FDA Home Page | CDRH Home Page | Search | CDRH A-Z Index | Contact CDRH

Human Factors and the FDA's Goals: Improved Medical Device Design

Introduction

As a practicing physician, I can relate and remember my own experiences involving use error. For most of my career, I have worked in the emergency room part time. One of my earliest experiences with use error made such an impression I submitted a report of it to the Journal of the American Medical Association.'

We admitted a man in his sixties experiencing myocardial infarct. I ordered a dose of prophylactic lidocaine to reduce the ventricular tachycardia and fibrillation. This medication is usually administered in a 100-mg intravenous bolus, but the nurse accidentally picked up a prepackaged syringe of 2,000 mg of lidocaine and injected it instead. The patient had a convulsion and then became astolic. Although we tried for more than 45 minutes to revive him, he died.

In this case, the design and packaging of two very different doses of medication were so similar, it was easy to see why a user could pick up the wrong syringe and administer a lethal dose. Two other cases of accidental overdose of this medication were reported as well. Because all of these incidents were reported to the Bureau of Drugs, the product was changed to avoid future errors in administering lidocaine.

Purpose

Industry, the medical community, and the Food and Drug Administration (FDA) need to better identify and define the problem of human factors to the public, both in broad terms and with respect to specific devices in specific situations. We need to focus our efforts where we can have the greatest impact on health. For many types of devices, poor human factors designs are responsible for more accidents than mechanical or electrical failure.

For example, many years ago, FDA determined that the unexplained anesthesia deaths of several healthy patients were caused by a sticky valve that administered an overdose of anesthetic gas to patients. We marshalled many resources to analyze this problem and developed an interim "fix" so this type of machine could be used while the manufacturer produced a new component that would work properly.

This mechanical failure caused several deaths; however, use error in anesthesia results in more than 100 deaths each year. While there has been a tremendous improvement over the past decade, we can do more. We can have a much greater impact on public health if each of us reduces the use errors in this one specialty-that is, anesthesiology-than if we put all our efforts into the fine tuning of design features for anesthesia machines. And of course, this isn't the only medical area.

The Harvard Medical Practice Study reported results of a study on patient injuries in New York hospitals during 1984. According to the study, nearly 4% of patients suffered unintended injuries caused by treatment they received while in the hospital. If the rates of unintended injury are similar in other states, then more than 1.3 million people in the United States are hurt each year while they are hospitalized.2 Although nearly half of these injuries are drug related, many are related to the use of medical devices.

The more complex the specialty, the greater the likelihood of accidental injury to the patient. According to Lucian Leape, the Harvard Medical Practice Study found that the highest rate of adverse events occurred in vascular surgery, cardiac surgery, and neurosurgery.2

What is the lesson from all of this? First, we have already put much effort into the design, composition, and testing of individual products with the result being a low incidence of device failure. But this is occurring against a background of disease outcome, medical care availability, and medical misadventure, that makes problems with device product failure pale in comparison.

We now need to take a hard look at medical misadventure and see where product operation interface factors are contributing to the likelihood of a bad outcome. Already, some documentation of the problem exists, but many of us believe that explicit problem scoping and statistical evaluation draw more attention to this problem.

Incentives

Furthermore, we must show health care facilities the incentives for addressing these problems. Unfortunately, humanitarian incentives are not a sufficient motivator in our society. We need to show the economic impact of accidental injuries caused by use error and the economic benefits of preventing them. So far, there has been little financial incentive for hospitals to examine use error from a systems perspective and to put resources into correcting these problems. Currently, additional treatment required by accidental injuries is simply added to the patient's bill. If hospitals had to take care of these complications for free, they might have some economic incentive to resolve problem of use error.

In the current health care setting, Health Maintenance Organizations and hospitals operating under medical payment schemes are the facilities most likely to respond to use error as an economic issue. The key for them is cost containment-they don't want to pay for unneeded complications. We must show them it is worth their while to take a systems approach to unintended injuries and to determine how to prevent problems with user device interface and inappropriate therapy.

Even if institutions become conscious of the extra treatment costs they can prevent, they will still not be giving proper weight to the human impact of use errors. For the patient who is accidentally injured through use error, there is no way to alleviate the pain, suffering, and loss of time from job and daily life. The only current financial incentive to an institution to eliminate use errors is the fear that the patient will sue. While patients may receive economic compensation, the problem that resulted in the injury is not resolved. Those of us in the public sector need to show there is a realistic expectation that intervention will reduce the impact of the problem.

Methods

How do we know we can do something about use error? We know because the Center for Devices and Radiological Health (CDRH) has a long history of involvement in human factors design. For many years, we've been involved in human factors issues associated with administering anesthesia. Currently, we're involved in looking at the next generation of anesthesia machines and making suggestions about alarm systems, displays, and other aspects of the operator-device interface. These can make a difference in lowering use error.

CDRH staff have also worked to improve the use of the self-blood glucose monitor through better human factors. By 1989, we had received 2,200 reports of erroneous test results from self-monitoring of blood glucose. As you know, problems with such test results could have severe repercussions for insulin-dependent diabetics. Through analysis of the use and design of several models of monitors, we found that often the design didn't take into consideration the age and manual dexterity of most patients, or the visual problems that many diabetics have.

