Technology

First was a diagnostic arthroscopy. Then randomization. Active arm surgery included removal of all loose, weak fragments as well as unstable meniscus with preservation of as much of the meniscus as possible.

For the sham surgery, a standard arthroscopic partial meniscectomy was simulated. To mimic the sensations and sounds of a true arthroscopic partial meniscectomy, the surgeon asked for all instruments, manipulated the knee as if an arthroscopic partial meniscectomy was being performed, pushed a mechanized shaver (without the blade) firmly against the patella (outside the knee), and used suction. The patient was also kept in the operating room for the amount of time required to perform an actual arthroscopic partial meniscectomy.

Outcomes were three pain scores. Their names aren’t critical. But all three were not statistically different. Critically, patients could not identify their treatment arm; blinding had worked.

In the 10-year radiological assessment. 81% of the patients in the operative arm had progressed to have arthritis vs 70% in the placebo arm. This barely missed statistical significance. More patients in the placebo arm reported satisfaction (84% vs 75%). More patients in the surgery arm progressed to high tibial osteotomy or total knee replacement (12% vs 4%).

The first thing that pops in my head is: what if these researchers (and patients) were not so bold. An ineffective procedure would have remained popular, with costs into the billions, and if the trends in this study are true, increasing numbers of people would progress to knee replacement. AKA: harm.

What if the surgery that fixed your knee did no better than fake surgery? EconTalk host Russ Roberts speaks with Dr. John Mandrola about a striking clinical trial in which patients who received sham knee surgery (a real incision, but no actual repair) did as well or better than those who had the actual procedure — one performed 700,000 times annually in the US. The conversation ranges from the power of placebo and nocebo effects (how expectation of harm can cause real suffering) to the broader philosophy of “medical conservatism” — the idea that humility, watchful waiting, and honest counsel often serve patients better than the knife. Mandrola argues that financial incentives, professional identity, and language itself (“bone-on-bone,” “the widowmaker”) conspire to push patients toward interventions that can do more harm than good.

An international team of researchers found that adults with moderate to severe depression who took part in twice-weekly monitored Nordic walking sessions experienced substantial reductions in depressive symptoms, with the most dramatic improvements occurring during the first half of the 10-week program.

By the end of Week 10, between 35% and 53.6% of walkers had achieved what would be clinically described as remission.

The term seasonal affective disorder, or rather its catchy acronym SAD, is so popular that it’s used in casual conversation. Steve LoBello, a psychologist and researcher at Auburn University at Montgomery, set out to do his own assessment of the nationwide scale of SAD—annual depression that follows a strict seasonal cycle, typically occurring in fall and winter and receding in spring and summer. LoBello and his team analyzed data from the CDC’s behavioral risk-factor survey, which asks hundreds of thousands of Americans each year about their health and well-being, including a separate screening for depression and anxiety, to see whether major depression rates followed a seasonal trend. “We expected cases to increase in the wintertime and then for that to subside starting in early spring and so forth, and there was nothing like that in the data,” LoBello told me of the study they published in 2016. “It was just flat as a pancake all the way through the year.” They also found no correlation between major depression and the respondent’s latitude (or hours of daylight). A couple of years later, in 2018, LoBello published another paper that found no correlation between even mild depression and the seasons. Still, the idea that we are all more likely to be sad and depressed in winter has dominated, and LoBello argues that that view is more steeped in folklore than science.

The first system, Mira, was built by academic researchers in Germany.

Given access to a simulated medical record, it can choose from more than 85,000 actions: tests, prescriptions, even hospital admissions. Across more than 500 emergency-department cases, it reached a diagnostic accuracy of about 87 per cent, against 78 per cent for a panel of six doctors. It was strongest on conditions with clear test results, such as pancreatitis and appendicitis.

The second, Amie, is Google’s, and runs on its Gemini model.

Tested against 21 UK GPs across 100 multi-visit cases, it matched them on clinical reasoning and produced treatment plans that stuck more closely to official guidelines. On a benchmark for tricky medication decisions, it came out ahead.