From Journal of the American Medical Association: About 500 doctors (and a handful of nurses) in training participated in a randomized controlled trial on communication skills. Half received thirty hours of intensive state-of-the-art communication skills training; the other half didn’t. Then they all saw a bunch of patients with really serious diseases, who rated the quality of the care they received. Doctors who got the training were rated no better than doctors who hadn’t, and in fact did significantly worse on the secondary outcome of how depressed their patients became about their illnesses. So even very very good communication skills training seems to have either zero or negative effect.
I would like to note for the record that I totally called this one.
To be fair, there is a large body of contradictory research. And if someone challenges me in the comments, I’ll be happy to go over why I trust this study more than some of the others.
But you know who agrees with me here? The New York Times. Their analysis is very, very close to my own heart:
In response to the growing recognition that effective communication with patients is a basic competency of our profession, and that doctors often have inadequate training in it, medical schools and hospitals have invested substantial resources over the past decade to teaching communication skills. […]
But can giving doctors a script for empathy actually make them more empathetic? Our patients know better. […]
Perhaps those residents participating in the study were too early in their training to lead these sorts of difficult conversations and participation in the study emboldened them in inappropriate ways.
It is also possible that, as we devote more time to teaching students and doctors effective communication techniques, we risk muting their authentic human voices, and instead of learning to connect they apply rote tools and scripts.
No communications course will magically transform lifelong introverts to hand-holders and huggers. At the same time, we must ensure that we are not converting people who genuinely care about their patients into people who only sound as if they care. Having physicians sound like customer service representatives is not the goal.
For those doctors who are emotionally challenged, communications courses can provide the basics of relating to other human beings in ways that, at the very least, won’t be offensive. But for the rest of us, we should take care to ensure the techniques and words we learn in such courses don’t end up creating a barrier to authentic human contact.
The doctors who wrote the NYT article (one of whom runs a leukaemia clinic! Talk about jobs that require communication skills!) go on to describe a horrendous communication skills class they attended, which somehow ended up being taught by a doctor who was legendary in the area for his total lack of communication skills and human empathy.
This, too, was close to my own heart. When I work with a new attending, I will very often check them out on ratemds.com or healthgrades.com to see what their patients think of them. These sites are not very valuable on the margin; you do sometimes get one guy who had to sit in the waiting room a little too long going on a personal campaign to destroy someone’s reputation. But on the tails, if there’s a doctor who is universally hated by every single one of her patients, this can be a strong warning sign.
And so when some of my attendings try to “correct” my communication style or give me helpful tips for “how to make patients like me”, it sometimes takes everything I’ve got not to retort “I think I’m going to limit myself to taking advice from doctors with better than a 1.0/5 stars patient rating.”
The doctor I’m working for now has a better than 1.0/5 stars patient rating. In fact, he has a 5.0/5 star patient rating. He has not tried to give me any advice on communication skills, which I am starting to notice is a common thread among doctors with excellent communication skills. But I have been trying to absorb what I can through osmosis.
Yesterday I watched him diagnose a woman with very advanced metastatic lung cancer – pretty much a death sentence. He showed the patient a CT image, told her she had cancer, and pointed out the location of the primary tumor in the right mainstem bronchus. Then he started talking about the anatomy of the bronchial tree.
The bronchial tree, for those of you not lucky enough to know, is the system of airways that gets oxygen into the lungs. It starts with the trachea, or “windpipe”, which divides at an area called the carina – around the level of the fifth thoracic vertabra – into the right and left mainstem bronchi. These bronchi then divide into smaller bronchi, which themselves divide into structures called “bronchioles”…
…and there is no point in continuing because by this point you are asleep. I do not blame you. No human being on Earth cares about the anatomy of the bronchial tree. I voluntarily went into medicine and even I have to make an extraordinary effort of will not to immediately change the subject to something less mind-numbing. If a communications skills tutor saw you discussing the anatomy of the bronchial tree immediately after diagnosing a patient with metastatic lung cancer, she would arrange to have you shot.
And I was thinking about this later, and I realized that what my attending was doing was providing white noise. If he had done the proper Communications Skills thing and said “And how does this make you feel?”, our patient would have been put on the spot. We would have been demanding some kind of coherent reaction to the diagnosis, and how in the world are you supposed to react to that five seconds after hearing about it? And if my attending had asked her if she needed a few minutes to collect her thoughts, she would have been socially obligated to say no, because it would have made her look weak and wasted a busy doctor’s time.
