SGU Episode 598
|
This episode needs: transcription, proofreading, formatting, links, 'Today I Learned' list, categories, segment redirects. Please help out by contributing! |
How to Contribute |
| SGU Episode 598 |
|---|
| December 24th 2016 |
 |
| (brief caption for the episode icon) |
| Skeptical Rogues |
| S: Steven Novella |
B: Bob Novella |
C: Cara Santa Maria |
J: Jay Novella |
E: Evan Bernstein |
| Quote of the Week |
In science it often happens that scientists say, ‘You know that’s a really good argument; my position is mistaken,’ and then they would actually change their minds and you never hear that old view from them again. They really do it. It doesn’t happen as often as it should, because scientists are human and change is sometimes painful. But it happens every day. I cannot recall the last time something like that happened in politics or religion. |
| Links |
| Download Podcast |
| Show Notes |
| Forum Discussion |
Introduction
- 2016 was mostly a bad year, but there were some good things
- Will SGU ever miss a week?
You're listening to the Skeptics' Guide to the Universe, your escape to reality.
Movie Review (3:54)
News Items
Better Outcomes from Female Doctors (30:13)
S: All right, Cara, start out the news items with - this one has been really all over the news in the last couple of days, about female doctors.
C: Yeah, okay, so there is a report that was just published in - do you pronounce it Jama when you're
S: Yeah, JAMA, JAMA
C: You do, okay. In JAMA Internal Medicine, actually. Here's the title. Let's start with that, because you've probably been seeing different headlines in popular news outlets. "Comparison of Hospital Mortality and Readmission Rates for Medicare Patients Treated by Male versus Female Physicians." Probably the headline that you're seeing over and over is, "Female Doctors Are Better For Their Patients," or "Patients Less Likely -
E: Superiority was the
C: Yes
E: word that I saw.
C: Superior to Male Doctors." Patients Less Likely to Die if they have a female doctor. All this kind of stuff. Let's talk about what they actually looked at. They decided to focus on medicare part B patients that were admitted to hospitals. And that was for a lot of different reasons. I think some of them were like, ease of records. But specifically, they did a kind of correction to their data by using these patients that were specifically admitted to hospitals, because they found that there would be less bias in their data.
If you're just going to your general practitioner, you chose that practitioner. So there might be all sorts of bias in whether or not you wanted a female versus a male doctor, and whether or not they wanted to choose you. Were you too sick for them to take on? Do you have other kind of reasons that they wouldn't want to take you on? But at a hospital, it's quite common that you can't pick your doctor. You're just kind of assigned to whoever's available.
So they wanted to make sure that that was kept clean and unbiased in that regard. There were a lot of other things that they did to control for these biases. So right there, they looked at medicare Part B patients in hospitals. And then they looked at two measures. They looked at mortality. Did they die within thirty days. And they looked at readmission. Did they come back within thirty days? They also extended that later in the analysis to sixty days, to see if the effect still held.
And then they looked at basically, was the doctor, the primary treating doctors, the first person that that patient saw, male or female? Admittedly, that was self-report, so they said they weren't able to control for any trans doctors, or any individuals who might have chosen to abstain from saying what their gender was, or might have chosen male versus female, even though that didn't really appropriately describe who they were.
But there you go; that's breaking down the title. So, it was a lot. They looked at a lot, a lot, a lot, of people. So let me find that. Twenty percent random sample ... okay. Beneficiaries sixty-five years or older. They were looking at people who were treated by general internists from January 1st, 2011 to December 31st, 2014. And it was a lot of people. It was one million, five hundred eighty three thousand twenty eight hospitalizations, and the amount of individuals was six hundred twenty-one thousand, four hundred twelve men, and nine hundred sixty-one thousand, six hundred sixteen women.
And that was for mortality. And then somewhat similar numbers for readmission. They also decided that they wanted to kind of parse out a lot of the data, to see if different things factored into whether or not men or women had better outcomes in their patients. And we're talking about male versus female doctors, not male versus female patients.
They wanted to see if, for example, eight common medical conditions, the eight most common that they found in the medicare data, that people actually were admitted to hospitals for, did these different conditions fare better? You know, did female doctors do better on heart patients versus male doctors doing better on kidney patients. Let's look at all of that to see if there's any sort of variability there.
And the outcome basically was that a very, very small, yet significant difference existed across almost every measure. Some of the measures had a small difference that was not significant, but, as reported in the study, all of the measures fared better on the female doctor side, versus the male doctor side. Or, in some cases, there was no difference. They never actually reported any situations, and tell me if I'm wrong, Steve, but I couldn't find any where male doctors fared better with patient outcomes
S: No
C: than male doctors.
S: No
C: Yeah, they never found any. But they did find that most of the measures showed a significant difference. Some of them showed very small, but unsignificant difference. And that significant difference was at a p value of less than .001. So, it was, you know, it was pretty stringent. And after they added all of these different analytic controls, like controlling for the condition, controlling for whether or not this was a hospitalization,
S: Age of the physician.
