Harvard Work Hours Health and Safety Group
A Scheduling Toolkit for Medical Professionals

HOME          MANUSCRIPTS         SAFETY RESOURCES        SLEEP RESOURCES         REPORT AN INCIDENT        CONTACT US

 
 
Graph of the decline in cognitive throughput over 50 hours awake

Number of Continuous Hours Awake

Most people are aware that the longer you stay awake, the sleepier you will become.  The number of continuous hours a person is awake also determines performance levels on simple tasks.

The figure to the left displays results from a lab study where subjects were asked to complete a test battery requiring them to add a series of two digit numbers. There is a steady decline in performance on this task over 50 hours of wakefulness with successful attempts falling below mean performance after just 15 hours (Jewett 1997).

The figure below shows the risk of a resident experiencing a motor vehicle crash following an extended duration shift, compared to a non-extended shift. The risk of having a crash is nearly 6-fold higher following an extended duration shift (Barger 2005).

Taken from Jewett 1997
 

Alarmingly, other research has demonstrated that sustained wakefulness of only 19 hours can lead to performance decrements equivalent to those of someone with a blood alcohol concentration of .05%. When wakefulness is extended to 24 hours performance on a simple reaction time task further declines to that of a person whose blood alcohol concentration is .10% (Dawson and Reid 1997).

Working for 24 hours straight is a common and accepted practice for physician housestaff in the US.

Table showing the risk of having a motor vehicle crash after an extended shift of >24h compared to a non-extended shift. The risk is an OR of 2.3
Figure taken from Barger 2005
Chart showing an increase in cardiac arrhythmias from 5.21 to 9.64 after an average of 7 and  1.8 hours of sleep in the prior 32 hours respectively
Figure above taken from Friedman 1971

In medicine, long duty schedules lead to lapses in attention while doing simple tasks such as placing medication orders and chart documentation. In 1971, long before the debate over housestaff schedules emerged, Friedman and colleagues found that sleep-deprived interns made significantly more errors than rested interns when examining ECG strips for cardiac arrhythmias (see the figure above).

More recently, Grantcharov and colleagues found that surgeons performed significantly worse on a laproscopic simulator task after a call shift of approximately 30 hours compared

Boxplot of errors on a laproscopic simulator before a long oncall shift and after one

to their performance before starting the shift (see the figure below to the left) (2001). Both of these studies provide solid evidence that small lapses in attention can have real consequences in hospitals.

In addition, the odds of a needlestick injury, of a reported fatigue-related harmful or fatal medical error, and of an on-the-job fall-asleep incident also significantly increased secondary to extended duration work shifts (Ayas 2006).

In hospitals, it can be very difficult to create housestaff schedules that minimize the number of continuous hours a person works (and therefore, the number of continuous hours a person is awake) due to the need for 24-hour coverage.  With careful planning, however, coverage can be designed in a way that allows housestaff adequate time for sleep prior to each overnight duty period. It is extremely important to design schedules that minimize the number of continuous hours awake, because the other determinants of sleepiness compound the detrimental effects of being awake for an extended duration.

 

Taken from Grantcharov 2001