A Walk in the Heat
The other day, my wife and I decided to take an after-dinner walk. We went in the evening to avoid the heat, at 7:15 pm, about as late as one can go out for a 2½ mile walk and still have time to shower, relax a little, maybe draft a random blog entry, and get ready for bed at a reasonable hour. When we started walking, the temperature was 94℉. When we finished, it was past eight, after sunset, and still 91 ℉.
I have lived in Mississippi for 14 years now, and I don’t remember sustained heat like this. It has been as hot, although almost never hotter (that day’s high of 102 ℉ was the hottest recorded temperature in Jackson in eight years), but the heat has never been so unrelenting. The weather is killing my potted banana plants. Bananas are tropical plants, but they, like most living things, depend on the air getting at least a little cooler at night. Being baked at 85 ℉ at midnight is not good for bananas or anything else that grows here. I expect to lose all my potted plants on my porch at this point — I can water them every day, but I can’t protect a banana plant from heat it wasn’t designed to handle. Even in the shade it withers. Eight days straight of 100-plus degree temperatures do that to a plant.
It is hard in the heat to avoid thinking about climate change. Of course, the numbers of a single day or even a week don’t prove the climate is changing. Weather is a localized event, changing daily, varying constantly. Climate, on the other hand, is a statistical phenomenon, a data trend over months, years, and millennia. Weather is hard to predict; long-term trends are more straightforward. What the weather forecaster on TV predicts over the next ten days is an educated guess. But not so with climate forecasts, which is why we should pay attention to them.
Climate is more predictable than weather because large data sets tend to be more consistent and predictable than small ones. This makes them easier to model with mathematics. For example, I can tell you if you flip a coin a thousand times, you will tally close to 500 tails (large data set), but I cannot tell you what your next coin toss will be (small data set). And that is why long-term climate models should be regarded as more accurate than your local weatherman.
One high temperature, or even one heat wave, does not mean climate change. But is anyone left in this country who looks at the extreme heat of this summer and isn’t concerned? The longer the heat waves, and the more frequently they come, the more likely global warming seems. One tails coin flip doesn’t mean something is wrong with the coin, but the more consecutive tails flips you get, the more likely there is something wrong with the coin. As the data set grows, the problem becomes more obvious.
In medicine, we sometimes joke about the n = 1 study. In a research paper, the number of subjects is usually expressed by the value n. Thus, an n = 50 study has fifty patients; an n = 1000 has a thousand patients. The more patients — the larger n is — the more accurate a study is likely to be.
An n = 1 study has only one patient in it — yourself. The joke is that an n = 1 study is just what happened to you. Suppose you feel bad and take ginseng, and a few hours later you feel better. This is an n = 1 study, anecdotal evidence, and from a scientific point of view, means nothing. It’s the story of what happened to you, but it doesn’t prove your experience applies to anyone else.
Walking one night in unbearable heat is an n = 1 study. It is an anecdote. But paradoxically, it is also a personal story. While the experiences of a single person do not matter much in the vast statistical models of science, n = 1 is the only way an individual can link a large amount of information to his or her own experience. I have looked at reams of data showing that cigarette smoking causes lung cancer, but the data meant much more to me when I first encountered a smoker with cancer. The wrinkled skin, the hoarse voice, the hyperexpanded chest of emphysema, the temporal wasting of malnutrition are all hallmarks of lung disease. Seeing them in a real person makes the statistics real. N = 1 is no longer a joke.
Hot days like we have experienced lately are n = 1 studies. They are moments when individuals can see the effects of climate change up close and personal, and incorporate the vast statistics of rising global temperatures into their personal stories.
Medicine is not simply about statistics. It is about patients, who are specific instances of diseases described in textbooks, derived from large databases. The statistics of a disease like Reiter’s syndrome (a rare kind of arthritis that can accompany certain infections) mean little to me until I see a patient that has Reiter’s; then the textbook becomes a story and I can incorporate that story into my thinking in a way no number of hours of study can. A doctor can only incorporate diseases into patient stories when he knows the statistics and how to apply them in real cases. Without the story and the real patient, Reiter’s is just a technical term. On the other hand, without the background statistics, a patient with mysterious arthritis is just a patient with mysterious arthritis. And nothing is learned.
Some people will walk in the heat and make a constructive n = 1 connection. Others will just say, well, it’s been hot before — it happens. But if doctors all said, “There’s another case of mysterious arthritis. It’s happened before,” where would we be?
