Latest / Truth in Motion with Dr. Jocelyn / Half of What? Average and Risk

Transcript

Half of What? Average and Risk · Truth in Motion with Dr. Jocelyn

0:02
SPEAKER_00: Welcome to Truth in Motion with Dr.
0:05
SPEAKER_00: Jocelyn, where insight, clarity, and wisdom help you thrive.
0:09
SPEAKER_00: Hi friends, glad you're here with me today.
0:13
SPEAKER_00: Yesterday we asked who was counted.
0:17
SPEAKER_00: We saw a giant poll miss an election because it counted the wrong people, and a famous aspirin study that counted only men.
0:27
SPEAKER_00: A number is only as good as the people behind it.
0:31
SPEAKER_00: Today we look at two small words that carry enormous weight.
0:36
SPEAKER_00: The word average and the word risk.
0:40
SPEAKER_00: In October of 1995, British health authorities sent out a warning.
0:45
SPEAKER_00: A newer generation of birth control pills, it said, carried twice the risk of dangerous blood clots compared with the older pills.
0:55
SPEAKER_00: Twice.
0:56
SPEAKER_00: A hundred percent increase.
0:58
SPEAKER_00: The story was everywhere.
1:00
SPEAKER_00: Women were frightened, many stopped taking their pill, some overnight.
1:06
SPEAKER_00: Here is what the headline didn't say.
1:09
SPEAKER_00: On the older pills, about one woman in 7,000 would develop a clot.
1:14
SPEAKER_00: On the newer pills, about two women in 7,000.
1:22
SPEAKER_00: That is what quote unquote doubled meant.
1:25
SPEAKER_00: And what happened next was real.
1:28
SPEAKER_00: In the year that followed, England and Wales saw about 26,000 more pregnancies than the year before, and more than 13,000 additional abortions.
1:40
SPEAKER_00: A trend that had been falling for five years turned around.
1:44
SPEAKER_00: The numbers weren't false, the risk really had doubled, but the way it was told changed thousands of lives.
1:53
SPEAKER_00: Why does this matter?
1:55
SPEAKER_00: Because every one of us is making decisions from these words, whether to start a medication, whether to stop one, whether to have the screening, the procedure, the surgery.
2:06
SPEAKER_00: And the same true fact can be told in a way that makes us run towards something or run away from it, scare us off a pill, or sell us one.
2:17
SPEAKER_00: Many of us have walked out of an appointment with a percentage ringing in our ears.
2:22
SPEAKER_00: This cuts your risk by half.
2:25
SPEAKER_00: Without anyone ever saying half of what?
2:28
SPEAKER_00: Or heard that the average person recovers in six weeks and wondered what was wrong with us at week 10.
2:36
SPEAKER_00: An average can make a perfectly normal body feel like a failure.
2:41
SPEAKER_00: When we don't know what a number means, someone else is making the decision for us.
2:47
SPEAKER_00: Here's where I stand.
2:49
SPEAKER_00: I believe every patient deserves to hear risk in plain counts.
2:54
SPEAKER_00: Out of a hundred people like me, how many does this help?
2:59
SPEAKER_00: How many does it harm?
3:01
SPEAKER_00: If a practitioner, a company, or a headline can't answer that, the percentage they're giving me is a sales pitch, not information.
3:11
SPEAKER_00: And I'll say it plainly.
3:12
SPEAKER_00: Relative risk is the language of marketing.
3:16
SPEAKER_00: It makes small effects sound big.
3:18
SPEAKER_00: When a drug company has a choice between 44% and about one in a hundred, we already know which one goes in the ad.
3:27
SPEAKER_00: That's not a reason to refuse every medicine.
3:31
SPEAKER_00: It's a reason to insist on the whole picture, so the choice is truly ours.
3:36
SPEAKER_00: And it's one more reason I start with the gentlest, safest option first.
3:42
SPEAKER_00: Because when the benefit of the strong option turns out to be small, the cost of it matters even more.
3:49
SPEAKER_00: Let's start with the word average.
3:53
SPEAKER_00: In 1954, a journalist named Daryl Huff wrote a little book called How to Lie with Statistics.
4:00
SPEAKER_00: On Monday, we talked about who later paid him, but his book still teaches something true.
4:07
SPEAKER_00: In one chapter called The Well Chosen Average, he describes a single neighborhood.
4:13
SPEAKER_00: A real estate agent trying to sell you a house tells you the average income in the neighborhood is$15,000 a year.
4:21
SPEAKER_00: In the 1950s, that's wealthy.
4:24
SPEAKER_00: Later, the same neighborhood is fighting a tax increase.
4:28
SPEAKER_00: Now a neighbor tells you the average income there is only$3,500.
4:34
SPEAKER_00: Same people, same incomes, same year.
4:38
SPEAKER_00: As Huff wrote, both are legitimate averages, legally arrived at.
4:44
SPEAKER_00: How?
