Five years ago, IBM announced that its supercomputer Watson would revolutionize cancer treatment by using its artificial intelligence to digest and distill the thousands of oncology studies published every year plus patient-level data and expert recommendations into treatment recommendation. Last week, a report published by STAT News shows that, years later, IBM’s hubris and hype have crashed into reality.
I’ve discussed so-called “right-to-try” laws, which promise to speed experimental drugs to terminally ill patients, but which in reality are about weakening and bypassing the FDA. Now über-Libertarian Peter Thiel is trying a new tactic to bypass the FDA by organizing an offshore clinical trial of a new herpes vaccine based on dubious science and not overseen by an IRB to protect patients. Both right-to-try and this trial are different fronts in the same fundamentalist free market war on FDA regulation.
Earlier this month, cancer quacks everywhere were touting a study that suggests that chemotherapy administered before breast cancer treatment can stimulate the spread of cancer, pointing to it as evidence that chemotherapy doesn’t work and even makes cancer worse. In reality, the study was far more nuanced. It didn’t show that chemotherapy doesn’t work (quite the contrary) but does point to ways we can make chemotherapy more effective.
Quacks love to invoke experts who made predictions that turned out to be wrong or point to Galileo or Semmelweis as examples of scientists whose findings were rejected by the scientific or medical establishment of the time, as though poor prediction or rejection by the establishment means there must be something to their science. Guess what? As Michael Shermer put it, heresy does not equal correctness.
Dr. Sara Gottfried, MD, wants you to know she is a doctor. She also wants you to know that you can “reset” your hormones and genes with food and saunas. In the case of the saunas, she’s put the preclinical cart before the clinical horse and extrapolated animal and early molecular epidemiological data off of a cliff.