[Dusts off cobwebs] — June 6, 2016

[Dusts off cobwebs]

  1. 90 percent of nurses indicated the EHR process changes diminished their ability to deliver hands-on care at the same effectiveness, yet only 5 percent of hospital leaders indicated EHR replacement had impacted care in a negative way.
  2. Buy this for me, please.
  3. Do people understand when we tell them they are dying?
  4. Also, this.
Links. — May 27, 2016


  1. Massachusetts prisoners placed in solitary confinement after meeting with state legislators about prison reform.
  2. Practicing medicine for the common good. I don’t think Rothbard is the answer, but this is important. “Care for the common good is a moral imperative, I agree, and resources must be used mindfully and as sparingly as possible.  But the common good is not an idealized concept that can be entrusted to bureaucrats.  The common good emerges when benevolent physicians help individual patients within the constraints of their legitimate resources.”
  3. Bookmarking for later: coercive citation in academic publishing.
  4. Bookmarking for later: dialysis for AKI.
  5. The Nathan Heller on activism and identity politics at Oberlin. Have read once. Need to read again and again. These are hard questions and Heller makes a good-faith attempt to understand them.


Morning Links — May 26, 2016
Quick thoughts on professionalism — May 25, 2016

Quick thoughts on professionalism

In my post last week about the Florida gun law, I mentioned that we get bad laws and other attempts to regulate and standardize physician behavior when we fail to do it ourselves.  That psychiatrists, eg, failed to report sexual misconduct is, I think unsurprising. Reporting your colleagues to authority bodies is hard.

This article from JAMA gives a little more insight into what people say keeps them from reporting:


One thing I’ve been thinking: reporting systems are fatally flawed.

Having difficulty putting this into coherent English. But, in order to work, any reporting system must be used. And if people are afraid to do that because they don’t trust whatever authority receives a report to deal with it appropriately and don’t trust their colleagues not to resent them or seek retribution, then people will report infrequently and only more egregious behaviors, and the reporting process will largely be punitive.  It’s a self-reinforcing spiral.

If you want to build a system where people report, you first have to build a system where transparency, open & critical peer-to-peer feedback and self-policing are expected, valued, and encouraged by everyone. Remediation when needed must be both effective and normalized.  And if you have all that, do you still need the reporting system?

Links —
Sunday Afternoon Links — May 22, 2016

Sunday Afternoon Links

  1. Should the Rio Olympics be cancelled or postponed because of Zika and more?
  2. IRBs need infosec expertise.
  3. Workers of the world, unite! But also, maybe try the practical approach before you accuse your CEO of being a fascist for making you do ICD-10 training, just sayin’.
  4. The case against the case against sex robots.
  5. In 1987, 65% of psychiatrists surveyed knew patients who had been taken advantage of sexually by past therapists, but only 8% reported it.
This is not a post about Florida. — May 21, 2016

This is not a post about Florida.

Update: The JGIM has just published an article on this that looks promising, but I haven’t had time to read it yet. The abstract claims ” existing data on physician-initiated conversations with patients about guns support a positive prevention effect.” Below I link to AAP guidelines that reference two negative RCTs on this subject.  Obviously, need to go to the data.  Don’t think it changes the thrust of that part of my argument, which is that there are probably good and bad ways to talk about guns and we should look at that question carefully.

The full 11th Circuit Court of Appeals is preparing to re-hear Wollschlaeger v. Governor of Florida next month, the legal challenge to Florida’s Firearm Owners Protection Act that prevents doctors from talking to patients about guns. A panel of three judges has upheld the law three separate times under differing levels of First Amendment scrutiny.  First things first: this is a bad law, and the decision to uphold it is bad – Dahlia Lithwick and Eugene Volokh talk more about the implications for speech more generally.  And the Harvard Law Review has an interesting commentary on moral conscience in professional practice – more on this later.

But even if the law’s upheld, I think lot of the writing about it from medical corners has been unhelpful.

I: Yes, you can still talk to your patients about guns.

Last week, the NEJM worried that “If FOPA is upheld, other states may enact laws regulating physicians’ speech pertaining to firearms. If so, physicians’ ability to counsel patients about gun safety will be significantly compromised.” Writing in the New York Times in 2014, Aaron Carroll laid out the case for anticipatory guidance about safe gun ownership in pediatric practice, and voiced similar concerns.

Screen Shot 2016-05-21 at 10.16.31 AM

A recent piece in Annals of Internal Medicine, on the other hand, suggests that the law does not prevent doctors from asking about guns in situations where there is particular relevance to patient care.  They offer this table as a guide.