4:45
SPEAKER_00: The first number is the mean.
4:48
SPEAKER_00: You add up every family's income and divide by the number of families.
4:53
SPEAKER_00: A few very wealthy families who kept weekend homes there pulled the number way up.
5:00
SPEAKER_00: The second number is the median.
5:03
SPEAKER_00: Line every family up from the lowest to the highest and take the one in the middle.
5:08
SPEAKER_00: Half earn more, half earn less.
5:12
SPEAKER_00: There's a third kind too, the mode.
5:15
SPEAKER_00: That's simply the most common value.
5:18
SPEAKER_00: The income more families have than any other.
5:23
SPEAKER_00: Three kinds of average, three different numbers.
5:26
SPEAKER_00: And whoever is talking gets to choose which one we hear.
5:31
SPEAKER_00: So when we hear average, the first question is which average?
5:36
SPEAKER_00: Now let me tell you how one man used that question to save his own hope.
5:42
SPEAKER_00: In July of 1982, a Harvard scientist named Stephen Jay Gould was diagnosed with abdominal mesothelioma, a rare and serious cancer.
5:54
SPEAKER_00: He went to the medical library and read that the median's survival was eight months after discovery.
6:01
SPEAKER_00: Eight months.
6:02
SPEAKER_00: Most people would hear that as I have eight months to live.
6:06
SPEAKER_00: Gould heard it differently.
6:08
SPEAKER_00: He knew what a median is.
6:10
SPEAKER_00: It means half the people lived less than eight months and half lived longer.
6:15
SPEAKER_00: And he knew something else.
6:17
SPEAKER_00: Nobody can live less than zero months.
6:20
SPEAKER_00: So the people below the median are all squeezed into those first eight months.
6:26
SPEAKER_00: But the people above it can stretch out for years.
6:30
SPEAKER_00: He looked at himself, young, diagnosed early, with good care available.
6:36
SPEAKER_00: He had every reason to believe he could be in that long tale.
6:41
SPEAKER_00: He wrote about it in 1985 in an essay called The Median Isn't the Message.
6:47
SPEAKER_00: One line from it has stayed with doctors and patients ever since.
6:51
SPEAKER_00: Quote, variation is the hard reality, unquote.
6:56
SPEAKER_00: Stephen Jay Gould died in 2002, 20 years after his diagnosis.
7:03
SPEAKER_00: An average describes a crowd.
7:05
SPEAKER_00: It doesn't decide what happens to any one person in it.
7:10
SPEAKER_00: You are not a median.
7:12
SPEAKER_00: Now let's look at the word risk.
7:15
SPEAKER_00: In 2008, a study called Jupiter was published in the New England Journal of Medicine.
7:22
SPEAKER_00: It tested a cholesterol drug called Rosuvastatin, sold as Crestor, in people who had normal cholesterol but a raised marker of inflammation.
7:34
SPEAKER_00: Nearly 18,000 people took part.
7:37
SPEAKER_00: The trial was paid for by AstraZeneca, the company that makes crestor.
7:42
SPEAKER_00: The headline result, the drug-reduced heart attacks, strokes, and related events by 44%.
7:50
SPEAKER_00: 44%, that's a number that gets written on whiteboards and printed in ads.
7:56
SPEAKER_00: Now let's look behind it.
7:58
SPEAKER_00: About 8,900 people took the drug.
8:02
SPEAKER_00: About 8,900 people took a placebo.
8:05
SPEAKER_00: In the placebo group, 251 people had one of these events.
8:12
SPEAKER_00: In the drug group, 142.
8:16
SPEAKER_00: Out of every 100 people, that's about 2.8 on placebo and about 1.6 on the drug.
8:24
SPEAKER_00: A difference of a little more than one person in a hundred.
8:28
SPEAKER_00: The researchers' own figures put it this way.
8:32
SPEAKER_00: About 95 people would need to take the drug for two years for one person to avoid one of these events.
8:48
SPEAKER_00: And there's more.
8:50
SPEAKER_00: The trial was planned to run for up to five years.
8:53
SPEAKER_00: It was stopped early at a median of under two years.
8:58
SPEAKER_00: Because the benefit looked so good.
9:01
SPEAKER_00: Which means we don't know what longer use in these people looks like from that trial.
9:06
SPEAKER_00: And in the drug group, 270 people were diagnosed with diabetes, compared with 216 on placebo.
9:16
SPEAKER_00: 44% is true.
9:18
SPEAKER_00: One in a hundred is also true.
9:21
SPEAKER_00: Only one of them tells us what it means for us.
9:25
SPEAKER_00: Let me make this real.
9:27
SPEAKER_00: Picture two different conversations in two different exam rooms about the very same drug built on the very same study.
9:35
SPEAKER_00: In the first room, the patient hears this.
9:38
SPEAKER_00: There's a medication that lowers your risk of heart attack and stroke by 44%.
9:44
SPEAKER_00: I'd recommend starting it.
