E111: The Intellectual Domain of Risks

Kirtly: How do you know that something is risky? Is it your tummy that has butterflies? Does your heart feel funny or skip a beat? Of course, that's all part of how we're wired as humans. If you haven't listened to our Physical Domain of Risk, you might want to. But when you talk to your financial advisor, or to your doctor, risk is often given in terms of a number.

We are working our way through the 7 Domains of Risk. In this episode, we'll be talking about the Intellectual Domain of Risk. Of course, we could talk about risky intellectual endeavors, weird research, whatever, but we are going to talk mostly about numbers that are used in your healthcare.

I'm Dr. Kirtly Jones from obstetrics and gynecology at University of Utah Health. I'm a reproductive endocrinologist and had a clinical career in family planning, infertility treatment, and advanced reproductive technology and menopause. There are lots of numbers in there, and I spent a lot of time with medical students trying to help them understand the clinical numbers of risk.

With me is my co-host, Dr. Katie Ward. She has a doctorate in nursing practice and a PhD in anthropology. She knows a boatload about statistics. Now, the listener should be asking, "How big is that boat? How full or how heavy is that load? And what statistical methods are the weightiest?'

Katie: Or is the boat named Titanic? And who is the captain? Seriously, the Titanic, honestly, is sort of a good metaphor for this because the passengers did not know what questions to ask before they boarded, like, "How many lifeboats?" And the captain and the cruise line had incentives that they were not sharing. So we're going to spread that metaphor.

Kirtly: Okay, thanks. Thanks a lot, Katie. So when we think about risk in medicine, we don't usually use the Titanic as a metaphor, but we often use the word "chance." And the risk and chance often have emotional weight, just like the Titanic made you have a sinking feeling in your tummy.

Anyway, we use emotions. There are emotional responses in their numbers, but we're going to start with numbers. Let's start with rolling dice. Maybe you're playing Yahtzee or some dice game, and you roll one die and you want to get a three. The chances of rolling three, unless the dice are rigged, is one in six. There are six sides to the dice. The roll is random and the chance is 1 in 6, or 16.666%, and the 6s would keep going on.

People mostly understand one in six. They can imagine six. They can probably imagine 100, so percents are something most people can get, sort of. Percents are the amounts out of 100. But if I tell you the chances of getting 3 on your dice is 16.66666%, you're going to give me a look. So maybe percents aren't for everybody.

But many studies have shown that we aren't very good at imagining 1,000. Can you hold 1,000 people in your mind? How about 10,000? People aren't so good at thinking about chances when we get over 100. Of course, some people are very good, but most are not.

Katie: This is a feature of how our brains evolved, right? So for most of human history, any risk we needed to pay attention to came with a face attached to it, human or animal or otherwise. But we were really concerned with, "Is this person trustworthy? Is that person taking more than their fair share?"

We're really good at social reasoning. We can tell if someone's cheating us or read their intentions or weighing fairness. But we didn't really evolve to read a table of probabilities or even understand dice.

There's actually a famous cognitive psychology experiment called the Wason Selection Task, and it's kind of fun. People are given an abstract thing, like a deck of cards, and there's a rule, something like if a card has a vowel on one side, then it's going to have an even number on the other side. Most people, having to look at this line of cards and guess what's on the other side, get it wrong, even if they're trying.

But if you take that same logic structure and give it a human face, so if the person is drinking alcohol, they must be over 21 and you're the person responsible for checking IDs, people can sort out what the rule is. Same logic, but we're better at doing it if it's faces.

That's what our brains evolved to do. And I think that the challenge for us, in both health communication or as consumers of healthcare, is to personalize that so we can picture it.

So if you say a room full of 1,000 women, you're right, people can't really imagine that. But if you say, "Picture a room with 100 women in it," then it gets a little bit more personal.

I think we're getting better at this in medicine. I've seen people pull out charts with 100 or 1,000 little stick figures. I think they still need a face, though.

Kirtly: Yeah, and then there are some risks that are so small. Let's go back to the year 2005 or thereabouts. Before that, about 40% of postmenopausal women were taking some kind of hormone replacement, 40 women in that room out of 100 ladies.

The Women's Health Initiative tried to figure out the risks and benefits of menopausal hormone therapy. They randomized women to hormones or a pill with no hormones. They didn't let the women or the doctors know if they were taking hormones. This is called a double-blind placebo-controlled randomized trial. It's kind of the gold standard for how we know whether something is risky or not.

