Latest / Truth in Motion with Dr. Jocelyn / Who Was Counted?

Transcript

Who Was Counted? · Truth in Motion with Dr. Jocelyn

• 0:09
  SPEAKER_00: Hello to all of my fellow travelers out there.
• 0:12
  SPEAKER_00: I'm glad you're here.
• 0:14
  SPEAKER_00: Wherever you're listening from, thank you for spending time with me today.
• 0:20
  SPEAKER_00: Let's dive right in.
• 0:22
  SPEAKER_00: Yesterday we saw that doubt can be manufactured.
• 0:26
  SPEAKER_00: One industry wrote that doubt was its product, and we learned three questions to carry.
• 0:33
  SPEAKER_00: Who is telling me this?
• 0:35
  SPEAKER_00: Who paid for it?
• 0:36
  SPEAKER_00: And what would they gain if I believed it?
• 0:39
  SPEAKER_00: Today we asked the most basic question behind any number.
• 0:44
  SPEAKER_00: Who was counted?
• 0:46
  SPEAKER_00: In 1936, a popular magazine called the Literary Digest ran the biggest election poll America had ever seen.
• 0:56
  SPEAKER_00: They mailed ballots to 10 million people.
• 0:59
  SPEAKER_00: 2.4 million sent them back.
• 1:02
  SPEAKER_00: 2.4 million answers.
• 1:05
  SPEAKER_00: The result was clear.
• 1:07
  SPEAKER_00: Elf Landon would beat Franklin Roosevelt with about 57% of the vote.
• 1:14
  SPEAKER_00: Roosevelt won with about 62%, one of the biggest landslides in American history.
• 1:22
  SPEAKER_00: How could 2.4 million answers be so wrong?
• 1:27
  SPEAKER_00: The magazine had built its mailing list from phone directories, car registrations, and magazine subscriptions in the middle of the Great Depression.
• 1:38
  SPEAKER_00: People who owned phones and cars and subscribed to magazines were not a fair picture of the country.
• 1:45
  SPEAKER_00: And the people who chose to mail back a ballot weren't either.
• 1:49
  SPEAKER_00: Meanwhile, a young pollster named George Gallup took a much smaller survey.
• 1:55
  SPEAKER_00: Far fewer people, but he worked to choose a cross-section that looked like the country, and he called the election correctly.
• 2:04
  SPEAKER_00: A huge number, the wrong people, a small number, the right people.
• 2:10
  SPEAKER_00: That's today's lesson in two sentences.
• 2:13
  SPEAKER_00: Why does this matter?
• 2:15
  SPEAKER_00: Because every health claim we hear is based on some group of people.
• 2:20
  SPEAKER_00: Every dose, every guideline, every study show, if that group looks like us, the answer may fit us.
• 2:29
  SPEAKER_00: If it doesn't, the answer may not.
• 2:31
  SPEAKER_00: And most of the time, no one tells us who was in the room.
• 2:36
  SPEAKER_00: Many of us have had the experience of a treatment that was supposed to work and didn't, or a dose that was supposed to be right and wasn't.
• 2:45
  SPEAKER_00: Sometimes the reason is simple.
• 2:47
  SPEAKER_00: People like us were never counted.
• 2:50
  SPEAKER_00: Here's where I stand.
• 2:52
  SPEAKER_00: I believe medicine tends to be designed around whoever was counted.
• 2:57
  SPEAKER_00: And for a long time, the people counted were a narrow slice of humanity.
• 3:02
  SPEAKER_00: When you weren't in the study, you're being asked to trust an answer that was never tested on you.
• 3:09
  SPEAKER_00: That isn't a reason to reject research.
• 3:11
  SPEAKER_00: It's a reason to ask every single time who was in it.
• 3:16
  SPEAKER_00: And I'll say something else.
• 3:18
  SPEAKER_00: Size impresses us.
• 3:20
  SPEAKER_00: A study of 20,000 people?
• 3:23
  SPEAKER_00: Millions of users?
• 3:25
  SPEAKER_00: But size never fixes a sample that's tilted.
• 3:29
  SPEAKER_00: The literary digest proved that.
• 3:32
  SPEAKER_00: More of the wrong people just makes the wrong answer louder.
• 3:37
  SPEAKER_00: In 1981, researchers invited male doctors across the United States to join a study on aspirin and the heart.
• 3:46
  SPEAKER_00: More than 22,000 men took part.
• 3:49
  SPEAKER_00: It was called the Physician's Health Study.
• 3:52
  SPEAKER_00: In 1989, the results came out.
• 3:55
  SPEAKER_00: Men who took aspirin had 44% fewer first heart attacks.
• 4:01
  SPEAKER_00: The finding was so strong that the aspirin part of the study was stopped years early.
• 4:07
  SPEAKER_00: It was a landmark, and for years it shaped advice about aspirin for heart health.
• 4:14
  SPEAKER_00: But notice who was counted?
• 4:16
  SPEAKER_00: Men, only men.
• 4:19
  SPEAKER_00: It took until 2005 for a large trial to answer the same question in women.
• 4:26
  SPEAKER_00: The women's health study followed nearly 40,000 women over age 45 for about 10 years.
