 1990s conducted rigorous reviews of all Medicare hospitalizations and concluded that there was so little overutilization and so few quality problems that it would not pay to continue the program.  The only reason that managed care companies exist today is by political convention.

2.  An effective teacher like an effective doctor is too complex to model - When that happens only indirect measures or "crude proxies" (p. 208) can be used to estimate effectiveness.  In medicine like teaching - the proxy measures are incredibly crude.  They generally depend on diagnosis, poorly account for comorbid illness, and the outcome measures are heavily influenced by business rather than medical decision making.  The best examples are length of stay parameter and readmission parameters.  Every physician knows that there are set payment schedules based on the supposed ideal length of stay for a particular illness.  The business influence in the discharge decision is so malignant these days that non-physician case managers are present to pressure physicians into discharging patients.  If the discharge beats the length of stay parameter - the hospital makes money.  I sat in a meeting at one point and asked the obvious question: "OK - we have completed the discharge checklist - do we know the outcomes?  How do the patients do when they are discharged by this process?  How many of them die?"  Dead silence followed.  Most people would be shocked to hear that what passes for evidence based medicine is often a checklist that has no meaning in the real world.  Making the points on the checklist is good for advertising though.

3.  There is a lack of transparency in the overall process - The teachers in WMD who were blindsided by the algorithm were never told how that conclusion was reached.  I encountered the same problem in a managed care organization when it was clear to me that administrators with no knowledge of psychiatry were telling us what to do.  In some cases, "consultants" were brought in to write reports to confirm the most recent administrative edicts.  When I asked my boss if I could talk with the people sending out the edicts I was informed that there was a "firewall" between clinicians and upper management.  This lack of feedback is another critical dimension of algorithms gone astray.  If you are writing an algorithm biased toward a business goal - why would you want feedback from clinicians?  Why would you want any humanity or clinical judgment added especially in the case of psychiatric care?  Let's just have a dangerousness algorithm and leave it at that.  Those are the only people who get acute treatment, even though it is patently unfair relative to how the rest of medicine works.             

Big Data is good for science.  We can't do elementary particle physics or genomic analysis very well without it.  Big Data is also good for business is much different ways.  There are clearly people out there who cannot resist buying items online if the Amazon algorithm shows it to them enough times across a number of web pages.  Big data in business can also come up with billing algorithms that have less to do with reality than making a profit.  Similar programs can be found for employee scheduling, performance analysis, and downsizing.  The problems happens when the business biases of Big Data are introduced to science and medicine.  Those techniques are responsible for an array of pseudoquality and pseudoaccountability measures for physicians, hospitals, and clinics.

Unfortunately physicians seem to have given up to the political conventions that have been put upon us.  Some administrator somewhere suggests that quality care now depends on a patient portal into an electronic health record and a certain number of emails sent by patient to their physician every month.  Across the country that will result in hundreds of millions of emails to physicians who are already burned out creating highly stylized documentation that is used only for billing purposes.  Terabytes of useless information that nobody will ever read again - the product of a totally subjective billing and coding process that started over two decades ago.  Is there any data that email communication is tied to the effectiveness or technical expertise of the physician?  I doubt it.  I worked with great physicians long before email existed.

It is about time that somebody pointed out these manipulations provide plenty of leverage for the management class in this country at the expense of everyone else.  It is well past the time that doctors should be confronting this charade.  

George Dawson, MD, DFAPA


1.  Cathy O'Neill.  Weapons of Math Destruction - How Big Data Increases Inequality And Threatens Democracy.  Crown Publishing Group.  New York, NY, 2016.  I highly recommend this book for a look at the other side of Big Data.  It is written in non-technical language and is very readable.


The photo at the top is a Server Room in CERN By Florian Hirzinger - (Own work (Florian Hirzinger)) [CC BY-SA 3.0 ( or GFDL (], via Wikimedia Commons"   href=""><img width="512" alt="CERN Server 03" src=""/></a>

  CERN Server 03