So instead he started talking about all this anatomical stuff, stuff that no human being could reasonably be expected to listen to, and it was his way of saying “Go on, tune me out, pull yourself together, and then when I’m done we can have a discussion about cancer and what to expect from it.”
And it worked. I’m not saying there aren’t ninety-nine ways this could have gone horribly wrong, but in this case and for this patient, it worked. And he knew it was going to work because he’d been seeing this patient for twenty years.
And I have no particular moral to this anecdote, or thoughts about how it relates to teaching communication skills. I just thought it was interesting. But…
Economist James Miller recently gave the following advice to LW college students pondering lucrative earning-to-give careers:
If you have a high IQ and are good at math go into finance. If you have a high IQ, strong social skills but are bad at math go into law. If you have a high IQ and a good memory but weak social skills, become a medical doctor.
There are a couple of different ways to have weak social skills. There is the way where you’re scared of large groups and being the focus of attention. There is the way where you dress unfashionably and smell bad. There is my personal preferred way, which is to have no idea what to do at a party and end up sitting in the corner really awkwardly and not being invited back. But there’s also the way where you consider human emotions an annoying inconvenience and prefer to just plow through them like a rhinoceros in a china shop.
If you are the last one, and you insist on going into medicine, do everyone a favor and become a pathologist, a radiologist, or a surgeon who works with a really good anesthesiologist.
Because communication skills classes won’t save you.
> If you are the last one, and you insist on going into medicine, do everyone a favor and become a pathologist, a radiologist, or a surgeon who works with a really good anesthesiologist.
I wonder if anesthesiologist is one to add to the list? Your patients tend to be unconscious for long periods of time…
Surgeon is traditionally listed as one, but might not be true because a lot of surgeons follow their patients up before and after surgery.
Anaesthesiologist is not traditionally listed as a good job for people without social skills. I think it might be because you have to deal with the patient before surgery and start the process of getting them anaesthetized, which is usually pretty scary.
There are also special challenges for anyone who does paediatric anaesthesia or obstetric anaesthesia. This second one is especially nerve-wracking because a lot of the time it is requested when some well-intentioned “natural childbirth” plan didn’t work, and so you have a woman giving birth who’s in terrible pain and demanding you fix it RIGHT NOW when anaesthesia is a pretty delicate procedure that takes at least a little time to set up.
‘But there’s also the way where you consider human emotions an annoying inconvenience and prefer to just plow through them like a rhinoceros in a china shop.’
As someone who has worked with- and only with- public relations guys, lawyers and historians* my experience has always been that this sort of person gets sidetracked to the back offices very quickly after the complaints come in. I get that you can’t just stick all the physicians in hidden cubicles, but there’s got to be a solution. Maybe you could start a program where these sort of doctors wear menacing sound-proof ritual garb while near the patients, and then well-spoken and smiling medical interns do the social interactions!
‘The doctors who wrote the NYT article … go on to describe a horrendous communication skills class they attended, which somehow ended up being taught by a doctor who was legendary in the area for his total lack of communication skills and human empathy.’
…on the other hand, this seems to happen with alarming frequency. My suspicion is that the people who teach these courses are chosen based on suboptimal qualifications. I strongly suspect that the bulk of them get chosen by the hosting organizations on the basis of ‘who isn’t too busy to prepare this?’
You’re totally right about the communications classes not fooling anyone. I’m not convinced that you can’t train (some? most?) people to be empathetic and natural speakers, but probably you can’t do that with a few hours of continuing-education training. You certainly won’t learn anything approaching ‘natural’ from any program a reputable business will approve for its employees.
~
*My ideal career path, of course, is to become an evil bureaucrat with excellent business cards and many artifacts of loathsome power.
“My suspicion is that the people who teach these courses are chosen based on suboptimal qualifications. I strongly suspect that the bulk of them get chosen by the hosting organizations on the basis of ‘who isn’t too busy to prepare this?”
Not infrequently, teaching spots in medical schools are filled by those professors who wanted to do research but whose grants didn’t end up getting funded. From the med school’s perspective, it’s a way of at least getting a few lectures worth of value out of someone who didn’t happen to bring in the big bucks to the department this funding cycle.
