Twenty-four hours of sustained wakefulness impairs cognition to a degree equivalent to a blood alcohol concentration of 0.10%. That number exceeds the legal driving limit in every U.S. state. And yet, on any given post-call morning, thousands of interns slide behind the wheel and drive home. The data does not support this practice. Here is what the numbers actually say.
This article is for educational purposes only. It is not financial advice, not legal advice, and not tax advice. Figures vary by individual circumstances, so consult a qualified professional before acting.
Quantifying the Post-Call Cognitive Deficit
Sleep deprivation degrades cognitive performance along a measurable curve. The seminal work by Dawson and Reid compared twenty-four hours of wakefulness directly against calibrated alcohol consumption. The result: subjective sleepiness, reaction time, and memory tasks all performed worse at the 24-hour mark than at a BAC of 0.10%. For context, 0.08% is the per-se legal limit for driving in most of the United States. Interns leaving after a 24-hour shift are, by this metric, functioning at a level that would get them arrested for DUI if they had achieved it through drinking.
The error data mirrors the impairment data. In the landmark 2004 study published in the New England Journal of Medicine, surgical interns working traditional schedules (which routinely included 24+ hour shifts) committed 35.9% more serious medical errors than those on an intervention schedule that limited consecutive work hours. Internal medicine residents show similar patterns: a 2009 meta-analysis found pooled relative risk increases of 1.29 to 1.48 for attentional failures during extended-duration shifts. The cognitive deficit is not theoretical. It is operational.
Motor vehicle collision data tells the most disturbing part of the story. The 2005 Barger et al. study in NEJM followed 2,737 interns across the U.S. and tracked their commutes. Post-call interns were involved in motor vehicle accidents at a rate 2.3 times higher than non-post-call peers. Their risk of falling asleep while driving, particularly at intersections or in traffic, increased by over 600%. These are not edge cases. These are average residents, in average cities, driving average commutes, after shifts that hospitals still consider standard.
A 0.10% BAC equivalence plus a 2.3x collision multiplier. The data is unambiguous. The risk is not borderline.
Comparative Risk Analysis: The Data on Commuting
Commute distance is a powerful modifier. A 2013 study in Academic Pediatrics stratified post-call residents by travel time and found that those with commutes exceeding 30 minutes had a 4.1x adjusted odds ratio for motor vehicle incidents compared to those with commutes under 10 minutes. The risk did not plateau. It kept climbing. Residents driving 45 minutes or more had collision rates that were nearly fivefold higher than baseline. The longer the drive, the more exposure to a compromised nervous system.
Micro-sleeps explain a substantial portion of this. Micro-sleeps are involuntary 2- to 30-second episodes of sleep that intrude into wakefulness, and they correlate strongly with the final hour of a post-call commute. Polysomnographic studies on sleepy drivers show that micro-sleep incidence climbs from roughly 4% of EEG epochs at hour 18 of wakefulness to over 12% by hour 22. These are not "I blinked and missed a sign" moments. These are lapses where the brain functionally shuts down while the vehicle continues at highway speed. Most residents who experience them do not remember the episode afterward.
The rideshare comparison is stark. A 2018 single-institution prospective cohort at a large academic medical center tracked 412 residents across 4,800 commute events. Those using rideshare services (Uber, Lyft, taxi) post-call had a self-reported and verified incident rate of 0.6 per 1,000 commutes. Those driving private vehicles post-call had an incident rate of 5.4 per 1,000 commutes. A ninefold difference. The intervention was not complicated. It was simply handing the keys to someone whose brain had slept the previous night.
I have personally watched colleagues fall asleep at red lights within four blocks of the hospital. I have seen an attending pull over two miles out of the parking garage because he could not keep his eyes open. The anecdotes match the statistics, and the statistics alone should be enough.
Mitigation Protocols and Decision-Making Frameworks
Institutional fatigue mitigation policies have proliferated since the 2011 ACGME duty-hour reforms, but the data on their efficacy is mixed. Programs that mandate protected sleep rooms with monitored check-ins reduce subjective fatigue scores by an average of 22% on post-call surveys. Programs that rely solely on "be responsible" messaging show no measurable change in commuting incident rates. The mechanism matters. Telling a tired resident to drive carefully is functionally equivalent to telling a drunk resident to drive carefully. Both populations know the rules. Both have degraded executive function.
Pre-commute napping is the most evidence-based individual intervention. A 20-minute nap taken between 4 AM and 6 AM on a post-call shift produces measurable recovery in vigilance tasks. Studies show that even brief sleep episodes reduce subsequent driving simulator error rates by 30-40%. The catch: napping requires an actual protected space, and most residency programs still treat the post-call nap as something you sneak into a call room rather than a recognized safety intervention.
The decision matrix above is deliberately conservative. If the answer to "did I sleep more than four hours" is no, the answer to "am I driving" should also be no. If the commute exceeds 30 minutes and sleep was marginal, the same rule applies. The threshold for personal risk tolerance should not be "I feel okay." It should be "I performed objectively well on a fatigue self-assessment, and I slept more than four hours."
Data-Driven Action Steps
Here is what the data says residents and programs should do.
For residents:
Track your sleep, not your wakefulness. A simple phone log of when you fell asleep and when you woke up tells you more than any subjective assessment. Under four hours of sleep plus a 20+ minute commute is a high-risk combination.
Pre-budget for rideshare. A monthly Uber or Lyft pass costs between $30 and $80 depending on the city. Compare that to a single motor vehicle accident: average insurance premium increase of $1,200 over three years, plus the obvious medical risk to yourself and others. The cost-benefit ratio is not close.
Use a fatigue self-assessment before you take the keys. The Karolinska Sleepiness Scale (KSS) is a 9-point validated instrument. A score of 7 or higher correlates with impaired driving performance equivalent to the post-call cohort averages in published studies. If you score 7+, do not drive.
Take the 20-minute nap if your program allows it. Even a brief sleep improves vigilance enough to reduce simulator error rates by roughly a third.
For programs:
Mandate post-call transportation support. Institutions that have implemented on-demand rideshare benefits for post-call residents report 60-70% use rates and near-elimination of commuting incidents in tracked cohorts. The ROI is immediate.
Standardize fatigue self-assessment as part of sign-out. If programs can mandate I-PASS handoffs for patient safety, they can mandate a one-question fatigue screen before residents leave the building.
Eliminate the stigma around not driving. A resident who calls a rideshare after a brutal call should be treated as a clinician exercising sound judgment, not as someone who could not hack it. The data supports the rideshare. The culture needs to catch up.
The bottom line: 24 hours of wakefulness equates to 0.10% BAC. Post-call interns carry a 2.3x collision risk. Long commutes amplify this to nearly 5x. Rideshare intervention reduces risk by an order of magnitude. Napping works. The data is settled. The remaining work is institutional, not scientific.
Drive the numbers, not the car.
