Too Little Sleep Makes Jack a Dull Surgeon

More than one quarter of surgeons reported being “impaired” as a result of being tired.

Despite new limits on resident work hours almost half of the orthopedic residents at two of Harvard’s hospitals said they were fatigued during work hours, and 27 percent said they were impaired by lack of sleep.

Moreover, the doctors-in-training said they average just a little over 5 hours’ sleep daily during work weeks, Frank McCormick, MD, of Massachusetts General Hospital, and colleagues reported in the May 21 issue of Archives of Surgery.

The impairment reported by 27 percent of the orthopedic residents who participated in the 2-week assessment, was “as severe as that expected from a blood alcohol level of 0.08 percent,” McCormick wrote.

He concluded that the fatigue-related impairment reported by the residents was “predicted to increase their risk of medical error by 22 percent (individual range 7 percent to 49 percent) compared with well-rested historical control subjects.”

The findings emerged from a prospective cohort study conducted over a 2-week period at two academic tertiary care centers; 27 of 33 eligible orthopedic residents completed the study. Sleep was recorded via actigraphy and residents also completed daily questionnaires, which were used to assess mental fatigue.

Most of the residents (52 percent) were in the second year of the 5-year long orthopedic training program, 18 percent were in the third year, and 15 percent in year four. Two first-year residents (6 percent) and three fifth-year trainees (9 percent) rounded out the sample.

Unlike residents from earlier generations, none of the study participants said they were moonlighting to earn extra income, and 85 percent said they had less than an hour a day for reading.

As might be expected, the residents were regular caffeine users — 72 percent had a least one cup of coffee a day and 65 percent had at least one caffeinated soda daily.

Although all residents reported some degree of sleep deprivation, the authors said that those on one particularly grueling service averaged only about 4 hours’ sleep daily.

“The two residents who participated in the study when on this service were less than 80 percent effective during 86 percent of their time awake and were less than 70 percent effective during 74 percent of their time awake. Their predicted risk of medical error due to fatigue was increased by 43 percent,” McCormick wrote.

In general, residents reported less sleep when they were assigned to night shifts — an average of 5.1 hours versus 5.7 hours when working days.

That 40-minute difference in sleep was apparent when mental effectiveness was compared: night-shift workers “were functioning at less than 70 percent mental effectiveness during 32 percent of their time awake, while day-shift residents were functioning at this level during 17 percent of their awake time.”

That sleep differential correlated with a 24 percent increase in the risk of error due to fatigue for night-shift workers compared with a 19 percent increase in risk of fatigue-related error for day-shift residents.

That difference in error risk is especially relevant “because the night-float resident serves as the inpatient responding physician among 50 percent of the time that a patient remains hospitalized,” McCormick and colleagues wrote.

In an invited critique also published in Archives of Surgery, Brown University pediatric surgeon Thomas F. Tracy, Jr., MS, MD, commended McCormick and colleagues for measuring fatigue through monitoring of movement and use of analytic tools that have been validated by the military.

Tracy concludes that any serious attempt to limit patient exposure to risk is likely to require serious debate about work hours that may require a sea change from the traditional acceptance of “night-float systems.”

The authors cautioned that their analysis has a number of limitations and potential biases, including “the high proportion of night-float residents enrolled,” which may have resulted in an overestimation of both fatigue and risk of error.

They pointed out, for example, that because orthopedic residents also rotated on outpatient services for sports, and foot and ankle clinics, residents in those rotations were underrepresented in the analysis.

And although there was a high rate of participation among eligible orthopedic residents, the brief nature of the assessment — just two weeks — meant that 40 percent of the residents in the training program were not eligible.

Since the study was limited to orthopedic residents, it is not known if the findings can be generalized to other specialties or even other surgical specialties.

Finally, the authors measured only the potential for errors rather than measuring actual errors.

By Peggy Peck