So if you ever got a sense that your lecturer would really rather be back in the lab and didn’t want to be there teaching, it could be because they would really rather be back in the lab and didn’t want to be there teaching.
I think this is the key problem. The actual principles of interpersonal interaction are massively unethical to think about, much less communicate to others. And so we are left to mostly get by on intuition. We have a word for people who teach communication skills, it’s “PUA”.
Well, GOOD sales training seminars do it too; they just reserve it for their money-makers.
Jordan D. wrote “As someone who has worked with- and only with- public relations guys, lawyers and historians* my experience has always been that this sort of person [i.e., the rhino in the OP] gets sidetracked to the back offices very quickly after the complaints come in.”
I ended up looking for a job after my quantum-simulations-in-chemistry Ph. D. in the midst of my divorce, and ended up in reasonably high end tech support (not laptops or anything, but workstations and clusters handling things like credit card databases) for IBM. I had some misgivings about it — I do not think of myself as a great customer service sort of person, and I doubt many people who know me do either. Indeed, I and they are more likely to think “rhino.” (Though reasonably disciplined and sensible, so I didn’t fear a lot of outright customer complaints. Just customers who weren’t properly soothed or encouraged otherwise pleased by soft people skills.) But as best I could tell, things went very well for me with customers (which was not a given, since they did not for every person who worked with me). Perhaps it was an artifact of the main metric they used to evaluate us — we had no particular control over which calls we took, and once we took it we were judged on satisfactorily (from the customer’s point of view) resolving the issue, no matter how many more interactions it took. Something involving e.g. generating sales leads might have been much harder for me to do well. Possibly that was the wrong metric for them to use, but it seems like a plausible choice for a business with worried customers trying to fix potentially major problems.
The sample size was not huge, though, because I only worked there a couple of months. In one curiously eventful week my IBM shift boss announced that our mandatory lunchtime meetings were not to be put on our timesheets, my supervisor at the contracting company informed me that even though they hadn’t thought to tell me when they hired me it was mandatory for me to start studying and testing for certifications on my unpaid time, my floor boss at IBM sent out a memo saying we should all round our timesheets down in the company’s favor (I have wished many times that I had saved it — part of the explanation involved time that we spent going to the bathroom), and one of the recruiters I had applied with when looking for this job pulled my name out of some file and called to ask if I was available for a significantly-better-paid developer position at Nortel.
I don’t know whether I’d make customers happy in a field where they were less preselected for a stressful situation, but I think conveying competence and seriousness went a particularly long way when almost all the customers were calling about something they were quite worried about. So if you have someone who’s basically capable and polite but a little too impatient with human folly to fit in well in the airbrush section of your body shop, and whose talents don’t really match working the back office, you might consider shifting that person to the major collision repair section.
I’ve taken communications training classes that were not about following a script, but about kindness and genuineness. That seemed to be good stuff. (I haz Aspergerz, so maybe not all of it stuck.)
I wonder if the lesson here is “doctors get sucky communication training” rather than “communications training sucks.”
I care because there’s some small probability that I will teach communication skills classes at some point. (This might be a bad idea…)
I was thinking that. I mean, I think teaching sympathetic communication to doctors IS hard, because it’s so much harder to be nice when you’re incredibly rushed and have to tell someone they’re dying of cancer.
But when Dale Carnegie told people “be nice to people and you’ll make lots of money” and “be nice to people and you’ll make them feel nicer about themselves” the first message got a lot more traction than the second, which suggests maybe you can teach communication skills, but only with effort, and only to people who genuinely want to be good at it?
Alternatively, become a surgeon with a really bad anesthesiologist.
>Doctors who got the training were rated no better than patients who hadn’t
Pretty sure you mean “than doctors who hadn’t”.
Also, why does the preview thing tell me that I can’t put linebreaks between my paragraphs?
EDIT: especially when it turns out I CAN?
Of all the various medical specialists I worked with after my stroke, the specialty I was most consistently impressed by in terms of their communication skills was neurologists.
Admittedly, I only worked with four all told, which is not a meaningful sample, but I can’t think of any other class of doctor I worked with where of four doctors there were no clunkers.