• 4:34
  SPEAKER_00: Half took low dose aspirin and half took a placebo.
• 4:39
  SPEAKER_00: For women, aspirin did not lower the risk of a first heart attack.
• 4:44
  SPEAKER_00: The numbers were almost identical, about 1% in each group.
• 4:50
  SPEAKER_00: It did lower the risk of stroke a little, but the heart attack benefit seen in men simply wasn't there.
• 4:57
  SPEAKER_00: 16 years.
• 4:59
  SPEAKER_00: For 16 years, women were living with an answer that had been found in men.
• 5:05
  SPEAKER_00: And notice one more thing about the men's study.
• 5:08
  SPEAKER_00: Before the main trial began, more than 33,000 doctors started a trial period.
• 5:14
  SPEAKER_00: Only about 22,000 went on to the rest of the real study.
• 5:20
  SPEAKER_00: The trial period was a way of keeping the people most likely to stick with the pills.
• 5:25
  SPEAKER_00: That's a common practice, and there are reasons for it.
• 5:29
  SPEAKER_00: But it means the study counted people who had already shown they could take the pills and who stuck with them.
• 5:38
  SPEAKER_00: When a sleeping pill met a woman's body, here's another story about who was counted.
• 5:44
  SPEAKER_00: For years, a popular sleeping pill called Ambient or Zulpadem was prescribed at the same starting dose for men and women, 10 milligrams.
• 5:56
  SPEAKER_00: Then in January of 2013, the FDA told the makers to cut the recommended dose for women in half, from 10 milligrams to 5.
• 6:07
  SPEAKER_00: Why?
• 6:08
  SPEAKER_00: Because women clear the drug from their bodies more slowly.
• 6:12
  SPEAKER_00: That meant some women were waking up with enough of it still in their blood to impair them the next morning, especially behind the wheel of a car.
• 6:22
  SPEAKER_00: Women had been taking it for about 20 years before the dose was cut in half for them.
• 6:29
  SPEAKER_00: Think about what that means in a real life.
• 6:32
  SPEAKER_00: A woman takes the dose her doctor prescribed exactly as prescribed.
• 6:37
  SPEAKER_00: She sleeps, she gets up, makes breakfast, and drives her kids to school.
• 6:42
  SPEAKER_00: She has no idea that her body is still carrying enough of the drug to slow her reactions.
• 6:48
  SPEAKER_00: She did everything right.
• 6:50
  SPEAKER_00: The dose was simply built on the wrong body.
• 6:55
  SPEAKER_00: And it wasn't an accident that women were missing from so much early research.
• 7:00
  SPEAKER_00: In 1977, the FDA issued guidance that kept women who could become pregnant out of the earliest stages of drug testing.
• 7:09
  SPEAKER_00: That guidance wasn't changed until 1993.
• 7:14
  SPEAKER_00: Sometimes the problem isn't who was counted, it's how they were counted.
• 7:19
  SPEAKER_00: For years, Colgate ran ads saying that 80% of dentists recommend Colgate.
• 7:26
  SPEAKER_00: 80% sounds like a landslide.
• 7:29
  SPEAKER_00: In 2007, the advertising watchdog in the United Kingdom ruled the claim misleading.
• 7:36
  SPEAKER_00: Why?
• 7:37
  SPEAKER_00: The survey had let dentists recommend several toothpaste brands, not just one.
• 7:45
  SPEAKER_00: So a dentist who named Colgate along with three other brands was counted as recommending Colgate.
• 7:51
  SPEAKER_00: The ad made it sound like 80% of dentists chose Colgate over every other brand.
• 7:59
  SPEAKER_00: The survey never showed that.
• 8:01
  SPEAKER_00: Same people, a different question, a very different number.
• 8:05
  SPEAKER_00: And Colgate was told not to use the claim again.
• 8:10
  SPEAKER_00: So here's what we can ask of any number before we let it guide us.
• 8:15
  SPEAKER_00: Who was counted?
• 8:16
  SPEAKER_00: Men, women, children, older people, people with one health problem, or the complicated real life mix most of us carry.
• 8:26
  SPEAKER_00: Who was left out?
• 8:28
  SPEAKER_00: Sometimes on purpose, sometimes by accident?
• 8:31
  SPEAKER_00: The phone directory in 1936?
• 8:34
  SPEAKER_00: Women in early drug trials.
• 8:37
  SPEAKER_00: Often older people and anyone with more than one health condition.
• 8:42
  SPEAKER_00: Who chose to answer?
• 8:44
  SPEAKER_00: People who mail back a ballot or fill out a survey can be very different from the people who don't.
• 8:50
  SPEAKER_00: Think about who answers a survey about a supplement, often the people who loved it and the people who hated it.
• 8:58
  SPEAKER_00: The ones in the middle rarely bother, and a survey that only hears from the delighted will always sound delightful.
• 9:06
  SPEAKER_00: Who dropped out or was weeded out before the study began?
• 9:11
  SPEAKER_00: If the people who had side effects quit halfway through and only the people who finished are counted, the treatment can look gentler than it really is.
