![[personal profile]](https://www.dreamwidth.org/img/silk/identity/user.png){kind=small}
Last month, I read Lisa Sanders's Every Patient Tells a Story. Sanders is talking about diagnosis, how it can be difficult, and how the process has changed over the last 20-30 years. The "annual physical" is a thing of the past, and many doctors doubt that a physical exam is ever worth doing, given modern medical testing (from blood tests to MRIs). Other doctors regret the loss of the physical exam, but suspect it's too late to save it: you can't teach skills you never learned. Sanders falls somewhere in the middle: not that everyone should have an annual physical, but that there are things the doctor will learn from talking to the patient, watching them walk, and so on, and that these are things worth doing if the diagnosis isn't obvious.
The book is organized as a series of case studies: Sanders has a magazine column on tricky diagnoses. Tricky depends on context, of course: she discusses someone whose Lyme disease took ages to be diagnosed because she was in the southeastern U.S., dealing with doctors who filed Lyme as a zebra if not a quagga ("if you hear hoofbeats, think horses, not zebras": this is a useful starting point, but sometimes there's a zebra lurking in the background).
At the end, Sanders recommended "How Doctors Think," by Jerome Groopman, for readers who wanted to follow the subject further. Groopman writes about exactly what the title says: the ways doctors are taught to think, how that works well, how it can go wrong, and ways to get the thinking back on course. A lot of diagnosis is pattern-matching: this cluster of symptoms looks like such-and-such. Often, that works.
There are inherent problems in this (and they aren't limited to medicine). One problem is confirmation bias: if you think the answer is X, you'll look for evidence that supports that idea, and often stop when you find enough. A related problem is that we tend to be looking for whatever we just saw: this is fine when you're looking at an epidemic of H1N1 influenza, or the twelfth heat stroke case of the week, but it means a doctor may not think of some other disease that has the same symptoms she's noticed.
Once we have an answer, we tend not to question it. In medicine, that can mean that if the doctors think a patient has an autoimmune disorder, and the first two treatments they try don't work, they'll suggest a third, risky treatment, rather than consider that maybe the problem is something else. (This is one of Groopman's examples, an infant adopted from Vietnam whose diet in the orphanage was missing a crucial nutrient.) There's also a tendency (again, not just in medicine) to assume that there is a single problem: if someone is diagnosed with diabetes, the doctors won't look for another diagnosis, they'll try to explain all symptoms as related to the diabetes.
Another problem is that doctors don't like saying "I don't know" or "we really can't do anything for this." That can lead to lots of tests, possibly past the point of likely usefulness. It can also lead to unnecessary or harmful surgery. On this, Groopman talks about himself as a patient, specifically with regard to a back problem and some hand problems. He also notes that no surgical treatment for back pain, on average, is better than placebo.
Groopman has some suggestions for dealing with the communication problems. For example, he advises doctors who are having trouble diagnosing or treating something to ask the patient to describe the problem as if she'd never met him before. From the patient side, he suggests asking "what else could this be?" if the first treatment isn't working.
He also notes that doctors—and patients—tend to judge people using social cues that can be problematic in the medical setting. For example, a patient who has a psychiatric disorder (or has been diagnosed with one) will be taken less seriously. He describes someone with an eating disorder who almost died because, when she was eating 2500-3000 calories a day to try to put weight on and failing, the medical staff assumed she was lying about her diet. It turned out that yes, she had an eating disorder. She also had celiac disease, and the "pack on the carbs" approach was making things worse. (I'm not remotely surprised here, but it's good to have doctors aware of this problem.) It also affects what doctors someone will go to: a cancer patient may choose an oncologist based on personality match, so people who want to dive in and fix everything will get more aggressive treatment, not just because they ask for it, but because they choose doctors who suggest it.
Groopman and Sanders are both practicing physicians, and build a lot of their books on anecdotes of their own patients, and patients they were told about. In a chapter on how money is affecting the practice of medicine, Sanders talked about a testosterone salesman who went from trying to wine and dine her to trying to bully her, saying "I need you to write three prescriptions a week for the next month." Completely disregarding what her patients might need. Yes, testosterone levels decline with age; that does not mean that most older men should be getting testosterone supplements. (This is one where the "normal" levels are probably useless, because of the amount of variation not just between men, but in a normal man's testosterone level over a 24-hour period.)
