In My Opinion
Human Factors Researcher Alarmed by Deaths During PCA
For approximately 17 years, I have been conducting research in human factors engineering mainly in the process control industries. Thanks to the prompting of Dr. John Doyle, an anesthesiologist and biomedical engineer, I and my colleagues (particularly Ms. Laura Lin) have been conducting research on how to improve the design of medical devices from a human factors engineering point of view since 1992. From the beginning, the Lifecare 4100 PCA infusion pump was selected as a testing model for our research. Our interest was to show the value of human factors in medical device design in general and not to selectively target a particular device or manufacturer. During our 8 years of research, we have become quite familiar with the human-computer interface for operating the Lifecare 4100.
Given the information provided in safety alerts and reports over the past 3 years1, 2, I view the Lifecare 4100 to be an example of inadequate human factors engineering. The Lifecare 4100 can be operated in one of three configurations, one that is pre-activated by the manufacturer before the product is shipped and two that can be activated by hospital biomedical engineering departments after they receive the product. When nurses begin to program the drug concentration in one of the latter two configurations, the Lifecare 4100 screen initially shows a concentration value of 0.1 mg/ml. Users can either accept this initially displayed value or they can modify it by using the arrow controls on the pump. In my opinion and that of others1,2, the safety-critical issue with the design is that the Lifecare 4100 offers this minimum morphine drug concentration as the first choice. If users mistakenly accept the initially displayed minimum morphine value of 0.1 mg/ml instead of changing the concentration setting to a correct (and higher) value (1.0 mg/ml and 5.0 mg/ml are common concentrations of morphine), then the machine will “think” that the drug it is to infuse is less concentrated than it truly is. The result would be an erroneously large liquid volume (and total morphine dose) being administered to the patient (e.g., 10 or 50 times the intended amount). Indeed, several deaths during the use of Lifecare 4100 have been reported.1, 2 In response to these reports, researchers have emphasized the importance of improving safety by incorporating human factors engineering principles into the design of medical devices.3
From a human factors standpoint, the concern is that the user of the Lifecare 4100 who erroneously accepts the first concentration presented can set in motion a series of steps that may result in the delivery of a morphine overdose to the patient. The position that “An active decision is made… This ensures people are looking at the drug type, the drug concentration”4 is inconsistent with existing scientific knowledge in human factors engineering. Because there is no simple correspondence between overt behavior and mental processes, the fact that someone presses a button does not ensure anything about what they have thought about or looked at.5 Furthermore, from a human factors engineering viewpoint, a position that “one of the keys to a successful patient-controlled analgesia (PCA) program is a well-educated clinical staff knowledgeable in PCA delivery”6 is necessary but not sufficient. There is overwhelming evidence to show that well-intentioned, well-educated, and attentive individuals inevitably sometimes make mistakes. This is true, no matter how much training they receive. More or better training will not completely compensate for inadequate attention to human factors engineering.
As the recent Institute of Medicine report emphasizes, “to err is human.”7 Medical professionals are not immune from making mistakes. Effective design of medical equipment recognizes this human frailty. Furthermore, effective equipment design attempts to minimize the consequences of errors that fail to be identified by the user. In this regard, I believe that the initial presentation of the minimum medication concentration value can have an effect opposite to that desired. Admittedly, the potential negative consequence of this design feature is counterintuitive and it is understandable that it could have been initially overlooked or not appreciated. However, once the potential problem is recognized and reported1,2,3,4, there are compelling reasons to activate existing mechanisms (i.e., response of the manufacturer and/or Food and Drug Administration) for avoiding future “accidents.” I am concerned that these existing mechanisms have not appreciated or recognized the human factors engineering design problem in the Lifecare 4100. As a result, patients remain at risk for accidental drug overdoses and associated morbidity and even mortality.
I hope that this experience with Lifecare 4100 might stimulate the Food and Drug Administration to adopt an increasing interest in the need to regulate the human errors that may be introduced by medical equipment. Patient safety cannot be achieved merely by ensuring that medical devices are safe when operated as intended. Human factors engineering should be a federally regulated aspect of design in medicine, as it is in other safety-critical industries (e.g., nuclear power and commercial aviation). For medical device designers, it is important to be aware of the counterintuitive relationship between minimum drug concentration value and patient safety in this, and perhaps other medical devices. The commitment of medical device manufacturers to patient safety should be rewarded for being willing and able to learn from unforeseen design flaws and make subsequent changes to better ensure patient safety.8
Dr. Vicente is Professor and Director, Cognitive Engineering Laboratory, Department of Mechanical and Industrial Engineering, University of Toronto, Ontario, Canada.
1. ECRI. Abbott PCA Plus II patient-controlled analgesia pumps prone to misprogramming resulting in narcotic overinfusions. Health Devices 1997;26(10):389-391.
2. Institute for Safe Medication Practices. Evidence builds: Lack of focus on human factors allows error-prone devices. ISMP Medication Safety Alert 1999; 4(15).
3. Weinger MB, Pantiskas C, Wiklund ME, Carstensen P. Incorporating human factors into the design of medical devices. JAMA 1998; 280: 1484.
4. St. John P. Drug pump’s deadly trail. Tallahassee Democrat, Sunday, May 28, 2000. Available URL: www.tdo.com/news/local/0528.loc.pump.htm.
5. Norman, DA, Categorization of action slips. Psychological Review 1981; 88: 1-15.
6. Letter dated October 29, 1997 from the Hospital Products Division of Abbott Laboratories.
7. Institute of Medicine. To err is human: Building a safer health system. 1999. Washington, DC. National Academy Press, 1999.
8. Leape LL. Error in medicine. JAMA 1994;272:1851.