Screen Shot 2016-05-21 at 10.21.43 AM.png

So, can pediatricians talk to parents about guns and gun safety or what?  It’s worth going to the law itself and the Court’s decision to uphold it.  Here’s the law, in part:

Screen Shot 2016-05-21 at 10.42.13 AM

The plaintiffs challenge that the “relevance” clause I’ve highlighted is too vague to be enforceable.  The Court spends a fair amount of time clarifying this (starting on page 28).  The gist is that by the plain meaning of the words, the law only implies that doctors should not ask about guns when they know or have good cause to believe that the question is not relevant.  So Annals gets it mostly right.

Here’s the Court again:Screen Shot 2016-05-21 at 10.48.54 AM.pngThis, I think, is what Annals misses. Nobody wants to be subjected to an investigation by the Board or face a potential malpractice suit. And although I think the court is pretty clear that if you are practicing according to the accepted standards you are likely to be protected, the law still introduces uncertainty to physician practice, and empowers patients to respond to seemingly reasonable questions with costly legal or professional complaints. So even with the assurance of protection, doctors are going to be less likely to talk to patients about guns. What do we do about that?

II. How did we get here?

In the case that sparked the legislation, a woman brought her 4-month old to a pediatrician in Ocala, Florida.  He asked her if she kept guns in the home. She asked him why he was asking. He may have explained that it was a routine question about preventative health and safety, or he may have said nothing, depending on whose account of the interaction you believe.  At the end of the visit, he told her to find a new pediatrician. She went to the news.  A number of other people had similar stories, including Florida state legislators.

Accidental injury, homicide and suicide are major causes of death, especially in otherwise healthy young people. Statistically, children in homes with guns are at greater risk of harm.  Although there are ways to mitigate those harms – lockboxes, trigger locks, storing ammunition separately from the gun – many people don’t use them.  The American Academy of Pediatrics “Bright Futures” guidelines recommend counseling parents about these safety techniques, and that the safest homes are ones without guns.

It’s a little bit curious that those same guidelines note two RCTs that have looked at whether counseling patients about gun safety leads to safer gun practices.  The answer is no.

One of the arguments made by the court is that it is “common sense that a doctor has no right to express to a patient her religious, political, or moral beliefs, – all irrelevant to medical care”. The Harvard Law Review commentary I mentioned above defends the right of doctors – under existing law and precedent – to cultivate their own practice style and to set for themselves the boundaries of personal and professional identity.  “As all people do throughout their careers,” they write, “a doctor must decide how important it is for her to communicate her personal beliefs to clients, what she is willing to risk in order to do so, and whether she is willing to pay the market price for her choice.”

It’s a nice theory, but how does it work in practice?

Let’s go back to Ocala.  The pediatrician in question, Dr. Okonkwo, was medical director of his practice.  Here’s how he explained his decision to terminate the mother who refused to answer his questions about guns:

Screen Shot 2016-05-21 at 11.06.52 AM

So, this is bullshit.

Building trust is an essential skill for a doctor. It is something to be earned, not due to us because we wear a white coat and have a diploma on our wall.  But there’s more: the kinds of patients who don’t instinctively trust their doctors, and refuse to divulge personal information simply because they were asked to, are also the kinds of patients who are marginalized, vulnerable, and have reasons to be distrusting.  They are the kinds of patient whose insurance is less likely to reimburse well.

I’d argue this doctor is staking his practice on objectionable moral grounds, even if he is within his legal rights. But what’s the “market price” for that? We don’t know how to tell a good doctor from a bad one in real time, and even if we did, it would be hard to get that information to patients in a way that was both reliable and fair.  As it stands, doctors are free to develop a practice “style” based on their moral principles and scientific beliefs, etc, but let’s say you want a doctor who believes she should have to earn your trust, and has some facility at doing with  patients with varying interests and values.  Where will you find that information? Unless you know someone who knows someone, you’ll learn it from trial and error, at the cost of time and money. Worse, you’ll get trapped with a doctor you don’t like.  If you’re also poor and sick, you have little recourse.

So is it any wonder we get bad laws and angry patients? Or other cumbersome attempts at standardization of practice?

III. Doing the same thing over and over and expecting it to be different.

If counseling patients about gun safety doesn’t make them safer gun owners, why do we keep doing it? Why do we insist on the right to keep doing it?

It’s not just guns that are the problem here. Educating vaccine-hesitant parents about the clinical data on vaccine safety and efficacy can backfire. Even on issues that aren’t politically polarized, it turns out that sometimes it takes more than an explanation to get people to change their mind or behavior. But the dogma here seems to be that if we proclaim the truth louder it will be more convincing.  A better approach might be to separate the practice of medicine a little from the notion that we are simply promoting scientific truth.  In reality, we’re selling health. And to do that we need to know what sells.

