Key Point:

Communication, teamwork, and simplifying and automating complex processes are essential to improving the safety of patient care in the ICU.

BALTIMORE—Due to its complexity, the ICU is a breeding ground for potentially dangerous errors. One study, for example, found 1.7 errors per patient daily in the ICU.¹ “Of these errors, 29% could have caused clinically significant harm or death,” related study author Peter J. Pronovost, MD, PhD.

Dr. Pronovost, Associate Professor of Anesthesiology and Critical Care Medicine at The Johns Hopkins School of Medicine in Baltimore, recently offered practical recommendations for improving safety in the ICU.² Better teamwork and communication and simplifying and automating complex processes are the foundation of his approach.

TO ERR IS HUMAN

“In medicine, we must allow ourselves to be human and accept that we will make mistakes—a bitter pill for some physicians to swallow,” stressed Dr. Pronovost. Thus, the objective must be “to deliver care that is harm-free rather than error -free,” he said.

The aviation industry may provide a useful model for safer health care delivery because of its well-defined hierarchy, redundant processes, and industry-wide focus on safety. In aviation, there is a strong commitment to preventing errors or at least making them visible before they cause harm. That usually is not the case in health care, according to Dr. Pronovost.

Health care may become more like aviation by improving teamwork and communication, concepts that are inadequately addressed during medical school and clinical training. Dr. Pronovost suggested that the current model of teamwork in health care—physicians give orders and nurses take them—may need to be recalibrated in favor of a system in which all staff are taught to speak up when they have concerns and to acknowledge personal and organizational vulnerabilities.

However, effective communication requires that a message is both sent and received, Dr. Pronovost emphasized. Thus, those in positions of authority should receive training that enhances their ability to listen to colleagues who voice concerns.

Another aspect of communication, trainee supervision, has substantial room for improvement, according to Dr. Pronovost. In a large survey of caregivers in the United States, United Kingdom, and New Zealand, only 58% of respondents reported that trainees in their discipline were adequately supervised.

To better ensure the competency of a fledgling physician, one might ask his or her peers and nurses if they would recommend that physician to a family member or friend. “Simulation methods may provide another, although more costly, method to evaluate competency,” Dr. Pronovost added.

SIMPLIFY AND AUTOMATE

In aviation, checklists increase safety by making key activities simpler and more redundant. In health care, similar checklists could be used to reduce errors by breaking complex diagnostic and therapeutic processes into a series of simple yes-or-no tasks. To prevent venous air embolism during central line removal, for example, the health care provider could follow a checklist that includes placing the patient in bed in the Trendelenburg position, having the patient perform the Valsalva maneuver, removing the catheter, and sealing the insertion wound with an occlusive dressing.

“Checklists that enlist both physicians and nurses to ensure adherence to evidence-based therapies are particularly helpful,” said Dr. Pronovost. “If physicians do not adhere, nurses could intercede, analogous to a flight engineer stopping take-off if events deviate from the prescribed sequence.” Obviously, checklists are only beneficial if they are used, he noted, and automating broken processes may yield poor results.

ADDRESSING ERRORS

When an error occurs, the organization must ensure that the entire staff learns from the mistake. “Few hospitals do this,” Dr. Pronovost stated. However, he maintained, it can be done without exposure to undue risk by first recognizing that the failure to improve and the hazards of recurrent errors are far worse than managing the immediate risk of a specific incident. “One possibility is to selectively discuss cases in which harm did not occur or in which a settlement [was reached] with or without litigation,” he suggested. Learning can be facilitated by assuring staff that individual performance will not be considered when discussing errors.

Although detailed root-cause analysis of mistakes is common, there is not much evidence that it actually helps to improve the safety of patient care. An alternative that, incidentally, is required by the Joint Commission on Accreditation of Healthcare Organizations is failure mode and effect analysis. This analysis involves examining a process prospectively to forecast errors. However, it “has not been adequately tested in health care, and in its current form is impractical for routine use, especially for frontline staff,” Dr. Pronovost asserted.

—Timothy Begany

References
1. Vincent C, Taylor-Adams S, Stanhope N. Framework for analyzing risk and safety in clinical medicine. BMJ. 1998;316:1154-1157.
2. Pronovost PJ, Wu AW, Sexton JB. Acute decompensation after removing a central line: Practical approaches to increasing safety in the intensive care unit. Ann Intern Med. 2004;140:1025-1033.

Return to table of contents