Hacker News Reader

Comment by olavk | Hacker News Reader

This is really fascinating (and troubling) statistics, but I can imagine what the explanations can be? Some of the possible explanations suggested in the article does not seem to correspond to the result. For example, if female doctors are better in general, then it should also affect the survival rate of male patients. If current medical research favors male patients, then it should affect the survival rate regardless of the gender of the doctor. If you are better at treating your own gender, then it should also affect women doctors treating men.

View on HN

The article does provide a explanation that would seem to correspond to the result.

> Still, she adds, the study raises many troubling questions about the treatment of women in the ER, “like the concern there’s a systematic bias where male physicians are not listening to female patients’ complaints as readily as [those of] a man.”

Simply put, if social conditioning causes men to take the concerns of male patients more seriously than female patients (OR inversely, social conditioning makes female patients less likely to share potentially relevant information with male doctors in the first place), it would explain the difference pretty easily.

Personally, I always prefer to have a female doctor when given the option. My theory is that even if only 0.1% of male doctors harbor a bias against female patients, and even if it's only a mild bias in most cases, I can avoid that chance pretty much completely by going to a female doctor. I guess it's possible for female doctors to be sexist against women, but it seems very unlikely given that most female doctors themselves would have experienced sexism against women at some point and thus would be more aware of it and more invested in overcoming it.

Note, also from the article:

"In the new study everyone was more likely to survive if they saw a female physician, and a study(1) published last year in JAMA Internal Medicine indicated all patients of female physicians had lower mortality and hospital readmission rates. “It seems that the female doctors practice in a better way or outperform male doctors,” says the JAMA study’s first author, Yusuke Tsugawa, an assistant professor of medicine at the David Geffen School of Medicine at University of California, Los Angeles."

1) https://jamanetwork.com/journals/jamainternalmedicine/fullar...

That directly contradicts this:

> The researchers divided 500,000-plus cases into four categories: male doctors treating men; male doctors treating women; female doctors treating men; and female doctors treating women. “All of those are statistically indistinguishable except for male doctor–female patient,” says Brad Greenwood, an author on the study and a data scientist at the University of Minnesota.

> That directly contradicts this

No, it just means in this study's data, there was no statistically significant signal that there was a difference between the groups treated by a male doctor vs a female doctor.

That doesn't mean there's not a difference in outcomes, just that one isn't suitably demonstrated by this data, but it's also a result over a completely different subset of the population than was looked at by the study in the GP's link (which was of hospitalized Medicare patients age 65 or over).

Actually if you look at their full sample data you'll see that women physicians had higher survival rates for both make and female patients. It's just that only female patients were statistically different from zero. And note that just because female-female was statistically significant and female-male was not, that doesn't by itself imply that female-female was statically distinguishable from female-male. (Though it may have been, I don't understand all the controls.)

I also wonder if it might be because men are trained to complain less and "man up". But the whole concept of "hysteria" has a long medical history so we certainly can't ignore that very possible reality.

Actually. I think it's the opposite in this case. That is, these are heart attacks that end up in __ER__ not all HA's. The bias is that women don't have HT's so __if__ they end up in ER then it's likely to be bad.

Can you explain what you mean by hysteria and provide some salient evidence, please?

Sure. "Hysteria" being a nonsense medical diagnosis from the past that's basically the poster child for sexism in medicine (https://en.wikipedia.org/wiki/Female_hysteria).

I bring it up because I'm suggesting that one possible explanation is that maybe this study is revealing yet remaining sexism in medicine. Just like the whole nonsense concept of hysteria, maybe professionals still discriminate by assuming women are just being "hysterical". Not consciously, but more as an artifact of western medicine and society.

I also strongly reccommend this fantastic podcast about the topic: http://www.maximumfun.org/sawbones/sawbones-hysteria

I hope this satisfies your question. I suspect I was just not clear enough that I was not legitimately suggesting hysteria as an actual explanation.

Thank you, this podcast episode looks worthwhile. I did misread your intent, I’m afraid; otherwise the history of hysteria is familiar to me.

But shouldn't this effect be across all health issues then, not just heart attacks?

