Chapter 7. Human Factors and Errors at the Person–Machine Interface

Until now, we have discussed several paradigm shifts required to improve patient safety. The dominant one, of course, is replacing an environment based on “blame and shame” with one in which safety is viewed as a top priority and systems thinking is employed effectively. Second is an awareness of the impact of culture and relationships on communication and the exchange of information. This chapter will introduce another lens through which to view safety problems: how human factors engineering (HFE) can improve the safety of person- machine interactions and the environment in which healthcare providers work.

To prime ourselves for a discussion of HFE, let us consider the following scenarios:

• An obstetric nurse inadvertently connects an opiate pain medication intended for an epidural catheter into a mother's intravenous line, leading to the patient's death.1 A subsequent review demonstrated that the bags and lines used for epidural and intravenous infusions were similar in size and shape, leaving nothing other than human vigilance to prevent a bag intended for epidural use from connecting to an IV catheter or hub.
• A hospitalized elderly man dies after a heart attack suffered while in a monitored bed in one of America's top hospitals. A later investigation reveals that the main crisis monitor had been turned off, and multiple lower level alarms—including ones that showed that the patient's heart rate was slowing dangerously for nearly half an hour before his heart stopped—weren't noticed by the busy nursing staff, who had become so inured to frequent false alarms that they suffered from “alarm fatigue.”2
• Modern multichannel infusion pumps are routinely used in the ICU to administer multiple medications and fluids through a single central line (Figure 7-1). This often results in a confusing tangle of tubes that cannot be easily differentiated. No surprise, then, that a busy ICU nurse might adjust the dose of the wrong medication.
• Medications are often stored in vials in doses that are highly concentrated. This means that they frequently need to be carefully diluted before being administered. For example, a vial of phenylephrine contains 10 mg/mL, while the usual IV dose administered to patients is 0.1 mg—one-hundredth of the dose in the vial! Inadvertent administration of full strength phenylephrine can cause a stroke.
• An elderly patient dies when a Code Blue team—responding to the call for help after a cardiac arrest—is unable to connect the defibrillator pads to the defibrillator.3 A subsequent analysis showed that over the years the hospital had accumulated more than a dozen defibrillator models on its floors, so providers were often unfamiliar with the models at hand and incompatibilities were common.4

In each of these examples, significant hazards resulted from people interacting with products, tools, procedures, and processes in the clinical environment. One could argue that these errors could have been prevented by more careful clinicians or more robust training. However, as we have already learned, to minimize the chances that fallible humans (in other words, all of us) will cause patient harm, it is critical to apply systems thinking. In the case of person–machine interfaces, this systems focus leads us to consider issues around device design, the environment, and the care processes that accompany device use. The field of HFE provides the tools to accomplish this.

This chapter was coauthored by Bryan Haughom, MD.

Human factors engineering is an applied science of systems design that is concerned with the interplay between humans, machines, and their work environments.5,6 Its goal is to assure that devices, systems, and working environments are designed to minimize the likelihood of error and optimize safety. As one of its central tenets, the field recognizes that humans are fallible and that they often overestimate their abilities and underestimate their limitations. Human factors engineers strive to understand the strengths and weaknesses of our physical and mental abilities and use that information to design safer devices, systems, and environments.

HFE is a hybrid field, mixing various engineering disciplines, design, and cognitive psychology. Its techniques have long been used in the highly complex and risky fields of aviation, electrical power generation, and petroleum refining, but its role in patient safety has only recently been appreciated.7–9 In applying HFE to healthcare, there has been a particular emphasis on the design and use of devices such as intravenous pumps, catheters, computer software and hardware, and the like.

Many medical devices have poorly designed user interfaces that are confusing and clumsy to use.10,11 According to the U.S. Food and Drug Administration (FDA), approximately half of all medical device recalls between 1985 and 1989 stemmed from poor design. FDA officials, along with other human factors experts, now recognize the importance of integrating human factors principles into the design of medical equipment.11–13

In Chapter 2, I introduced the concept of forcing functions, design features that prevent the user from taking an action without deliberately considering information relevant to that action. The classic example of a forcing function was the redesign of automobiles that prevented cars from being placed in reverse if the driver's foot was off the brake. In healthcare, forcing functions have been created to make it impossible to connect the wrong gas canisters to an anesthetized patient or prevent patients from overdosing themselves while receiving patient-controlled analgesia (PCA). Although forcing functions are the most straightforward application of HFE, it is important to appreciate other healthcare applications, ranging from improving device design to aiding in device procurement decisions to evaluating processes within the care environment.

