More Physicians Are Staying Put — But Not Enough
Over one in five physicians still wants to cut back on clinical hours, and nearly one in six is planning to leave their organization — even as both numbers have quietly improved since the COVID-19 peak.
A serial cross-sectional study of 37,112 physicians across 160 US health systems found that intent to reduce hours (ITR) fell from 25.6% in 2022 to 22.5% in 2024, while intent to leave (ITL) dropped from 19.9% to 15.1% over the same period. Progress, yes — but the authors are careful about calling it a win: "current rates represent mixed progress with favorable overall trends in ITL but unfavorable trends in ITR," they write.
Here's the sneaky part: female physicians were more likely to want to cut back (OR 1.11) but less likely to want to leave (OR 0.93) than male peers. Physicians 20+ years into practice had the highest odds of both. Part-timers, counterintuitively, were more at risk for both outcomes than full-timers.
The "why" likely involves workflow inefficiency, compensation dissatisfaction, and staffing instability — the factors physicians most consistently cited as potential reasons to reconsider, across both outcomes. These aren't mystery variables; they're known, addressable, and expensive to ignore: replacing one departing physician runs $500K–$1M.
The trend is real. The problem isn't solved.
Source: JAMA Network Open
Your Brain Is Lying to You About That Oncoming Car
Approaching sounds feel closer than they are — and a loudness model may explain why, no evolutionary bias needed.
The "adaptive looming bias theory" has long held that the auditory system evolved to overweight approaching threats, giving us a built-in margin of safety. A new study in Proceedings of the Royal Society B doesn't demolish that idea — but it does offer a simpler explanation that fits the data just as well.
Eleven blindfolded participants judged distances of virtual approaching and receding tones and white noise across three ranges (1.2 to 24.8 m). Approaching sounds were consistently judged as closer than equivalent receding ones, strongest at close range — consistent with the bias theory. But an ISO-standard loudness model predicted the same pattern with notable accuracy (r = 0.976 for tones, 0.966 for noise), no evolved bias required.
The mechanism: the loudness model has asymmetric time constants — rising levels build loudness perception faster than falling levels dissipate it. As the authors note, "it is not necessary to invoke the adaptive perceptual bias theory to account for asymmetries in loudness and distance judgements between approaching and receding sounds."
The authors are careful to flag that these findings are specific to anechoic, level-only conditions — real-world environments are messier.
For clinicians thinking about hearing aid compression or auditory rehabilitation: distance misperception for dynamic sounds may have a simpler psychoacoustic basis than previously assumed.
Source: Proceedings of the Royal Society B
Irregular Bedtimes May Signal Cardiac Risk — At Least in Short Sleepers
Forget what time someone goes to bed. The real signal might be how much that time shifts night to night.
A Finnish cohort study tracked 3,231 middle-aged adults for over a decade and found that irregular bedtimes were associated with about twice the risk of a major adverse cardiac event — though the confidence intervals were wide (HR 2.01, 95% CI: 1.00–4.01), and the finding just crossed the threshold for statistical significance. The association only emerged in participants sleeping under 7 hours 56 minutes per night. In longer sleepers, irregular timing showed no meaningful association.
Here's the subtle part: average bedtime didn't differ between those who had events and those who didn't. It was the variability across 7 consecutive nights that carried the signal. Irregular sleep midpoint showed a similar pattern. Wake-up time irregularity? Nothing.
The authors are careful about mechanism, noting only that "maintaining consistent sleep-onset timing may be more critical for supporting sleep quality and cardiovascular recovery than consistent wake-up time alone."
This is hypothesis-generating, not practice-changing — but for shorter sleepers, bedtime consistency may be a reasonable conversation to have.
Source: BMC Cardiovascular Disorders
Dental Schools Are Still Using Cadavers — And Nobody's Really Measuring Whether It Works
Forty-four studies. Fifty years of literature. And still no agreed-upon way to measure whether practicing on a cadaver actually makes a better dentist.
A new scoping review from Newcastle University surveyed everything published on cadaveric training in dental education from 1972 to 2023. The vast majority of studies — 82% — involved anatomy teaching, not hands-on surgical skill development. Local anesthetic delivery, exodontia, and suturing appeared in only a small handful of papers each.
The most common outcome measure across studies was student self-reported confidence — useful for understanding anxiety, but the reviewers note there are "limitations on using self-reported skill acquisition in its translation to be a surrogate of objectively improved clinical ability."
The tricky part: cadaveric training rarely lives in isolation. It's typically embedded in larger multi-modal curricula, making it genuinely hard to attribute outcomes to any single component.
As the authors note directly: "There was no universally recognised outcome measure used to consider the translation of cadaveric learning to practical impact on the development of surgical skill and subsequent delivery of patient care."
For educators thinking about curriculum design, the infrastructure for cadaveric training exists — the measurement framework to evaluate it meaningfully is still a work in progress.
Source: British Dental Journal
The clinical literature. Applied to the patients in your waiting room.