Hell, I can’t even say that about therapists, which you would think select for communications skills. (I mean physical/occupational/speech therapists here. I’ve only ever talked to two “therapist” therapists, one of whom turned me off after 30 seconds.)
Oh… the above was me.
I afford to you more credibility than I afford to the New York Times, whose science reporting is frequently very poor. (Examples on request, but I don’t want to derail.)
As a layperson on the subject of science, I’d be really interested in knowing good, reliable sources for actual scientific research as well as examples of such poor reporting, for contrast. Being one to drift along the blogosphere, the chances of information being misrepresented can be quite high, as evidenced by the ‘Lies, Damned Lies and Social Media’ series of posts.
I don’t think there’s any great substitute for having the technical/statistical literacy to critique reports the way Scott does, but there’s a whole industry devoted to cranking out interesting garbled misrepresentations, and Scotts are few. There are a lot of good science bloggers who do much better work–Derek Lowe, for example–but the technical bar is usually a lot higher for understanding them. There’s no royal road to geometry.
An important defensive tool is simply to remember the Gell-Mann Amnesia Effect. Pay special attention to news reports on matters you actually know something about, remember how many mistakes you find, and take that as your prior for news reports on matters you aren’t personally familiar with.
Rev’s Law is also relevant.
Having said all that, Wikipedia is actually not a bad source on average for material of no political interest.
As an example of really abysmal science reporting in the NYT, consider Obesity Rate for Young Children Plummets 43% in a Decade.
There are two big problems with this article.
First, the 43% claim is wrong eight ways from Sunday. It putatively applies to 2-5 year olds; there was no significant change found in younger or older groups of children. Because of this, the significance test was done incorrectly (roughly, they didn’t account for the Green Jellybean Principle), and the authors of the original report admit that in the text, though they don’t actually try to fix it. There are other, more technical problems with the claim, which this margin is too small to contain.
Second, and more absurdly, the NYT tries to somehow imply that Michelle Obama deserves credit for this already questionable conclusion. But the ‘change’ was over a decade, between the 2002-2003 survey and the 2011-2012 survey, and MO didn’t start her nutrition campaigns until 2010!
(ummm I mean Derek Lowe etc. are much better than the average NYT science reporter, not that they’re much better than Scott)
Science reporter Nicholas Wade of the, um, NYT, is excellent. His main beat is Genetics. Wade has a book coming out in a month or two on Human Biodiversity, which may well make him former science reporter at the New York Times.
The less political the subject, the better is Wikipedia as a source. No surprise then, it is highly unreliable on matters pertaining to Global Warming.
I like Harold McGee, too. Perhaps I should have said average NYT science reportage.
Hi. You are still the best writer on the web. Good stuff.
Have you seen this study? Because it strikes me as pretty relevant:
http://www.nejm.org/doi/full/10.1056/NEJMoa1204410#t=article
Patients had terribly inaccurate ideas about what palliative care was supposed to do, and were apparently displeased when doctors managed to correct those misunderstandings.
“In multivariable logistic regression, the risk of reporting inaccurate beliefs about chemotherapy was higher among patients… who rated their communication with their physician very favorably, as compared with less favorably (odds ratio for highest third vs. lowest third, 1.90; 95% CI, 1.33 to 2.72).”
and
“Our results also provide evidence that physicians have some ability to influence patients’ understanding. The observed association between inaccurate beliefs about the likelihood of cure and higher ratings of physician communication suggests a link between physicians’ communication behaviors and patients’ understanding of treatment benefits. This suggests that patients perceive physicians as better communicators when they convey a more optimistic view of chemotherapy.”
Since the primary outcome of the study you’re writing about is ” patient-reported quality of communication”, I suspect there’s something similar happening here, which would explain why giving providers communication training made patients more depressed. We’re lucky, in fact, that the communication training didn’t make QOC ratings significantly worse.
I get your point, and the study I’m citing does go into that a little more than I did. But one thing jumps out at me:
“The risk of reporting inaccurate beliefs about chemotherapy was higher among patients… who rated their communication with their physician very favorably, as compared with less favorably”
Wouldn’t this mean that there’s also a mechanism by which communication skills training could decrease depression?