For example, many hospitals are now approaching the challenge of increasing the frequency with which providers clean their hands partly as a human factors problem (Chapter 10).14,15 While these institutions continue to work on education and observation, they also ensure that cleaning gel dispensers are easy to use and strategically located throughout the hospital wards. In fact, a whole field of patient safety–centered hospital and clinic design has emerged, and some buildings have been constructed using human factors principles.16,17

Despite these early success stories, HFE remains conspicuously underused as a patient safety tool, for reasons ranging from the lack of well-defined avenues to report and correct design or process flaws within hospitals to the natural tendency of highly trained caregivers to feel that they can outsmart or work around problems.9 With the dramatic growth in the volume and complexity of man–machine clinical interactions, the probability that human workers will make mistakes has escalated, as has the importance of considering HFE approaches.

One of the key tools in HFE is usability testing, in which experts observe frontline workers engaging in their task under realistic conditions—either actual patient care or simulated environments that closely replicate reality. Users are observed, videotaped, and asked to “talk through” their thought processes, explaining their actions as well as their difficulties with a given application or device. Engineers then analyze the data in order to fine-tune their design for the users, the chosen tasks, and the work environment.18

Software engineers and design firms now see usability testing as an indispensable part of their work, preferring to make modifications at the design stage instead of waiting until errors have become apparent through real-world use. Similarly, many equipment manufacturers and a growing number of healthcare organizations now employ individuals with human factors expertise to advise them on equipment purchasing decisions, modify existing equipment to prevent errors, or identify error-prone equipment and environmental situations.8,10,11 These trained individuals instinctively approach errors with a human factors mindset, asking questions about usability and possible human factors solutions before considering fixes involving retraining and incentives, interventions that may seem easier than device or environmental redesign but are generally far less effective.

Usability testing can be a complex process, requiring not only human factors experts but also extensive investigatory time and cooperation from users. A less resource-intensive alternative is known as heuristic analysis.11 The term heuristics was first mentioned in Chapter 6 in reference to the cognitive shortcuts that clinicians often take during diagnostic reasoning, shortcuts that can lead to errors. In the context of HFE, though, heuristics have a different connotation: “rules of thumb” or governing principles for device or system design. In heuristic evaluations, the usability of a particular system or device is assessed by applying established design fundamentals such as visibility of system status, user control and freedom, consistency and standards, flexibility, and efficiency of use (Table 7-1).18,19

During a heuristic evaluation, experts navigate the user interface searching for usability issues. In essence, analysts try to put themselves in the shoes of the end user, looking for error-prone functions or designs that might ultimately compromise safety. The information gleaned from these analyses becomes feedback to the design teams, who iteratively update and refine the design of prototypes. The ultimate shortcoming of heuristics, however, is that they are only as good as the analyst performing the evaluation. Without putting end users in real-time situations and observing their work, it is difficult to fully unearth problematic design issues. Nevertheless, heuristic evaluations can provide valuable information to a design team, or even to a hospital looking to review devices or systems they already own or are looking to purchase.8,11

Inability to truly stand in for the novice is only one problem that designers and engineers have in anticipating all the safety hazards of complex systems. In my interview with him, Donald Norman, the author of the best-selling book, The Design of Everyday Things,20 reflected on another real-world human factors issue:

[As an engineer], you focus too much, and don't appreciate that all the individual elements of your work, when combined together, create a system—one that might be far more error-prone than you would have predicted from each of the individual components. For example, the anesthesiologist may review beforehand what is going to be needed. And so he or she picks up the different pieces of equipment that measure the different things, like the effects of the drugs on the patient. Each instrument actually may be designed quite well, and it may even have been rigorously tested. But each instrument works differently. Perhaps each has an alarm that goes off when something's wrong. Sounds good so far. But when you put it together as a system it's a disaster. Each alarm has a different setting, and the appropriate response to one may be incredibly dangerous for another. When things really go wrong, all the alarms are beeping and the resulting cacophony of sounds means nobody can get any work done. Instead of tending to the patient, you're spending all of your time turning off the alarms. So part of the problem is not seeing it as a system, that things have to work in context. And that these items actually should be talking to each other so that they can help the anesthesiologist prioritize the alarms.21

The difficulty of anticipating all these interactions and dependencies is yet another powerful argument for usability testing, not only of individual devices but also of multiple devices as they interact in actual systems.

