Emergency department physicians are badgered by interruptions, many of which cause them to change tasks, leave tasks unfinished, and predispose them to making medical errors.
In a normal two-hour period, an academic emergency room physician may treat as many as 16 patients at the same time, but interact with as many as 132 individuals. In a community hospital setting, emergency room doctors treat as many as 12 patients simultaneously, and interacted with as many as 101.
Those are findings by University of Indiana researchers who used non-medical observers to chronicle tasks performed by 85 academic and community hospital ED physicians during their work shifts. Two unnamed community hospitals and two academic teaching hospitals participated.
By dividing tasks into work units as small as 30 seconds, researchers determined that the majority of ED physicians' activities are spent on indirect patient care, such as charting, reviewing medical records, interacting with consultants, and interpreting tests. The median amount of time spent on indirect patient care in academic settings was 64 minutes, while for community hospital ED doctors it was 55 minutes.
"This report is more groundbreaking, in that we found that in spite of moving toward electronic medical records and point of care testing systems, we haven't increased the amount of time ED physicians spend on direct care activities," says Carey D. Chisholm, MD, lead author an emergency room physician at Indiana University School of Medicine and Methodist Hospital in Indianapolis
That so little time is spent in direct patient interaction, or physically examining the patient, is a concern, Chisholm says.
"Facets of communication are facilitated when you have direct interaction. And the fact that so much of this care has to take place away from the bedside does not facilitate communication between the patient and the physician and their family."
The number of interruptions the ED doctors endured -- up to 19 per two-hour period for community hospital physicians with a median of six, and 32 for academic settings with a median of 12, twice as many -- should also be cause for concern, he says.
Even in the amount of time classified as that devoted to direct patient care, very little of it involved directly touching patients. For academic medical centers, the amount of time ranged from between one to 13 minutes, with a median of six minutes. For community hospital physicians, the range was from zero minutes to 20, with a median of seven minutes.
Chisholm noted that while some interruptions are unavoidable, some of them are unnecessary. Part of the problem centers with how inefficiencies are built into the system "to create unnecessary interruptions," he says.
He gave a common example: an ED doctor determines that he needs to admit a patient to the orthopedic service. "I call and leave a message, and now I have to sit by the phone for 2 to 30 minutes. And I may be in the middle of performing a physical on another patient when the service calls back, calling me away from the bedside. And that interrupts."
He suggests that EMR systems create electronic "virtual patient records accessible to multiple nurses and consultants for review and data entry simultaneously, allowing asynchronous information transmission and interaction without interruptions." Also, use of electronic message boards "could allow communication with consultants without the currently built-in telephone call interruptions."
In this way, Chisholm says, the ED doc could post information in the virtual medical record that would explain the patient's needs to the orthopedic surgeons on the service, "and they could reply in electronic format, rather than place a call back," avoiding a good number of these interruptions.
The researchers observed 203 two-hour periods and four hospitals, two of them community hospitals and two academic.
Why emergency physicians in academic teaching hospitals had twice as many interruptions is unclear. The authors speculated that some of them might be due to inquiries from students, or the higher overall acuity of illness necessitating discussions with multiple consultants.
Chisholm, however, noted that some reports suggest that when doctors are interrupted, they fail to return to the interrupted task one in five times. "There is evidence supporting the negative effect of interruptions on task performance and subject perception of stress," prompting providers to compensate with short cuts.
But aren't emergency room physicians of a personality type that thrives on fast-paced, ever-changing circumstances, capable of rapid "switchtasking?" Chisholm says that they are. But that doesn't get around the fact that "all the cognitive literature suggests that the human brain doesn't adapt well to interruptions. And at the baseline, it's something we should attempt to control better. It's not a matter of expressing frustration, or whining, it's that you're delivering care less effectively than you should. And it does end up contributing to the potential for medical error."