Yes, and I have strong feeling it probably is. It's a hard statistic to measure in other fields though. Fatalities due to heart failure are pretty open and shut. For other conditions, especially in fields such as psychiatry, the effect of doctors not taking the symptoms of women seriously is more likely to manifest in the form of misdiagnoses, which can't be easily detected or tracked. Furthermore, many criteria for diagnosing diseases were developed based on primarily male subjects, so some conditions (such as autism-spectrum disorders) are somewhat underdiagnosed in women due to confusion over presentation in female patients. That isn't necessarily the doctor's fault, but the outcome can make it hard to distinguish whether the fault is with a doctor who dismissed a patient's symptoms due to sex, or whether it's just insufficient diagnostic criteria.

as another point of data:

https://www.health.harvard.edu/blog/women-and-pain-dispariti...

"Simply put, if social conditioning causes men to take the concerns of male patients more seriously than female patients (OR inversely, social conditioning makes female patients less likely to share potentially relevant information with male doctors in the first place), it would explain the difference pretty easily."

But given the opprobrium currently attached to labeling someone "SEXIST" in our society, personally I'd like to see a bit more evidence, and some alternate explanations explored and ruled out, before we start wheeling that one out and tarring people with it.

The other problem with that is more practical... if it "truly is" sexism, then you also foreclose any chance of fixing it, because there's just no way you're going to go to doctors and make them precisely 10% less sexist than they are now. But if it is in fact something else, it may be fixable. (Or illusory.)

Lots of people have biases that they don't realize can impact the quality of their interactions. There's a frequent issue where people who speak English as their second language are more likely to be perceived as less intelligent. Certain thick regional accents also can cause that issue.

Part of the educational process for all professionals, especially doctors, includes learning to monitor for potential prejudices like that and taking steps to ensure it does not impact your ability to deliver high quality care (ie, ask more questions or have a colleague review your work if you are unsure).

I'm not tar and feathering people. But when women in our society are constantly saying "Some men don't take me as seriously as my male counterparts despite my equal (or superior) credentials" then maybe there is some truth to that which can bleed over professional interactions in hospitals as well? I've experienced it first hand myself professionally (never been to the ER) only a handful of times, but I don't think it's the norm, and I don't go into professional situations with men assuming it will happen. Still, when it happens, it sucks and I wish other men would talk to those guys and tell them why it sucks so those guys will do better.

We've gone a long way towards making sexism less of a problem for everyday women. Just because there are a few examples of it that have slipped by doesn't mean we give up. We just know more of what to watch out for.

Isn't he also trying to say that perhaps the bias that some women have against male doctors might also contribute to the statistic? In other words, women heart attack patients might be more stressed or worried with male doctors leading to higher mortality.

You yourself admit to being biased against male doctors.

"We've gone a long way towards making sexism less of a problem for everyday women."

I couldn't have asked for a better example of precisely what I was talking about. We are so primed to jump down people's throat because of SEXISM that by the time you reached the end of your post, you've already slipped into the assumption that it's the only possible explanation, and are lecturing the doctors in question about how sexist they are and how they should stop it.

It is counter to the principles of science to be so sure, though. It's not that large of an effect, even with the population size, and there's plenty of room for some other systematic error to be the problem.

But you've been programmed to assume that if I'm bringing up the possibility that a scientific paper may have a second explanation, then politically I must also be denying the existence of sexism in the universe and must therefore be downvoted and also lectured to about how important it is that we not be sexist.

But I reject your attempt to exert moral superiority, for the exact reasons I gave in the original post. What's important is not shutting down sexism and scoring moral preening points for having Right Opinions. What's important is finding the real cause, so we can address it. Science is filled unto overflowing with examples and inductive proof that when people go into something with a predetermined conclusion about the causes of some phenomenon, they are often wrong, and end up doing great harm because of it. If we have to choose between getting to yell at people because Sexism, or actually fixing the real problem, which is it going to be?

Unfortunately, as your own message demonstrates, which I have no reason to believe is particularly exceptional, I know the answer. In 2018, it's getting to yell at people Because Sexism, and to hell with whatever the real reason may be. We already know the real reason, and it's Sexism. Because in 2018, even if the real reason isn't sexism, the real reason is still sexism, you just must not be looking hard enough.