We’re not accidentally picking up that patients with worse diseases rate their doctors as worse communicators, because it’s an RCT so we avoid these kinds of correlational issues. We could posit an effect where doctors with more communication skills training are more honest and therefore patients learn worse news and become more depressed, but the most plausible route from communication -> honesty is that patients get along better with their doctor and so talk to them more. And it looks like that wasn’t happening here.
Am I missing something?
I’m not sure. Maybe? I feel like you’ve gone wrong, but I can’t figure out where. Did you just misread?
Here’s the thing: in the study I cited, patients who said their doctors were better at communication understood their condition worse. So what I’m saying is, when the news is bad (as in both these studies), maybe good communication makes patients unhappy? And they then rate their doctors as bad communicators who made them unhappy?
The thing is that most people, when they say “communication skills,” mean both “successfully communicating information” and “making people feel good.” But in bad-news situations, those goals are in tension. Maybe it would be better to try to control for how well patients actually understood what they were being told– give them a little quiz after the encounter, and correct for the results?
Okay, this makes sense and I agree.
Thinking to myself: it looks like you’re proposing that good communicators tell patients how bad their chances are, and bad communicators don’t. This makes the good communicators’ patients more depressed. It also means that the bad communicators’ patients like their doctors more because they have happy fuzzy associations of hearing good news from them.
My only two concerns about this theory are – first, that it’s not immediately obvious why good communicators (as in doctors who took a communication skills course) would necessarily be more honest about their patients’ chances. Perhaps the course “emboldened” them (as the NYT blog put it) to think they could get away with breaking bad news more than they actually could? Second, that it requires that patients confuse “My appointment today went well because I heard good news” with “my appointment today went well because my doctor is a good communicator”, which seems to go a little beyond what your study supports and which is plausible but by no means certain.
My interpretation was that both sets of doctors tried to be honest, but the ones with communication training were perhaps better at getting their point across. The study I cited does pretty well establish that patients get the wrong idea a lot, and I don’t think it’s because most of their doctors were trying to lie to them.
I’m concerned that your study doesn’t distinguish between breaking bad news in such a way as to make patients as comfortable as possible with it (good communication), and breaking bad news in such a way that patients don’t fully understand the badness of it (one type of bad communication). You and the NYT seem to think training caused doctors to fail more at the first thing, but I’m proposing that maybe they just successfully avoided the second thing.
On your subject matter of bosses insisting that interactions with patients be in a contrived script to avoid all possibility of surprises of any sort, and doctors being expected to sound like call-center reps (personally, I wish that call center reps would sound a bit less like call-center reps…), that seems to be part of a big trend across practically all parts of society these days. In a totally different area, this blog post laments the lack of spontaneity in TV late-night talk shows these days; they’re all thoroughly planned and scripted to avoid all possibility of embarrassing surprises. That seems to be what Our Corporate Overlords want… their goals of avoiding liability risk, promoting a consistent marketing image, and cutting costs by outsourcing everything to the lowest-price places (and hence getting bottom-level workers of low skill who can’t be trusted to show any personal initiative) have resulted in a world where everything seems phony and contrived, like it’s all made of cheap plastic.
I have no good solutions, but I think Scott is right that scripting responses is a bad idea for people with some sense of humanity.
As a prosecutor who handled a lot of vehicular manslaughters and a dozen or so murder cases, I have worked with lots of grieving people from all walks of life. I think there are some interesting parallels.
I think the basics are kind of trainable – express sorrow for their loss, always return calls, explain the basics of the criminal justice system, don’t make promises you can’t keep, be very patient, repeat key concepts, and tell the truth.
But it sounds like the training is to find the canned answer that best fits the question. That strikes me as a bad idea, unless you really suck at this (as Scott expressly notes). I know of very little such training in my profession. There are trained victim advocates/counselors, but for legal answers, you need the prosecutor.
I know there’s a perception that good-communicating doctors are better at their job. In my limited experience, there may well be a real skill correlation. Maybe not communicating things clearly is because the doc is not good at what he does. (For lawyers, it’s an easier question: If your lawyer can’t communicate clearly, she’s almost definitionally bad at her job.)
Calling it “unteachable” seems a bit premature. Have we seen what happens when capable people incentivized to actually succeed make a serious effort?
http://whatshouldwecallmedschool.tumblr.com/post/75404194322/empathy-training
I am regular visitor, how are you everybody? This piece of writing posted at
this site is actually fastidious.