When the study was finished, newspapers reported that there was an increased risk of breast cancer and blood clots in women who took hormones.

Just stop there. "What kind of increased risk?" you should ask. A big increased risk? A small increased risk? And who gets to say what's big or small?

Some newspapers reported that the increase in breast cancer was 20%. Many women and doctors, I'm afraid to say, interpreted that as meaning that the risk of breast cancer was 20%. That means if you took hormones, 20 out of 100, check your division, 1 in 5 women would get breast cancer if they took hormones.

Women stopped taking their hormones and doctors stopped prescribing them. After the reports from the Women's Health Initiative, the percent of postmenopausal women taking hormone therapy was 5%, 5 in 100, down from 40 in 100.

At the time, I was very proud of "The New York Times," read by a lot of my patients. I, of course, had the smartest patients on the planet who actually reported the numbers correctly. And when you hear that there's a 20% increased risk of something, you're supposed to ask, "Twenty percent of what?" or, "Compared to what?"

My patients came in wanting to stay on their hormones. They knew that the usable number to understand risk for breast cancer in women who took hormone therapy was one extra breast cancer per 1,000 women per year. That's the real number. One extra breast cancer per 1,000 women per year.

It actually was bigger than that. I think it was 1.2. So I rounded it. I always tried to use round numbers when I talked to patients because studies show that most people do not like fractions and they don't like percents and they don't like parts of numbers like 1.2, so I round it.

Katie: Let's talk about another study that has a 20% risk. There are two concepts here I think that are important. One is the fear of cancer, right? Even if the risk is a 20% increased risk, the fact that it's something you're afraid of, that's compelling in one direction. But if somebody tells you that a drug has a 20% benefit, that's also compelling.

Now, we tend to be more loss-averse, so the 20% risk of something bad happening weighs heavier than the 20% chance of something good happening, which does not explain gambling, but we'll come back to that.

But I'm going to confess when I did something like the doctors did after the WHI, because, again, I think that this illustrates just being human.

About three years ago, I was at a menopause meeting, and this was right when the GLP-1 medications like semaglutide and tirzepatide were just hitting the market for weight loss. They'd been available for diabetes, but they were starting to come out for weight loss. And there was this famous trial kind of like the WHI called the Select Trial.

The conference was buzzing about it. Everyone was talking about it. The speakers were excited, and the drug reps were excited. The people were excited. And so I was hearing it everywhere that these GLP-1s reduced cardiovascular risk by 20%.

I came home and I sent you the study, Kirtly, and you sent me back the back-of-the-napkin math and said, "I'm not impressed."

Kirtly: That's me. I rain showers on everybody's good stories. Oh, dear.

Katie: I had some egg on my face, because you know what? I hadn't actually done the math. This is similar to the WHI situation where this was a trial that did technically show a 20% reduction in major cardiovascular disease effects. But the actual numbers were that it was 6.5%. The rate of these bad effects, like a heart attack or death, were 6.5% in the people taking the medication versus 8% in the placebo group.

The difference is, given the negative effects only happen in 8% of people, you can drop it. That's not nothing, and it's certainly important if you're at risk for cardiovascular disease, but that's very different than, "Everybody's going to get a 20% reduction in their risk." And so it's all in how you present it. I was glad that you reminded me that there's hype, and then there's math.

Kirtly: There's hype, and then there's math. I think it's hard for people because there's now . . . Even a recipe is like, "This is the recipe. I can't stop making it every week." There's so much hype about everything these days.

Well, there are a couple of ways to talk about risk in medicine that I think are useful to patients, and one is numbers needed to treat. And that's not a phrase that people may get used to, but that's a useful concept. How many patients would have to do something to avoid one problem?

How many people would have to be treated for hypertension? Which is usually asymptomatic so people don't feel bad, and sometimes the medicine for hypertension makes them feel bad, so there better be a good reason.

Or my favorite is mammograms. Getting your breast squished is uncomfortable. Ladies, you know that. Guys, you could imagine a sensitive body part squished between two plates. And scheduling a mammogram and waiting for the answer can make some ladies anxious.

How many women have to have a mammogram to save one life? And the answer is 1,000 women have to be screened for mammograms to save one life from dying of breast cancer.

