Lit Matters #2: Speed versus Safety

Emergency physician productivity varies with site familiarity, shift timing, and operational crowding more than with simple years in practice. In a large national emergency medicine dataset, faster physicians did not have higher 72-hour return rates, suggesting that speed and short-term safety are not necessarily in conflict.

Emergency Physician Productivity Signals

• Site-specific experience effect: Productivity rose with longer tenure at the same facility, with gains continuing out to about 5 years, suggesting local workflow knowledge matters as much as raw post-residency experience.
• Early shift performance drop: Patients per hour fell after the 6-hour mark, reinforcing that fatigue and shift design measurably erode throughput rather than productivity staying flat across a long shift.
• Overnight and weekend slowdown: Night shifts produced fewer patients per hour than day shifts, and weekends lagged behind Mondays, a useful reminder that schedule context changes expected output.
• Boarding pressure penalty: More boarded patients, defined here by ED length of stay over 6 hours, were linked to lower physician productivity, underscoring boarding as a throughput problem not just a bed problem.
• Crowding and staffing paradox: More physicians on shift and more active ED patients were associated with lower individual patients per hour, a counterintuitive operations signal we get into in the episode.

Speed and Safety Tradeoff

• Bounceback proxy for safety: Higher productivity was not associated with more 72-hour returns, making the common speed-equals-missed-diagnosis assumption harder to defend, at least by this metric.
• Return with admission signal: Faster physicians also did not show an increase in 72-hour returns requiring admission, which weakens the argument that higher throughput reflects riskier dispositions.
• Small absolute difference: The return-rate difference slightly favored more productive physicians, about a 0.1% absolute reduction, though that should be read as reassuring rather than definitive proof of quality.
• Limits of bounceback metrics: Seventy-two-hour returns are a practical quality marker but an incomplete one, because diagnostic error and patient-centered outcomes can still be missed. We cover that caution in the podcast.
• Operational training implication: If efficient clinicians are simply more efficient, then onboarding, retention, and scheduling around experienced site-specific physicians may be a safer lever than assuming speed itself is the problem.