I'm flattered by your interpretation of my comment as one which allows me to "exert moral superiority". I have made no judgements against your morals, only against the merits of your arguments. Now let me correct your other misconceptions.

>> "We've gone a long way towards making sexism less of a problem for everyday women."

> I couldn't have asked for a better example of precisely what I was talking about. We are so primed to jump down people's throat because of SEXISM that by the time you reached the end of your post, you've already slipped into the assumption that it's the only possible explanation, and are lecturing the doctors in question about how sexist they are and how they should stop it.

You misunderstand context. That last line was there to convey the sentiment that society HAS successfully addressed sexism in the past. It was a direct refutation of your claim:

> if it "truly is" sexism, then you also foreclose any chance of fixing it...

> But you've been programmed to assume that if I'm bringing up the possibility that a scientific paper may have a second explanation, then politically I must also be denying the existence of sexism in the universe and must therefore be downvoted and also lectured to about how important it is that we not be sexist.

1. Calling people "programmed" for disagreeing with you is rude, and hacker news readers tend to prefer polite discourse - this may have affected your votes.

2. You did not propose a second explanation at all. You simply said that the one I proposed should not even be investigated until every other alternative is ruled out. How can you claim to care about science but not even be willing to investigate a hypothesis that is supported by data, just because you don't like the idea that it relates to a political and social topic? Take your anger out on the people tar-and-feathering, not the people trying to figure out how to stop women from dying disproportionately due to preventable health issues. If you really have a strong theory on an alternative reason behind this data, then propose it and make your own case.

3. "lectured"? I disagree with your assessment of tone there. I simply re-iterated my argument with context that is more relatable to hacker news readers and shared a relevant experience.

The point is: Even if it's NOT sexism underlying the root cause of this particular issue, you appear logically biased when you dismiss the role it could play for no reason other than "I don't like when people blame sexism for things". Dismissing it out of hand due to a purely emotional appeal makes it seem like your analysis is troubled. It's as short-sighted and illogical as people who blame sexism for every little thing ever. Morality aside, the downvotes you earned were well deserved due to the poor quality of argument and the uncivil tone.

Also, "sexism" doesn't need to have the "s" capitalized.

> But given the opprobrium currently attached to labeling someone "SEXIST" in our society, personally I'd like to see a bit more evidence, and some alternate explanations explored and ruled out, before we start wheeling that one out and tarring people with it.

There's opprobrium attached to being uniquely sexist / otherwise bigoted, but I think there is relatively little opprobrium in saying that most men, statistically, behave in the following way because of social conditioning since childhood, and no man in the relevant population is specifically intending to be sexist. That's where the whole idea of "unconscious bias" comes from. (Arguably, the entire idea of "unconscious bias" training is to provide a way for companies to address these types of issues precisely without having to label their employees sexist and bring out the tar.)

Also, even if it did require tar, I think we would need to have a hard discussion about where the tradeoff between "more patients die, but doctors have good reputations" and "fewer patients die, but people think doctors are sexist" is. It does not seem obvious to me that we should choose the former.

> because there's just no way you're going to go to doctors and make them precisely 10% less sexist than they are now.

I don't see how this follows.

For instance, you can imagine pretty direct ways to work around both potential explanations: if it's social conditioning on the part of the doctor, have patients report pain on an objective scale (i.e., not "1-10", I believe there are studies that women statistically report lower scores than men; instead assign specific comparisons like "stubbing your toe" or "unable to focus on work" or whatever), and ask doctors to respond in standard ways based on the response. Again - assuming that this isn't doctors consciously choosing to ignore women's concerns - we should probably assume doctors are actually interested in giving their patients meaningful care. If the social conditioning is on the side of patients not reporting pain, make sure there are enough women medical professionals around and have a practice of women reporting their symptoms to a woman. (Since nurses are something like 90+% women this shouldn't be too difficult; it would just require asking male doctors to calibrate their responses to nurses' notes instead of asking questions de novo.)

Also, historically, we have seen examples of people becoming less sexist and less bigoted. The idea that women just do not have the mental capabilities to vote, for instance, has become an extreme minority position. It may take generations of change, but it's still a change that we can effect.