Now, of course, there are women at high risk for breast cancer for whom the numbers needed to screen to save a life would be less. I don't want real numbers to scare anybody from getting screened. But women actually sometimes overestimate how good mammograms are.

The other way of showing risk to patients is my favorite. It's called attributable risk. This is rarely used in talking to patients, but it's often the most understandable.

So there's a 30% reduction in death from breast cancer in women who get regular mammograms. And some studies suggest that reduction might be 40%, but I'll stick with one study. So 30%, you should be asking, "Thirty percent of what? What's the risk of dying?"

Well, the 25-year risk of dying from breast cancer is about 10 per 1,000 women who never get mammograms. So that room that's already crowded with 1,000 women in it and they never get mammograms, 10 of them will die from breast cancer.

Now, you take that 1,000-person room and every one of those get regular mammograms, 7 will die of breast cancer.

The attributable risk of dying from breast cancer, if you never get a mammogram, is 3 per 1,000 women over 25 years.

Does this make a difference in terms of you thinking about mammograms? People get mammograms based on their fear of their . . . maybe their sister got breast cancer, so that's a good reason to get screening because there is a genetic risk.

Maybe their neighbor got breast cancer. Well, that's not inherited and it's your neighbor, but the fear that you have is pretty powerful and that may drive you to get mammograms. But the numbers suggest that it's a good screening tool sort of and it's a good thing to do, but it's not a huge effect.

Katie: Yeah. And so let's go back to the GLPs too, because that's all the rage at the moment too.

Kirtly: Oh, yeah. One in six Americans . . . There's a room of six people and one of them will have taken this drug. It's very high.

Katie: I did do the math finally, and as we talked about, while it was technically a 20% risk reduction, it also was not that big. Thinking about the numbers needed to treat, you have to treat 65 people on these medications, which cost about $1,000 a month, and the trial looked at people for 36 months.

If you're an insurance company and you have to put 67 people on a drug that's $1,000 a month for 36 months to save one heart attack or death, it starts to make sense why they're not covering it.

So I think that factors into both the mammograms and whether or not medications immediately get picked up by insurance. What is the cost of doing that?

I think with mammograms, the cost may be to save those three additional deaths. The cost might be a little bit more in line than putting 66 people who didn't need a GLP-1 on it. So people are making those decisions as well.

Kirtly: So let's talk about how you talk to your surgeon, not your influencer, because you can't talk to your influencer. They just send stuff out. So you sign a paper that says you have a risk of infection, hemorrhage, and death. That's usually the bad stuff. There's a piece of paper when you go in for a procedure and that's usually infection, hemorrhage, death, and a few other things if you have a certain surgery.

It's important for you to ask your surgeon, if you want to, how much of that risk. It's okay to say, "Out of 100 surgeries of this kind, how many patients have a hemorrhage or how many have an infection? How many out of 100?" Now you have more informed consent.

If there's a particular complication that your surgeon talks about, and they say the risk of that complication is low, you don't know what low means to him, and he doesn't know what low means to you. He may think 5 in 100 is a low risk. You may think five in a million means low risk. You should be talking real numbers.

And of course, your ability to assess risk is completely personal. You may have a huge aversion to numbers. You may be so scared you don't even want to hear about the risks. There's often a box in the consent form that says, "I don't want to hear about the risk," and you can check that box.

It could be that your aunt died from complications of hip replacement, and you are completely terrified, and you know in your heart that the risks are so much greater because the only person you knew who had hip replacement was your aunt. So in your own experience, the risk of dying is 100%.

It's your job and your surgeon's job to talk about what you think the risks are and what your fears are. This may be your surgeon's 1,000th hip replacement, but hopefully it's your one and only.

Katie: I think the decisions that we ask people to make with knowledge that's hard to understand are challenging right now. I think all of us know someone who's had cancer, and especially if it's someone we love, like you were saying. Our own assessment is skewed by that, and that's only natural.

I work in menopause care, and I regularly see patients who carry one of several genes that significantly increase their risk of breast or ovarian cancer. And many of them know that they have that gene because they've watched a family member suffer through cancer or maybe even die from it. So then those numbers are really personal and visceral.

Now that we know about absolute risks and understand them a little bit, I want to talk about that a little just because I see this in my practice. So in the general population . . . You were talking before about how many people die from breast cancer. More people than that get it.

So 13 out of 100 women will develop breast cancer in their lifetime. That doesn't mean that they're going to die from it, but we do diagnose a fair amount of it. And the thing is, we're good at treating it now, right?