In addition to their use in the design and purchase of medical devices such as ventilators, programmable IV pumps, defibrillators, and anesthesia equipment, human factors principles should also be utilized in the design and implementation of advanced information technology systems (Chapter 13). With the increasing computerization of healthcare, this interface—between HFE and information technology—is becoming ever more critical, and the growing list of unintended consequences of information technology systems is partly an indictment of our lack of focus on this area.22,23

The consequences go well beyond inefficiency and workarounds. One prominent pediatric teaching hospital reported an increase in mortality rates after a commercial computerized provider order entry (CPOE) system went live.24 Although the findings are controversial and a subsequent study at another hospital (using the same CPOE system) found improved mortality,25 the authors' description of the problems is a useful cautionary note regarding the hazards of poor design and implementation. For example, they observed:

Before CPOE implementation, physicians and nurses converged at the patient's bedside to stabilize the patient. After CPOE implementation, while one physician continued to direct medical management, a second physician was often needed solely to enter orders into the computer during the first 15 minutes to one hour if a patient arrived in extremis. Downstream from order entry, bedside nurses were no longer allowed to grab critical medications from a satellite medication dispenser located in the ICU because as part of CPOE implementation, all medications, including vasoactive agents and antibiotics, became centrally located within the pharmacy department. The priority to fill a medication order was assigned by the pharmacy department's algorithm. Furthermore, because pharmacy could not process medication orders until they had been activated, ICU nurses also spent significant amounts of time at a separate computer terminal and away from the bedside. When the pharmacist accessed the CPOE system to process an order, the physician and the nurse were “locked out,” further delaying additional order entry.24

Human factors techniques can also be used to increase the safety of a working environment, such as by standardizing devices within a hospital to make training easier and increase reliability, carefully designing equipment and processes in a pharmacy to ensure that pharmacists always dispense the right drug in the correct dose, and optimizing a clinical work environment by providing adequate lighting, eliminating distractions (including excess noise), and ensuring that care providers receive adequate rest.

In the end, usability testing and heuristic analysis can be used together or separately to design safe, effective, and intuitive devices for use in complex clinical care situations. Let's now return to the clinical scenarios outlined at the beginning of the chapter. How could HFE principles be applied to prevent these unsafe conditions? Here are some approaches human factors engineers might consider for each scenario:

• To prevent an epidural–intravenous mix-up, a forcing function could ensure that medications intended for epidural use are physically incapable of connecting to intravenous hubs.26,27 An effective bar-coding system is another obvious solution. (Tragically—and ironically—the hospital where this error occurred had a bar-coding system, but its usability and reliability problems were well known and the nurses had learned to use workarounds, particularly in urgent situations.1,28)
• Alarm fatigue can be mitigated by careful selection of equipment and by implementing policies and procedures developed after consultation with nurses and other caregivers and detailed observations of workflow.2,29
• By employing standardized color coding and other techniques, nurses using multichannel infusion pumps can more easily differentiate IV lines.30
• Pharmacies can be equipped with high-reliability robotic systems that produce bar-coded unit-dose medications that can be double checked using bar-coding devices at the bedside before they are administered to the patient. Multidose vials are inherently risky and should be avoided where possible.31
• The hospital where this error occurred standardized its defibrillators (buying more simple, “goof-proof” ones) and increased the frequency of testing.3,4,32

A dramatic case study in the application of HFE to another patient safety problem, adapted from the book Set Phasers on Stun,33 is described in Box 7-1.

• HFE is the applied science of systems design. It is concerned with the interplay of humans, machines, and their work environments.
• Thoughtful application of HFE principles can help prevent errors at the person–machine interface.
• Usability testing and heuristic analysis aim to identify error-prone devices or systems before they lead to harm.
• The initial focus of HFE in healthcare was on the design of medical devices. Today it is broader, with efforts to try to create safe environments of care (i.e., designing hospital rooms around safety principles) and to decrease errors associated with poorly designed clinical information systems.