This gets back to a point I’ve promised will be fundamental for this blog. It’s not particularly useful to think of doctors as counselors of scientific truth, which patients then accept or reject.  All of our choices are between competing claims of authority and the grounds for accepting those claims. So from a physician’s standpoint, getting a patient to understand what the data says about vaccine or gun safety is only part of the job. The harder part is to give them reason to believe that we are trustworthy and reliable, reporting all the data, carrying no ulterior motives, capable of understanding and respecting people with very different values and priorities (on guns in particular, this piece by Dan Kahan and Donald Braman is very much worth reading).

Data alone won’t do it, and neither will firing a patient who doesn’t start off sharing those beliefs.



Links — May 20, 2016


  1. Do we spend too much time talking about homeopathy?
  2. Robert Centor does not like performance metrics. Related: why does the new sepsis measure define severe sepsis as lacate>2 when the evidence suggests the cutoff should be 4?
  3. 30% of women in academic medicine have experienced sexual harassment.
  4. Canada started taking demented seniors off antipsychotics. See what happens next. (Hint, they fell less, but also, their agitated and uncooperative behavior improved).  Would like to see this published, though.
  5. Rational suicide for the mentally ill in Canada? Also, friends don’t let friends visit xojane.


Surgical variability. For and against.

The ABIM Foundation moves its money offshore.  Best enjoyed by imagining that Richard Baron looks like Snidely Whiplash.

If you’re going to burn the system down, start with a brush fire. — May 19, 2016

If you’re going to burn the system down, start with a brush fire.

If you want to understand why doctors sometimes complain about being doctors, see this page by the Happy Hospitalist on CPT coding for admission notes (one presumes he is happy because he enjoys this stuff). The gist of it is that the government has developed a complex set of rules and language by which health care systems can communicate what they have done to and for patients that they should be paid for, and that language is arcane, cumbersome, and bureaucratic. Benefit to patient care is indirect and comes from translating medical practice into codes that can fed into data sets which can later be mined by analysts for trends and such; that has to be balanced by the up front costs of mind-numbing and tedious documentation, billing, and coding that could be better spent at the bedside.  Or golf course. Or …

Whatever. If you’ve ever tried to figure out what was going on with a patient by reviewing charts from before this era of forced thoroughness (or from old-school doctors who buck the system) you’ll know that part of the reason Medicare feels it has to nanny doctors is that has to – too frequently, we write shitty incomprehensible notes that only make sense to us at the time we write them.  There are other plusses and minuses of this system and I’m obviously not going to get into it here.

But I’m going back on service next week and have been thinking a little bit about the physical exam. Depending on who you ask, it’s an art that’s dying, dead, or not dead enough. It’s not taught well in medical schools, which means the next generation of teachers doesn’t know how to teach it well, which means, etc.  Also, why bother with a physical exam when we know that it’s often unreliable and, like, you’re going to get a chest x-ray and labs anyways, or maybe you’re going to learn how to use an echocardiogram probe that plugs into your iPhone, and with a thorough history those things are probably going to give you the answer you want.

My physical examination skills are not good, and haven’t gotten better with years of practice.  At the same time, my skill at talking to patients and interpreting test results has improved dramatically.  I can’t help but think that CPT codes have something to do with this.

Screen Shot 2016-05-19 at 8.17.52 AM

Here are the requirements for a comprehensive physical exam: nine organ systems or body areas with two bullet points in each. There’s something about a complete single organ examination, but I don’t know anyone who knows what that means, and I’ve never read a chart where that’s been documented.  For patients in the hospital, the expectation is to do this generally complete exam for every patient every day.  That’s insanity, and it reduces the physical exam from a test to be applied rationally and as the situation warrants, to a list of boxes to check off according to rote habit.  In fact, much of my physical exam course in medical school was precisely aimed at teaching us how to memorize and routinize the practice of doing a head-to-toe exam in as little time as possible. Again, crazy if your goal is to learn something about the patient and their disease. It’s helpful as performance art, maybe.

I had a cardiology attending once who said when he trained they used to try to guess the right atrial pressure by physical exam, and then verify it with a right heart cath at the bedside.  That’s, well, fucked up.  But also instructive.  A cardiac examination with the goal of answering the question “what’s the right atrial pressure?” is more useful than one with the goal of answering the question “you say you have chest pain so can I learn something about it by putting a stethoscope on your chest for a few seconds?”

Anyways, I could say more here about false positives and false negatives, but instead I’m going use this space as a commitment strategy. I’m going to ignore CPT coding requirements for the forseeable future, and try to use the physical exam as a rational tool of investigation. I’m guessing it’s the only way to really teach myself how to do it well.  Will use this blog to update on progress.



Hot Links. —