But for people who carry one of these gene mutations, that number increases significantly. Just to keep it simple, I'm going to talk about just one gene mutation, the BRCA1 gene. So that raises people's risk closer to 70 in 100 for breast cancer, and 40 in 100 for ovarian cancer.

Women, in this day and age, who carry that gene are offered risk-reducing surgery. And so that means contemplating do you remove your breast tissue or your ovaries when you're finished having children?

I'm going to focus on ovary removal for a second because I think that's a really hard choice for people. Removing your ovaries means you get instant surgical menopause. And no matter what we do, we can't fully replicate what your hormones were doing naturally for young women. So that surgical menopause can be pretty miserable: hot flashes, sleep disruption, sexual dysfunction.

And then beyond that, going without estrogen when you're young carries long-term risks of cardiovascular disease, osteoporosis, and even dementia.

Those additional conditions affect both the quality and the length of your life, just like the cancer would have. So the decision, I think, is really hard for people.

It's roughly you're weighing a 40-in-100 lifetime risk of ovarian cancer, which is also a 60-in-100 chance of never getting it, against a 100-in-100 certainty of surgical menopause and everything that comes with it. And I think that's a really hard trade-off for people.

For many of my patients, and I want to be clear about this, removing their ovaries is the right decision at some point. It just feels extra urgent if you've seen a loved one suffer through cancer and you're afraid.

I think in our community, we see people who are finished with their childbearing very young, so they're making this decision at a relatively young age. And oncologists will tell patients about the cancer prevention, but maybe not discuss the menopause aspect.

I just think these decisions are hard. There's no right or wrong decision, just what's right for you. But I think that this is where understanding your risks can really help, and talking to all the specialists that are going to be involved in your care as well.

Kirtly: Oh, I think so. Well, I want to segue a little bit to vaccines. And I get a vaccine for several reasons. Let's talk about the flu vaccine. So I get a flu vaccine, first, to decrease my chance of getting the flu. The risks to me of getting a flu vaccine are small, and now you should be asking, "What do you mean by small?"

Well, there are risks, which is the risk of something big, meaning dying from an allergic reaction or getting some weird neurologic diseases, one in a million. And we have great data on that because millions and millions of people get vaccinated. We know the data.

Now, there are side effects, meaning my arm is going to get sore and I might feel a little feverish for a day, or I might feel a little achy.

Anyway, the reason I get the vaccine for flu is because I don't want to get the flu. But secondly and more importantly for me as a healthcare provider, I don't want to give anyone else the flu.

The University Hospital requires employees to get vaccinated against the flu not just to decrease the risk that employees would get sick, but because we're taking care of the most vulnerable people dying from the flu: pregnant women, old people, people with decreased immune systems. We get the flu vaccine at the hospital so we won't bring this virus to vulnerable people who get very, very sick.

It's big huge thing in the country now about the measles vaccine. I gave my kid the measles vaccines partly so he wouldn't get measles and I might lose a few days of work, but I gave it so that he didn't give measles and give everybody, young person, young kid the measles, or God forbid a pregnant woman the measles.

Vaccines now have become so contentious, and I think that the data, the real numbers, are powerful and powerfully convincing, at least to me.

Katie: Yeah, they are to me too, and it's just been such an interesting thing to see evolve where vaccines are becoming controversial. And I think that's in part because we haven't seen the measles for . . . This current epidemic of measles is something we haven't seen in a long time, and so I think that's allowed us to pretend that it's not real or not out there.

I was just thinking back to the peak of the pandemic when we first got the COVID vaccine, and that was controversial. I remember bumping into a friend of mine, a colleague actually, and she was not enthusiastic about the vaccine and she suggested that we engage in a dialogue about vaccine research.

She said she was retired and she had time to do some research. And I was working pretty hard and maybe not researching that much, and I thought, "Well, great. This will be interesting, and we'll share learned information."

And so I went to Google Scholar and read some research studies that I thought were interesting. She sent me YouTube videos one after another after another. I had to write back and say, "We're not doing the same kind of research. In fact, you're not doing research. You're down a rabbit hole. You're not finding information. It's finding you." That really just has framed a phenomenon that I keep watching.

I think, Kirtly, you and I have spent our careers learning how to read a study, not just the abstract but the table on Page 9 that actually gives you the information to calculate the absolute risk reduction. And if I forget to dig that deep and I'm confessing that, you call me on it.