C: So let's talk about what the actual difference was. This was like a half a percent difference. Actually, slightly less than a half a percent difference. It was a .4 ...
S: .43
C: .45 percent?
S: .43
C: .43 percent. Here's what the write up said. Quote,
"The first question that everyone will ask is whether the size of the effect matters. I'm gonna reiterate what I said above. The effect size is modest, but important. If we take a public health perspective, we see why it's important. Given our results of male physicians had the same outcomes as female physicians, we'd have thirty two thousand fewer deaths in the medicare population. That's about how many people die in motor vehicle accidents every year. Second, imagine a new treatment that lowered thirty day mortality by about half a percentage point for hospitalized patients. Would that treatment get FDA approval for effectiveness? Yup. Would it quickly become adopted in the hospital wards as a treatment we should be giving our patients? Absolutely. So while the effect size is not huge, it's certainly not trivial."
So, I think that the authors effectively make the point that from their analysis, which is only the second of its kind, and the only previous analysis was not for hospitalized patients, and it was done with a way smaller sample size, they think that this is important enough to make sure that individuals know about it. But as we know in the kind of public, or general media, a lot of people are really sort of over-valuing what was said.
And there is some precedence for why they would say this. If you look in the introduction of this study, they actually do note a bunch of previous studies that show some important parameters that they think may be contributing to why female physicians have better outcomes. And one of them, which I really appreciated. It perked by ears up, was that there are previous studies which show that female physicians are more likely to adhere to clinical guidelines, and evidence-based practice.
S: Yeah, which is, in my mind, this is my bias, that's like, the big thing, the big glaring
C: Exactly
S: It could explain the whole effect right there.
C: So it's not about male versus female.
S: Yeah
C: It could. And that's what they were saying. They even said in their study, there's nothing inherent about male or being female that would explain this. So there must be something about the way that female physicians versus male physicians practice medicine that may be contributing. And I think the reason this is important, and the blog post does a better job of explaining this than the actual, you know, clinical study, which is full of a lot of jargon, is that it's at the very end of the blog post, and I think I'll just quote it again, because he says it better than I do.
"Another important point must be addressed: There's pretty strong evidence of substantial gender pay gap, and gender promotion gap within medicine. Several recent studies have found that women physicians are paid less than male physicians, about ten percent less after accounting for all potential confounders; and are less likely to be promoted within academic medical centers."
And they even mention in the study that a lot of female physicians will not be able to get certain positions because of worries about prenancy. You know, a lot of the things we see
S: Yeah
C: see in academia. So that said, I think it's that much more important, this information be out there, even if it didn't say female patients or female doctors fared better, but it said they fared exactly the same, I think it's just another piece of evidence for like, there's no reason female doctors should be paid less, or treated any differently within medical and academic institutions, because they have good outcomes, and that's the most important thing for a patient.
S: Yeah
C: So there you go.
S: So let me put the data into a little bit of perspective, a little bit of further
C: Yep
S: perspective. So yeah, the difference is very small, .43 percent. But whenever you're talking about death, obviously, even the slightest advantage is important, because, when you apply it to the whole society, you end up with numbers that are meaningful, like thirty-two thousand fewer deaths over the course of a year.
C: Yeah, it's funny, because we think of a hundred percent as like, a hundred people. So then we're more like, "Half a person!" But no, you've got to
S: No, it's a half a percent. But the
C: Yeah
S: other things is, these are all statistically significant, because you're dealing with 1.5 million hospitalizations. So that when you have numbers that high in the study, it's a bit of a double edged sword. Obviously, it makes it a very powerful study. But what it also means is that any subtle effect in there, even a subtle confounding effect is going to easily achieve statistical significance,
E: Yeah
S: because the power is so high, of the study. This is a not a randomized or controlled study, so patients are not being randomized to
C: Yeah
S: male versus female doctors, which means, it's impossible to completely eliminate confounding factors in one study. You have to look at this multiple ways, multiple different times to see if it holds up.
C: Which is weird, because they mention that you can't do a study like that.
S: No, you can't.
C: They actually said in the paper you can't.
S: You never will be able to do it. That's why
C: Is it unethical?
S: you have to triangulate as best you can. Well how are you gonna do that? I mean, with any
C: I mean, I guess within a single hospital
S: kind of real numbers. Yeah, but you're not gonna get to these kind
C: Yeah
S: of numbers.
C: That's true.
S: You're not gonna see this kind of effect size with that kind of study.
C: Yeah
E: We need robot doctors too.
S: Yeah
(Cara laughs)
S: So, yeah, to put that number into
B: Yeah, baby!
S: further context, the number needed to treat. So in other words, in order to prevent one death, you would need to have two hundred and thirty-three patients treated by female rather than a male doctor. So that helps it. A number needed to treat it is a good way to put things into perspective. If you look at it from the point of view of an individual patient, 'cause I know, the worry is that patients, the public's gonna see this and like, "Oh, I want a female doctor, not a male doctor."