9:46
SPEAKER_00: Who says no to that?
9:48
SPEAKER_00: Almost nobody.
9:50
SPEAKER_00: In the second room, the patient hears this.
9:53
SPEAKER_00: Out of a hundred people like you, about three would have a heart attack, stroke, or similar event over the next two years without this medication.
10:03
SPEAKER_00: With it, between one and two.
10:05
SPEAKER_00: So about one person in a hundred avoids that event.
10:13
SPEAKER_00: There was also a small rise in diabetes diagnosis.
10:18
SPEAKER_00: What would you like to do?
10:20
SPEAKER_00: Same drug, same study, same truth.
10:24
SPEAKER_00: Some people in that second room will still say yes, and that's a good informed choice.
10:30
SPEAKER_00: Some will say, let me first try changing how I eat, how I move, how I sleep, and check again in six months.
10:37
SPEAKER_00: And that's a good informed choice too.
10:40
SPEAKER_00: The difference isn't the answer.
10:42
SPEAKER_00: The difference is that in the second room, the patient is the one deciding.
10:48
SPEAKER_00: That's what I want for every one of us, not to be talked into something or scared out of something by a number we didn't understand, to be handed the real numbers and trusted with them.
11:01
SPEAKER_00: So how do we protect ourselves?
11:04
SPEAKER_00: Here's what I want us to carry.
11:06
SPEAKER_00: When we hear a percentage, we ask percent of what?
11:11
SPEAKER_00: Relative risk compares two groups to each other.
11:15
SPEAKER_00: Absolute risk tells us how many people out of a hundred or a thousand are actually affected.
11:22
SPEAKER_00: A drop from two in a hundred to one in a hundred can be called cut in half.
11:28
SPEAKER_00: It's also one person in a hundred.
11:31
SPEAKER_00: We ask over how long?
11:34
SPEAKER_00: One in a hundred over two years is different from one in a hundred over a lifetime.
11:40
SPEAKER_00: We ask how many people need to take it for one person to benefit.
11:45
SPEAKER_00: Researchers call this the number needed to treat.
11:49
SPEAKER_00: The closer it is to one, the stronger the treatment.
11:52
SPEAKER_00: When it's 95, we know most people taking it won't see that particular benefit.
11:59
SPEAKER_00: We ask, and what does it cost?
12:02
SPEAKER_00: Every benefit has a price.
12:05
SPEAKER_00: Side effects, money, time, what it does to the body over the years.
12:11
SPEAKER_00: The same honest count belongs on that side of the scale too.
12:16
SPEAKER_00: And when we hear average, we ask, which average?
12:20
SPEAKER_00: And how spread out are the people around it?
12:23
SPEAKER_00: These aren't hard questions.
12:24
SPEAKER_00: They are just questions most of us were never taught to ask.
12:29
SPEAKER_00: Here's what honest risk sounds like.
12:32
SPEAKER_00: Out of a hundred people like you, over five years, about this many will have this problem without the treatment.
12:40
SPEAKER_00: With it, about this many.
12:42
SPEAKER_00: And about this many will have this side effect.
12:45
SPEAKER_00: That's it.
12:46
SPEAKER_00: Plain counts, a time frame, both sides of the scale.
12:51
SPEAKER_00: In my treatment room, that's how I want to talk with every person about every option, including the ones I offer.
12:58
SPEAKER_00: When we can see the whole picture, we can choose.
13:01
SPEAKER_00: And choosing freely with our eyes open is where healing starts.
13:06
SPEAKER_00: This week we look behind the number.
13:10
SPEAKER_00: Here's today's practice.
13:12
SPEAKER_00: Find one health claim that uses a percentage, an ad, a news story, a label.
13:18
SPEAKER_00: Then try to turn it into plain counts.
13:21
SPEAKER_00: Out of a hundred people, how many before, how many after, over how long?
13:27
SPEAKER_00: If you can't find those numbers, write that down.
13:30
SPEAKER_00: That empty space is information too.
13:34
SPEAKER_00: Today we saw a doubled risk that meant one more woman in 7,000 and the thousands of lives changed by how it was told.
13:43
SPEAKER_00: We saw a neighborhood with two true averages,$15,000 and$3,500.
13:50
SPEAKER_00: We saw a scientist here eight months and live 20 years because he knew what a median is.
13:58
SPEAKER_00: And we saw a 44% benefit become about one person in a hundred.
14:03
SPEAKER_00: And we saw why it matters.
14:05
SPEAKER_00: The same true number can scare us or sell us.
14:09
SPEAKER_00: When we turn it into plain counts, we get the choice back.
14:13
SPEAKER_00: Tomorrow we look at pictures.
14:15
SPEAKER_00: How a graph can tell a story the numbers themselves don't.
14:20
SPEAKER_00: I'll meet you there.
14:22
SPEAKER_00: Keep the conversation going.
14:24
SPEAKER_00: Find your next steps in notes from a fellow traveler, my blog.