Also, even if we're talking about populations, it's not really the case that every person is equally sexist. If you want to make doctors precisely 10% less sexist, measure them for sexism and reject the ones that exhibit abnormally high levels of sexism.

People on HN frequently proclaim that we need self-driving cars now, now, now!! to prevent thousands of deaths caused by car accidents every day—even if it involves "breaking a few eggs" in the short term.

Meanwhile, I feel like those very same people would be very reluctant to "play the sexism card", even if it meant that thousands of women's lives could be saved, and even though this is a problem that has been pointed out and studied for a long time[1], and that it's potentially much easier to correct our biases with better training than it is to approach general AI with technology.

I really wish these trends were reversed.

[1] See Serena Williams' childbirth story for one recent example: https://qz.com/1177004/serena-williamss-terrifying-childbirt...

Life is not a conspiracy against your opinions, be carefull with accusations like that. Do not make up fictional enemies.

I'm guessing it's a combination. Heart attack signs and symptoms are different for men and women. But most of the awareness and training, especially in the past, has been more focused on the typically male symptoms.

Add to that the likely facts that 1) women doctors are, on average, more recently trained than men doctors; 2) women doctors are probably far more likely to be aware of the differences in symptoms between the traditionally discussed symptoms and those that are specific to women, simply from a self-interest perspective; 3) men are likely to be a higher percentage of heart attack patients, so the effect of fewer men being as aware of women-specific symptoms of heart attacks would be amplified by the relative frequency of the occurrences (ie, they see few female heart attacks and so the differences in appropriate treatment are less likely to sink in with someone who doesn't have a personal awareness ("what if this were me having the heart attack?")).

But yes I agree it's too bad these topics were skipped in the article in favor of less likely differences (such as "women are better doctors"--something that I wonder if it's controlled for age, but might just be the effect of the fact that women who do make it through medical school and continue to practice are probably just far more driven and serious about the career than similar men, because the obstacles women face in reaching the same point in their career are much higher, so the women who do make it through will naturally be better doctors than the "average" man doctor).

You're discussing things which are discussed in the abstract of the linked study... which makes it all the more frustrating that these things are omitted from the article.

> Patients treated by female physicians had lower 30-day mortality (adjusted mortality, 11.07% vs 11.49%; adjusted risk difference, –0.43%; 95% CI, –0.57% to –0.28%; P < .001; number needed to treat to prevent 1 death, 233) and lower 30-day readmissions (adjusted readmissions, 15.02% vs 15.57%; adjusted risk difference, –0.55%; 95% CI, –0.71% to –0.39%; P < .001; number needed to treat to prevent 1 readmission, 182) than patients cared for by male physicians, after accounting for potential confounders.

> 2) women doctors are probably far more likely to be aware of the differences in symptoms between the traditionally discussed symptoms and those that are specific to women, simply from a self-interest perspective;

Meanwhile here we are on a forum of randos from the internet, and half the thread is people noting that heart attacks have different symptoms in men and women. So what's the doctor's excuse?

I'm not saying this to argue with you, I mostly agree. The point is if that's the explanation, it's a pretty pitiful one.

Heart attack symptoms differ in men and women. The most common ones in men (chest and left arm pain) are less common in women.

Perhaps female doctors are more likely to be familiar with the symptoms of heart attack in women.

Additionally, I think there's a stereotype of a heart attack being a male affliction (probably because the difference in symptoms led to underdiagnosis in women in the mid-20thC). It's conceivable that male doctors are less likely to consider heart attack as a diagnosis for a woman even when the symptoms are typical of a heart attack?

I don’t have sources but I remember reading that heart issues manifest differently in women than men and that women weren’t taken as seriously when explaining their symptoms (indicative of heart issues) to male docs. I think that would explain why the reverse situation doesn’t arise.

Like this [0] issue, but for heart problems specifically? How pervasive is this...

[0] https://digitalcommons.law.umaryland.edu/cgi/viewcontent.cgi...

Yeah. Here I think is an article I’d read about it: https://www.theatlantic.com/health/archive/2015/10/heart-dis...

Reply on news.ycombinator.com