COVID19: How Do Scientists Think?
President Trump was criticized recently for saying disinfectants could be injected or ingested to kill the coronavirus. The criticism is unfair. He didn’t say that. He asked a question: “is there a way we can do something like that, by injection inside or almost a cleaning…” Clearly, his public speculation wasn’t helpful; hundreds of people took it as a serious proposal and called medical hotlines. The FDA, CDC and makers of disinfectants almost immediately put out warnings that such products were for external cleaning only.
The incident exposes the confusion many Americans have about science in regard to COVID19. Following the president’s remarks, one Facebook thread revolved around an argument that an antibiotic and a disinfectant are really the same thing (not true) since they both deal with infectious agents (true). Another point of confusion is around whether there is enough testing, with the president claiming we’ve tested more people than any other nation (true) but others arguing that it’s less than two percent of Americans (also true). Many don’t understand why it takes so long to develop a vaccine, unaware that a promising candidate must go through clinical trials to determine effectiveness and proper dosage before manufacturing much less distribution can be addressed.
One of the most controversial questions in which science is involved is whether already available drugs such as hydroxychloroquine and remdesivir are effective treatments. If you believe in the power of these medications, you can find evidence of people cured by them. If you doubt their efficacy, you can find evidence for that too. Beneath both arguments are misunderstood principles of scientific reasoning.
Press accounts of individual cures (or failures) are anecdotal evidence, which is easy to find. They are a simple form of correlation – this person took this drug and is now virus-free (or not). The example often given to young science students is the fact that roosters crow in the morning and people eat cereal for breakfast does not mean roosters cause cereal to be eaten. Correlation and causation are different. Science needs more than anecdotes, which can lead to asking scientifically testable questions but should never be confused with proof. For the same reason, lots of anecdotes don’t add up to anything but lots of anecdotes.
Pharmaceutical science advances through controlled studies. The fact that a doctor, for example, gave 100 patients a new drug and reports 80 of them got better does not answer critical questions: (a) how many would have gotten better without it?; (b) did the mere fact patients were told they were getting the drug help them improve (the placebo effect)?; (c) what demographic, medical, or other traits of the individuals might account for the results? Nor does the fact that lots of doctors are using a medication prove its effectiveness. In the eighteenth century, almost all doctors used bloodletting.
A controlled study deals with such issues. The gold standard is the double-blind, randomized control trial that is peer-reviewed and stands up to scrutiny. In such experiments (on which we still await sufficient results for hydroxychloroquine and remdesivir), patients are randomly assigned to either receive a medication or a placebo and neither they nor the doctor knows who got which (the latter to make sure the doctor does not skew treatments or interpretation of the drug’s effects to get a preferred outcome).
Yet another frustration Americans anxious for comforting COVID19 news have is that scientists disagree with each other or change their own minds. Consider the controversy about how far apart we need to be for social distancing. Is it 6, 20 or even 100 feet (the last emerging from a study of virus particles found in industrial smoke). Or consider the confusing messages about whether and when we should wear masks. As consumers of science, we want certainty. When scientists disagree or waffle, we lose trust in science. Yet good researchers are almost never certain. They are trained to question their own and others’ work. Science advances by admitting it could be wrong and a willingness to revisit its conclusions. Scientists seek not “truth” but testable hypotheses. Even when results are promising, they are reported only in terms of the probability they are correct, which is never 100 percent.
The coronavirus is educating us about scientific thinking. That’s good. A 2019 Pew Research survey showed only 60 percent of Americans “identify the need for a control group to determine the effectiveness of a new drug.” The survey revealed that the less formal education people had, the worse they did on science knowledge. In a 2015 study of science literacy in 71 countries by the Programme for International Student Assessment, the United States ranked 24th. On the National Assessment of Educational Progress, 12th grade Americans’ performance showed no improvement between 2009-2015, scoring 150 out of 300 both times.
Science has a critical role to play in dealing with COVID19. It cannot tell us when to “open up America.” That is a political and ethical question as well as a scientific one. But it can inform us if we understand and use it.
Photo Credit: CDC.gov