But we look at all those other things, like who got into the clinical trial and whether the population looks anything like our own patients and what journal it got published in and how rigorously it was reviewed. Was it even on people or was it just on animals? That makes a difference.

I know I've changed my mind when new data warrants it. That's what evaluating evidence actually looks like. It's slow and it's boring and it doesn't make good YouTube video content.

But I think what social media has really figured out is, again, back to how our brains are wired. We are wired to treat repeated information as though it's true.

Again, our ancestors, if they always saw a lion at a particular watering hole, that pattern was really important. If you saw one yesterday, that is way more urgent than what happened last year. So these are features of our brain that kept our ancestors alive, and social media has really weaponized them. So you see the same information over and over again and that makes it feel really important. Or if you've just seen it recently, that also feels more important than the thing you learned a year ago.

And social media is good at even . . . They're sort of understanding where you are in space and whether or not you're near a doctor's office, and it's feeding you information just in time to influence you in that moment. I think those are patterns about how social media works that it's really important to recognize.

And then there's another thing that just is super insidious that I want to mention while we're talking about risks, because I think the real risk to people is that they're getting bad information with an agenda.

One of the things I see over and over again on social media is a person that's got doctor in their handle, and they're standing in front of some just barely legible research paper. It's on a green screen behind them. And they're pointing at you and they're speaking very fluently and expertly about physiology and words you don't quite understand, and they're telling you, "Your doctor isn't telling you this."

I've started, honestly, taking screenshots of these when they come on my social media and I go look up the article now, and it often doesn't say exactly what the video claims.

But what I know is happening, because I see it coming into my clinic every week, is that people are really being influenced by this and they're being convinced . . . They see this whole fancy setup on a social media feed, and they've come to my office because they feel like they really need the supplement that was being sold at the end.

I just think that this whole social media ecosystem is a real risk in that people end up worried about things they can't control. Your guts do leak. It's how food gets across. Cortisol fluctuates, but it's not something you need to measure. And your brains and your bones are not eating themselves in menopause, but they are deteriorating as you age. But you don't need to be afraid of Zone 2 exercise.

These are all things my patients come in really worried about because they've seen it a lot, or they've seen it recently on their social media. And that worry is the thing that's really not good for you. The fear is costing people real money. All of it is actually using up healthcare resources that makes the system more expensive for everybody and makes it less available for people in need.

Kirtly: Oh, the wanting expensive blood tests for this, that. We could go on. Oh my gosh, it just looks like . . . This is very expensive. What happens if you have to pay for it out of your pocket?

I mean, this test that you want, it's not really a viable test for anything that's meaningful. So then they might go off and get the blood test from another non-organization that I don't really understand, and then they want me to interpret the data, and I say, "There is no data. I can't really tell you. There are no numbers here."

Katie: It happens all the time. And I think the thing that is most sort of intentional and worrisome to me is this phrase that "these are things your doctor won't tell you." And what that does is it's eroding trust in the scientific community. It is a risk, but it's an agenda to do that thing exactly, to seed doubt about your healthcare provider and the person that's trying to communicate risks. It's worrisome to me, Kirtly.

Kirtly: Yeah. Well, to round up . . . and not numbers now. We're going to finish our talk about the Intellectual Domain of Risk. Risks are often difficult to talk about. In some cultures, you can't even mention a bad thing happening, even if you're just talking about the 1-in-1,000 chance of that thing happening. To talk about bad things calls up that bad thing and makes it happen.

You need to let your clinician know what you need to know. If you don't understand the numbers, then ask them to explain in more simple numbers, and ask them if they can put it in a chance per 100. If that's too many people in the room, how about 1 in 10?

And I hope that the chances that you've listened to our Intellectual Domain of Risks, all about numbers, didn't turn you off numbers. It's our job in medicine to give you numbers you can use and let you get to choose what kind of numbers these are.

Thanks for listening to our 7 Domains of Risk. Check out our other domains. This is the only one in which we're using numbers.

Katie: Except the number seven, "7 Domains."

Kirtly: Oh, sorry. Sorry about that. Yes.

Well, you can get all of your 7 Domains of Risk and all of our "7 Domains of Women's Health" podcasts wherever you get your podcasts. Enter "7 Domains of Women's Health," or at womens7.com. And it isn't too risky. Thanks for being here.