• 9:21
  SPEAKER_00: Always ask how many started and how many finished.
• 9:26
  SPEAKER_00: And finally, the one that matters most.
• 9:29
  SPEAKER_00: Are they like me?
• 9:31
  SPEAKER_00: Now I want to be fair, good researchers know all of this.
• 9:35
  SPEAKER_00: And when they count well, it shows.
• 9:38
  SPEAKER_00: Gallup counted well in 1936, the women's health study counted women when it was finally done.
• 9:45
  SPEAKER_00: Counting well means choosing people who look like the people the answer will be used on.
• 9:50
  SPEAKER_00: It means saying plainly who was included and who wasn't.
• 9:55
  SPEAKER_00: And it means not dressing up a narrow sample as if it spoke for everyone.
• 10:00
  SPEAKER_00: So when you find a study that tells you exactly who was in it, men and women, young and old, healthy and not, that's a good sign.
• 10:09
  SPEAKER_00: It's being honest with you about the limits of its answers.
• 10:13
  SPEAKER_00: Why this brings me back to the person in front of me.
• 10:16
  SPEAKER_00: This is why I care so much about observing the person who's actually in front of me.
• 10:21
  SPEAKER_00: In my treatment room, no one gets left out of the sample.
• 10:25
  SPEAKER_00: The woman with three conditions and five medications counts.
• 10:30
  SPEAKER_00: The man in his 80s counts.
• 10:32
  SPEAKER_00: The person whose body doesn't behave like the textbook counts.
• 10:37
  SPEAKER_00: Studies give us averages from groups.
• 10:40
  SPEAKER_00: Careful observation tells us about one real person.
• 10:44
  SPEAKER_00: And we need both.
• 10:46
  SPEAKER_00: But when the study never counted people like you, your own record and a practitioner who pays close attention may be the best evidence you have.
• 10:56
  SPEAKER_00: Last week I invited you to look at your tongue and feel your pulse each morning and write down what you noticed.
• 11:03
  SPEAKER_00: That's a sample of one.
• 11:06
  SPEAKER_00: And it's the one sample that's guaranteed to include you.
• 11:10
  SPEAKER_00: No one weeded you out, no one asked you the question in a tricky way.
• 11:15
  SPEAKER_00: No one chose you because you owned a phone in 1936.
• 11:19
  SPEAKER_00: It won't tell you what's true for everyone.
• 11:22
  SPEAKER_00: That isn't its job.
• 11:23
  SPEAKER_00: Its job is to tell you what's true for you.
• 11:26
  SPEAKER_00: And when you bring that record into the room with a practitioner who knows how to read it, you've added the one person every study left out.
• 11:35
  SPEAKER_00: So the next time we hear studies show, we can ask, in whom?
• 11:40
  SPEAKER_00: The next time a dose is recommended, we can ask whether it was tested in bodies like ours, our sex, our age, our size, our other conditions.
• 11:52
  SPEAKER_00: That can be as simple as asking, was this tested in people like me, women my age, people who take other medicines I take?
• 12:01
  SPEAKER_00: A good practitioner won't be offended by that question.
• 12:04
  SPEAKER_00: A good practitioner will be glad you asked.
• 12:07
  SPEAKER_00: And the next time a number sounds overwhelming, 80%, 2 million answers, we can remember the literary digest.
• 12:17
  SPEAKER_00: A big number can still be counting the wrong people.
• 12:22
  SPEAKER_00: This week we look behind the number.
• 12:24
  SPEAKER_00: Here's today's practice.
• 12:26
  SPEAKER_00: Find one health claim that mentions a study.
• 12:29
  SPEAKER_00: Then go looking for who was in it.
• 12:32
  SPEAKER_00: You can often find it in the first few lines of a study summary or in a news story that describes it.
• 12:38
  SPEAKER_00: Write down three things, their sex, their age, and what health conditions they had or didn't have.
• 12:45
  SPEAKER_00: Then ask yourself honestly, are they like me?
• 12:49
  SPEAKER_00: And if they're not, how much weight should I give the answer?
• 12:54
  SPEAKER_00: Today we saw a poll of 2.4 million people miss an election by a landslide because it counted the wrong people.
• 13:03
  SPEAKER_00: We saw a famous aspirin study count only men and a trial in women 16 years later find a different answer.
• 13:11
  SPEAKER_00: We saw a sleeping pill dose the same for women and men until we learned their bodies handled it differently.
• 13:19
  SPEAKER_00: And we saw 80% of dentists shrink once we learned how the question was really asked.
• 13:26
  SPEAKER_00: And we saw why it matters.
• 13:29
  SPEAKER_00: A number is only as good as the people behind it.
• 13:33
  SPEAKER_00: When we know who was counted, we can decide whether the answer belongs to us.
• 13:38
  SPEAKER_00: Tomorrow, we look at two of the slippery words in health news, the word average and the word risk, and how quote unquote cut your risk in half can mean one person out of a hundred.
• 13:53
  SPEAKER_00: I'll meet you there.
• 13:56
  SPEAKER_00: Keep the conversation going.
• 13:58
  SPEAKER_00: Find your next steps in notes from a fellow traveler, my blog.