Sentences to ponder

Countless times, I have found that it is only during the physical exam that patients reveal what is truly on their mind. Whether it is the cough that they are reminded of now that I am listening to their lungs, or whether it is the domestic violence, the eating disorder or the genital symptoms that they feel comfortable revealing once we are in a more intimate setting — there is something about touch that changes the dynamic.

That is from Danielle Ofri’s interesting piece on the physical exam in medicine, via Jeffrey Flier.


Tyler dismissed the tactics of the pick-up artists in Neil Strauss's "The Game" (perhaps justifiably) in Tyler's Discover your Inner Economist. For what its worth, "touch" (kino as they call it), is a big part of the strategy.

I make no representations on whether they or Danielle Ofri are right or wrong... just sayin.

He categorically dismissed every single tip and strategy?

Touche. He did accuse Struass of peddling "snake oil" though...

That effect has been noted in the conversations of women with their hairdressers.

You would think that there would be more efficient ways of achieving this goal (such as anonymous online consultations). This reason was also given to justify routine pelvic exams, a rather uncomfortable procedure which shows no medical benefits.

People are not robots, and interaction via human touch is, for whatever reason, useful in forming a bond.

Software will eat this too. Oculus Rift VR + Haptic feedback.

...anecdotal evidence is a good place to start, but a bad place to end

The implications are obvious: program a robot to ask the questions and hire a low-skill technician to touch the patient while the robot does this.

I wonder if I were new at reading this blog if I would find this as hilarious as I do.

Does it work for airport security as well? "Now that you've grabbed my boob, I'll confess I'm hiding 105 ml of shampoo in my hand luggage."

What if instead of seeing a doctor for five minutes, you met with someone whose job is to elicit symptoms from you in a way that can be input into a computer. So they spend more time with you, they touch you, they know what questions to ask, they know how to rephrase the things you say - basically just trying to help you clarify exactly what's up in a way that can be put into a WebMD-type program that factors in symptoms, probabilities, and maybe genetic information too. It could then tell you which tests to run and connect you with the appropriate specialist, possibly remotely. Terrible idea?

Symptoms are only half of the equation. You would need someone with training ie a doctor to observe signs in a physical examination. I don't see how a computer could easily substitute for this.

My eye doctor's office is doing that -- the problem in their implementation is that the pre-work job requires a more highly-skilled person than they wanted to pay for. The guy left his computer on when he left to tell the doctor that I was ready and I looked at what he had written. He had misunderstood pretty much every thing I had told him and misspelled several important words; it was pretty clear to me that he didn't know much about eyes or vision. I told the doctor about this and showed him the errors but I don't know what the follow-up will be.

My old HMO used a physician's assistant in the same role and that was much more of a success because he'd had the training to have the context.

So it's really a matter of costs and control as you'd expect.

As a practicing physician, I find that routine physical exams are next to worthless. The author's and my anecdotal experiences aside, research demonstrates that routine physical exams are not associated with with any positive outcome compared to only doing physical exam in association with the presenting complaint. For anyone interested, search the Cochrane Review website for what meta analyses have to say about physical exams

Any search tips? Neither searching for "physical exam" nor browsing produced any results other than some interesting unrelated reviews.

Seems like it would be easy enough to run controlled experiments to test this hypothesis. Let's do that rather than relying on anecdotes.

If a doctor is doing a full general exam he / she is:

1. Not looking you in the eye (except the eye part), and

2. Quiet.

When the doctor is talking or is looking at you, it's really hard to bring up the symptoms you wanted to ask about.

Interesting about how the doctor in the interview keeps talking about how busy she is and the other doctors are. My alarm bells always ring when someone is telling this at my work, I don't want heroics, I want consistent dependable results, so if this is real then I need to either simplify their job or add more people. So if doctors are so busy and their analysis is compromised by lack of time, we should expand the supply of doctors. I mean we would be concerned if, say, airline pilots confessed that they didn't have the time to do their job properly because they are too busy. I think the current argument against expanding the supply of doctors is that there is a limited supply of people capable of being doctors, sort of like Investment Bankers argue that the reason they get paid so well is that there is a limited supply of people who could do their very intellectually demanding job. In another parallel with doctors, the IBs also are always claiming they are very over-worked. Isn't it more likely that in both cases what we have is someone in a rent seeking position, who is reluctant to share the rent with other people? I don't mean that doctoring is simply of no value, but just that doctoring and investment banking are areas where the lump of labor fallacy is not in fact a fallacy - double the number of doctors and you will halve their work (also halve their compensation) because the work (sick people) is really fixed. So the question is, is doctoring really so special that only a few gifted souls can do it or is it the case that doctors are operating a guild system to minimize rent sharing? In the same way as I wonder what value IBs actually add, I wonder what doctors can do that a basic expert system could not do. If that were true that diagnoses are not computable I would be worried that medical diagnoses were so dependent on the magical, un-computable skill of an individual doctor. How could you know if your doctor had the magical skill and was not just making it up as he or she went along? How would you teach for such a skill that can't by definition be coded? Its too scary a thought.

"because the work (sick people) is really fixed". Ha!!! Compare medicare and other utilization rates.

Comments for this post are closed