Unplanned ICU Transfers Often Preventable but Are Not Typically Due to Clinical Error

Terry Hartnett

April 12, 2010

April 12, 2010 (Washington, DC) — Transfers to the intensive care unit (ICU) that occur within 24 hours of presentation at the emergency department (ED) are primarily the result of a worsening of the initial condition or the development of a new problem. Very few of these transfers are due to diagnostic error or inappropriate treatment, according to a report presented here at Hospital Medicine 2010: Society of Hospital Medicine Annual Meeting.

The general consensus is that these transfers can be avoided, but a study presented by lead investigator Srinivas Bapoje, MD, MPH, a hospitalist at the Denver Health Medical Center in Colorado, suggests that improving triage and monitoring in the ED might result in immediate appropriate transfer from the ED to the ICU.

Dr. Bapoje told Medscape Internal Medicine that the accepted thinking on why so many patients end up in the ICU when they were not originally triaged there from the ED has been that clinical errors lead to unplanned transfers.

"Most of the patients in our study went to the ICU within 24 hours, and many within the first 6 hours," he said. "It seems clear that these patients likely met ICU criteria in the ED," said Dr. Bapoje. "We need to screen patients better."

The study involved 152 patients. The 2 most common diagnoses were respiratory failure (24%) and acute coronary syndrome (11%).

"Errors in care accounted for the transfer of 29 patients (19%), but in 15 of 29 patients (52%) the errors were in triage, because 14 of 15 (93%) met ICU admission criteria while still in the emergency department," Dr. Bapoje and his colleagues reported.

The study found 4 primary reasons for unplanned ICU transfers:

  • Triage errors from the ED. This occurred in 15 cases (10% of the total), 14 of which (9%) were due to overlooked severity; only 1 (0.7%) of the unplanned ICU transfers was due to a diagnostic error. None of the transfers were the result of inadequate assessment.

  • Worsening of the presenting condition. This was true for close to half of the patients in the study, with 73 patients (48%) experiencing worsening (68 of which were spontaneous). There were only 5 errors in assessment or treatment (3%).

  • Development of a new problem. A significant number (59 patients [39%]) experienced this situation.

  • Critical lab values. Only 5 cases (3%) fell into this category.

Dr. Bapoje told Medscape Internal Medicine that a rapid response team was used in the management of 106 of the study patients, but there was no recognized difference in outcome from those managed conventionally.

The researchers also found that mortality decreased for patients transferred to the ICU in the first 24 hours of hospital admission, compared with patients transferred after the initial 24 hours (4% vs 22%).

Still, Dr. Bapoje found that 29 (19%) transfers might have been prevented. "If the transfer occurs from the floor to the ICU, it adds to the cost and lengthens the stay," he said. "When the transfer occurs directly from the ED, it allows us to manage resources better," he noted.

A faculty member at the conference, Mary Jo Gorman, MD, FHM, founder and chief executive officer of Advances ICU Care, based in St. Louis, Missouri, told Medscape Internal Medicine that the results are not surprising, but she agreed that they point to the need for better observation and a quicker response to changes in the patient's condition.

"Hospitalists are increasingly being asked to fill the gap of intensivists in the ICU," said Dr. Gorman. "We need to closely monitor these patients and proactively intervene when necessary," she said.

Dr. Gorman's company monitors patients through a telemedicine system that puts a monitor and camera in each patient's room. The data feeds are monitored around the clock and allow for a quick response to any worsening or change in condition.

Dr. Bapoje said that although the deterioration of a patient's condition cannot be alleviated, the hospitalist who is based in the hospital around the clock can do a better job of monitoring and providing a quick response.

The study did not receive commercial support. Dr. Bapoje and Dr. Gorman have disclosed no relevant financial relationships.

Hospital Medicine 2010: Society of Hospital Medicine (SHM) Annual Meeting: Abstract 19. Presented April 9, 2010.