If you're going to read only one of these, I'd say start with Sanders.
[I have a large backlog of book posts I may never write; this one is prompted by the need to return the Groopman book to the library, and the Sanders went back a month ago.]
The book is organized as a series of case studies: Sanders has a magazine column on tricky diagnoses. Tricky depends on context, of course: she discusses someone whose Lyme disease took ages to be diagnosed because she was in the southeastern U.S., dealing with doctors who filed Lyme as a zebra if not a quagga ("if you hear hoofbeats, think horses, not zebras": this is a useful starting point, but sometimes there's a zebra lurking in the background).
At the end, Sanders recommended "How Doctors Think," by Jerome Groopman, for readers who wanted to follow the subject further. Groopman writes about exactly what the title says: the ways doctors are taught to think, how that works well, how it can go wrong, and ways to get the thinking back on course. A lot of diagnosis is pattern-matching: this cluster of symptoms looks like such-and-such. Often, that works.
There are inherent problems in this (and they aren't limited to medicine). One problem is confirmation bias: if you think the answer is X, you'll look for evidence that supports that idea, and often stop when you find enough. A related problem is that we tend to be looking for whatever we just saw: this is fine when you're looking at an epidemic of H1N1 influenza, or the twelfth heat stroke case of the week, but it means a doctor may not think of some other disease that has the same symptoms she's noticed.
Once we have an answer, we tend not to question it. In medicine, that can mean that if the doctors think a patient has an autoimmune disorder, and the first two treatments they try don't work, they'll suggest a third, risky treatment, rather than consider that maybe the problem is something else. (This is one of Groopman's examples, an infant adopted from Vietnam whose diet in the orphanage was missing a crucial nutrient.) There's also a tendency (again, not just in medicine) to assume that there is a single problem: if someone is diagnosed with diabetes, the doctors won't look for another diagnosis, they'll try to explain all symptoms as related to the diabetes.
Another problem is that doctors don't like saying "I don't know" or "we really can't do anything for this." That can lead to lots of tests, possibly past the point of likely usefulness. It can also lead to unnecessary or harmful surgery. On this, Groopman talks about himself as a patient, specifically with regard to a back problem and some hand problems. He also notes that no surgical treatment for back pain, on average, is better than placebo.
Groopman has some suggestions for dealing with the communication problems. For example, he advises doctors who are having trouble diagnosing or treating something to ask the patient to describe the problem as if she'd never met him before. From the patient side, he suggests asking "what else could this be?" if the first treatment isn't working.
He also notes that doctors—and patients—tend to judge people using social cues that can be problematic in the medical setting. For example, a patient who has a psychiatric disorder (or has been diagnosed with one) will be taken less seriously. He describes someone with an eating disorder who almost died because, when she was eating 2500-3000 calories a day to try to put weight on and failing, the medical staff assumed she was lying about her diet. It turned out that yes, she had an eating disorder. She also had celiac disease, and the "pack on the carbs" approach was making things worse. (I'm not remotely surprised here, but it's good to have doctors aware of this problem.) It also affects what doctors someone will go to: a cancer patient may choose an oncologist based on personality match, so people who want to dive in and fix everything will get more aggressive treatment, not just because they ask for it, but because they choose doctors who suggest it.
Groopman and Sanders are both practicing physicians, and build a lot of their books on anecdotes of their own patients, and patients they were told about. In a chapter on how money is affecting the practice of medicine, Sanders talked about a testosterone salesman who went from trying to wine and dine her to trying to bully her, saying "I need you to write three prescriptions a week for the next month." Completely disregarding what her patients might need. Yes, testosterone levels decline with age; that does not mean that most older men should be getting testosterone supplements. (This is one where the "normal" levels are probably useless, because of the amount of variation not just between men, but in a normal man's testosterone level over a 24-hour period.)
If you're going to read only one of these, I'd say start with Sanders.
[I have a large backlog of book posts I may never write; this one is prompted by the need to return the Groopman book to the library, and the Sanders went back a month ago.]