Group items tagged

Everyone knows that doctors must not discriminate on the basis of gender, sexuality, race, religion or national origin when they select or treat patients: It’s an obligation they accepted when they entered the health care profession. (That doesn’t mean they have to take all comers; they can turn away patients for various other reasons.) But should patients be able to choose clinicians on the basis of such attributes? The answer is: It depends.

Perhaps the greatest strength of the randomized controlled trial is in combating what’s known as selection bias. That occurs when groups being studied (intervention and control) are already significantly different after they are “selected” to be in the intervention or not. One of the most elegant examples of why we need such trials came recently in an examination of employer-sponsored wellness programs.

As the doctor Mark Tonelli has argued, distinct forms of knowledge can’t be judged by the same standards: what a patient prefers on the basis of personal experience; what a doctor thinks on the basis of clinical experience; and what clinical research has discovered — each of these is valuable in its own way. While scientists concur that randomized trials are ideal for evaluating the average effects of treatments, such precision isn’t necessary when the benefits are obvious or clear from other data.

Distrusting expertise makes it easy to confuse an absence of randomized evaluations with an absence of knowledge. And this leads to the false belief that knowledge of what works in social policy, education or fighting terrorism can come only from randomized evaluations. But by that logic (as a spoof scientific article claimed), we don’t know if parachutes really work because we have no randomized controlled trials of them.

Experiments, of course, are invaluable and have, in the past, shown the consensus opinion of experts to be wrong. But those who fetishize this methodology, as the flossing example shows, can also impair progress toward the truth. A strong demand for evidence is a good thing. But nurturing a more nuanced view of expertise should be part of that demand.

So what, exactly, did the algorithm “learn” about the process of dying? And what, in turn, can it teach oncologists? Here is the strange rub of such a deep learning system: It learns, but it cannot tell us why it has learned; it assigns probabilities, but it cannot easily express the reasoning behind the assignment. Like a child who learns to ride a bicycle by trial and error and, asked to articulate the rules that enable bicycle riding, simply shrugs her shoulders and sails away, the algorithm looks vacantly at us when we ask, “Why?”

How Will We Remember the Pandemic? Museums Are Already Deciding

How Is ‘Dune’ So Prescient About Climate Change? Thank This Native American Tribe.