C: Yeah
S: But yeah, but the probability of that making a difference for you is really slim.
C: Yeah
S: It's not worth worrying about. Also, the final thing to keep in mind is that the difference among male doctors, and among female doctors is orders of magnitude greater than this statistical difference between male and female doctors, right?
C: Um hmm, yeah, so this is just the average
S: Yeah
C: overall of them.
S: Having said all of that, I have absolutely no problem believing that female doctors have better outcomes in the inpatient setting. My own experience is, female doctors in general are fantastic, and especially, they, for some reason, and this, I think, whatever culture, they really operate really well in the inpatient setting. 'Cause there's so many interpersonal relationships that go into managing a patient on the inpatient side. And women, in my - this is totally subjective. It just seems that the female doctors are just so much better at that than the male doctors, on average, right?
C: And there's actually some evidence to support that.
S: Yeah
C: Because they show in the introduction, you know, often in introductions, it's a big literature review. So they look through other studies. And they show that literature has shown female physicians are more likely to adhere to clinical guidelines, provide preventive care more often, use more patient-centered communication, and perform as well or better on standardized examinations, and also, provide more psycho-social counselling to their patients than do their
S: Yeah
C: male peers. There's also studies, just of patient satisfaction with their physicians, and it shows that generally, female physicians score higher on patient satisfaction than the male physicians.
S: That's all interesting, and I think that with that, the most useful thing about this is, what can we fix? What could we improve.
C: Yeah, and what can we learn from the way, yeah.
S: If the adherence to guidelines, I mean, there's other independent evidence which shows that adherence to evidence-based guidelines leads to better outcomes. That's a little bit circular, 'cause, you know
C: Yeah
S: evidence shows that evidence is helpful. Yeah, okay.
(Cara laughs)
S: But, you know, I think it's totally legitimate. So I think, if, for whatever reason, women are adhering to guidelines more than their male counterparts, that may explain this. I just think, again, just in general, my observation is that women communicate on average better.
Now, the caveats the authors put in the study, and I think I have to point out, this is about internists on the inpatient service.
C: Yes
S: That does not necessarily mean that this would be the same for surgeons, or for other specialists. And this does not mean it would translate to the outpatient side. Outpatient practice is very different than
C: Yeah
S: inpatient practice. And the differences may not translate. So I would be very interested to see the similar studies done in other settings and other specialties, to see how it holds up. And that would be a great way to try to isolate, try to triangulate to, what are the critical factors here?
C: Yeah
S: What do they have in common? Or if they're different, what's different? And how does that inform it? So this is
C: Yeah, you could do a factor analysis or something.
S: This is one slice of the evidence. It's very interesting. But this has to be part of, obviously, a lot more studies, many more studies, to try to really get an idea of what this is telling us. But I do think, absolutely, just on this evidence alone, this undercuts any argument that women deserve to be paid ten percent less than their male counterparts.
C: Yeah, or that they shouldn't get promotions
S: Yeah
C: because they're less likely to contibute.
S: Yeah, very interesting.
Wet Ceres (44:04)
Light from Antimatter (48:31)
Purple Food (53:55)
Who's That Noisy (1:00:46)
- Answer to last week: Corn growing, sped up
Interview with Brian Switek (1:04:00)
Science or Fiction (1:16:27)
Item #1: A study found that subjects deliberately infected with hookworm had reduced symptoms of asthma and improved lung function. Item #2: Scientists described a fossil of a 520 million year old arthropod with an exquisitely preserved nervous system, showing the structure of the ventral nerve cord, ganglia, and individual nerves. ttp://www.pnas.org/content/113/11/2988 Item #3: Measures taken to preserve the prairie dog inadvertently precipitated the extinction of the already endangered black-footed ferret, the prairie dog’s main predator.
Announcement (1:31:13)
- NECSS 2017
Skeptical Quote of the Week (1:32:47)
"In science it often happens that scientists say, ‘You know that’s a really good argument; my position is mistaken,’ and then they would actually change their minds and you never hear that old view from them again. They really do it. It doesn’t happen as often as it should, because scientists are human and change is sometimes painful. But it happens every day. I cannot recall the last time something like that happened in politics or religion." - Carl Sagan
2nd Announcement (1:33:53)
- Year End Wrap Up episode. Episode 600 coming up.
S: The Skeptics' Guide to the Universe is produced by SGU Productions, dedicated to promoting science and critical thinking. For more information on this and other episodes, please visit our website at theskepticsguide.org, where you will find the show notes as well as links to our blogs, videos, online forum, and other content. You can send us feedback or questions to info@theskepticsguide.org. Also, please consider supporting the SGU by visiting the store page on our website, where you will find merchandise, premium content, and subscription information. Our listeners are what make SGU possible.
References
|
|
|