Hemodialysis is another area in which human factors can make a difference. In 1987, FDA, industry, and users developed a videotape called "Human Factors in Hemodialysis" on mishaps attributed to use error. Problems included improper temperature settings, incorrect priming of blood lines, and poor design for ease of cleaning the equipment.

In addition to projects focused on improving the human factors of a particular device, FDA has sent several safety alerts to the medical community to warn of accidental injuries resulting from inappropriate use or use error. For example, one of these alerts concerned human factors problems with infusion pumps; another concerned use problems with apnea monitor leads.

Program Elements

CDRH plans to pay more attention to human factors to ensure that devices are designed to met the needs of users in order to minimize use error and patient injuries that come from use error. This human factors program will be designed to address three basic elements.

First, errors in the use of medical devices are a significant cause of patient deaths and injuries. In 1985, a survey of nearly 300 private anesthesiologists found that 24% admitted to making an error that had lethal consequences. After all, medical professionals are human, and human beings are fallible no matter how they strive for a perfect practice.

Many factors in the medical environment increase the chance for error, including inadequate work space and inadequate work layout. For example, many use errors with anesthesia systems seem to result from the configuration of the various components. FDA has worked with industry and anesthesia professionals to develop a checklist to ensure that components are set up properly. The first checklist was developed in 1987, and a revised checklist was completed in 1992. Other medical setting factors that increase the chance for error include poor environmental conditions, poor supervision, and stress.

Errors seem to be greatly under-reported, most likely due to liability concerns. We need to take a systems approach to use error, not a who's-to-blame approach. The potential for developing a solution to use error is lost if the medical community, manufacturers, and government are unaware of the problems.

A second theme of FDA's program is that device design and labeling dictates how users interact with devices and influences the type and frequency of use error.

According to one report, the probability of omitting one task out of five decreases immensely when written procedures are used, and procedures with checkoff provisions reduce the potential for error even more. The design of labeling is integral to how a device works----actually, labeling is part of the equipment design. For example, many problems with blood glucose meters have been attributed to failure to follow directions. But field testing with diabetics showed that, in many cases, the labeling gave inadequate instructions for proper use of the meters.

At the same time, improvements in labeling cannot compensate for poor design. Again, blood glucose meters are a good example. Although poor labeling was part of the problem, the design of many meters contributed to use error as well. The designs failed to take into account the stages of the disease of the user. Tiny test strips had to be inserted into a machine by patients who were losing their fine motor skills. Many models displayed results with tiny numbers, even though visual problems are also a part of diabetes.

In addition to labeling, packaging may also be a problem. One type of blade used in intubation does not release easily from its packaging, yet this device is frequently used in the emergency room where it is needed in a hurry.

Third, FDA's program is based on the belief that proper attention to human factors principles and practices can help reduce, and in some cases prevent, use errors by ensuring that device features and operating characteristics do not invite or allow error.

Operating a medical device should be almost intuitive. Many hospitals are equipped with several brands or models of infusion pumps. A nurse should be able to use all of them properly without referring to a user's manual. A pump with a complex, multistep program is more likely to be programmed incorrectly because a nurse does not have time to figure out the difficult procedure. Use error is even more likely when such a pump is used in the home and is programmed by a patient's family.

As a device is developed and manufactured, it should meet good manufacturing practices (GMP) requirements. The new version of the GMPs emphasizes preproduction quality design, and human factors should be one of the considerations in designing a quality device.

Analysis and testing during the design process is critical. Use of simulations, mockups, and prototypes can provide insight into potential errors under real-use conditions. It is difficult to do actual "in use" testing, as it is intrusive to medical care in most cases; however, consumer testing is acceptable.

Conclusion

We know there are great benefits to be had if human factors can be designed and manufactured into medical devices. There is no doubt that customer satisfaction will improve. Health care professionals want to give patients the best possible treatment, and any product that minimizes error and accidental injury is a better product to use. Fewer errors will also reduce product and practice liability----another important consideration. For the manufacturer, sales of a device designed with user/machine interface in mind are likely to increase. And most importantly, human factors considerations result in safer devices and fewer injured patients. We can all appreciate that.

Despite our experience in this area, the CDRH's human factors program is at an early stage of development. We want to hear views from industry, the health care community, and others as we formulate our program. But tackling the problem of use error is not simply a government issue----it is an issue for the entire medical community. We invite you to join us in making health care safer for every patient.

References

1. BURLINGTON, B. and FREED, C. Massive overdose and death from prophylactic lidocaine. JAMA, 1980, vol. 243, p. 1036-1037
2. LEAPE, L. The Preventability of Medical Injury. In: BOGNER MS (ed). Human Error in Medicine. Hillsdale, NJ: Laurence Erlbaum Associates, 1994

Updated December 3, 1996

CDRH Home Page | CDRH A-Z Index | Contact CDRH | Accessibility | Disclaimer
FDA Home Page | Search FDA Site | FDA A-Z Index | Contact FDA | HHS Home Page

Center for Devices and Radiological Health / CDRH