It's time to reform work hours for resident physicians.

Last year, the American medical profession admitted that its century-old tradition of sleep-depriving physicians-in-training is unsafe. Working day and night, rather than transferring patient care to a fresh doctor, increases serious medical errors in ICUs by 36 percent, including a 460 percent increase in serious diagnostic mistakes. Twenty-four hours without sleep slows reaction time comparably to alcohol intoxication. Physician-trainees routinely fail asleep during lectures, on patient rounds, while examining patients and even during surgery.

Sleep deprivation impedes memory consolidation and degrades physicians' clinical performance to the 7th percentile of rested performance. After working more than 24 hours, resident physicians are 73 percent more likely to stab themselves with a needle and 168 percent more likely to crash driving home. One-quarter of them make most sleep-related errors, likely due to sleep disorders or differential vulnerability to sleep loss, which may be genetic. Yet resident physicians with medical conditions that put them and patients at greater risk from sleep deprivation caused by working conditions requiring 30-hour shifts twice per week are seldom provided with a reasonable accommodation.

In 2006, the Harvard Work Hours, Health and Safety Group reported that one in five resident-physicians admitted making a fatigue-related mistake that injured a patient. One in 20 admitted a fatigue-related mistake that resulted in a patient's death. As a consequence, a U.S. House committee requested the Agency for Healthcare Research and Quality to commission the Institute of Medicine to determine whether long resident work hours compromise patient safety. After studying decades of research and hearing expert testimony, the IOM concluded what many find obvious: the extended duty 30-hour shifts permitted in the current ACGME (Accreditation Council for Graduate Medical Education) limits on resident duty hours "promote conditions for fatigue-related errors that pose risks to both patients and residents." The IOM recommended that resident physicians should not work more than 16 hours without sleep, should not be awakened from sleep to treat patients and should not drive home after work shifts longer than 16 hours.

This is an important step toward reforming a U.S. tradition established at Johns Hopkins in the 1890s by William Haisted, whose cocaine addiction perhaps clouded his judgment as to how long physicians could safely work. New Zealand has limited physician-trainees to 16-hour shifts since 1985. The European Union limits work shifts to 13 hours, requiring 11 consecutive hours off daily. Unfortunately, organized medicine has stalled implementation of the IOM recommendations. When other arguments fail, the claim is made that resident work-hour reform is too costly, despite the reality that taxpayers already pay all costs of residency training many times over. Medicare funds teaching hospitals with about $10 billion annually for the 100,000 physicians-in-training in the United States--often paying nearly triple residents' salaries. Medicare and private insurers then pay those same hospitals again for the medical services residents render. Yet, despite these subsidies, no federal regulations limit the hours these physicians work, exposing them to chronic sleep deprivation that adversely affects their health, increasing their susceptibility to infection and to the risk of hypertension, obesity, diabetes, occupational injury and depression. While on such schedules, 20 percent of pediatric residents screen positive for clinical depression and 74 percent for burnout, a syndrome of mental exhaustion and personal detachment that develops in response to chronic job stress.

Yet 30-hour shifts are deeply entrenched. In 2001, the Association of American Medical Colleges urged hospitals to impose a 12-hour work limit in ICUs, to no avail. The ACGME staved off then-pending legislation by announcing work-hour limits that sanctioned twice-per-week 30-hour shifts rather than eliminating them, refusing to disclose work-hour compliance data from member hospitals, failing to provide whistleblower protection, abiding falsification of resident work-hour records, and allowing 84 percent non-adherence to current ACGME work-hour limits.

By the residents' own admission, tens of thousands of patients and residents are injured and thousands of patients lose their lives annually from fatigue-related errors. Many safety-sensitive workers have work-hour limits, including airline pilots and nuclear power engineers. Congress should implement meaningful resident-physician work-hour reform based on the IOM recommendations to prevent these unnecessary tragedies.

Charles A. Czeisler is director of the Division of Sleep Medicine at Harvard Medical School and chief of sleep medicine at Brigham and Women's Hospital. Read more at understandingsleep.org