Time to take a breath: Rebalancing the human and technical aspects of patient safety

It used to be a familiar joke: all doctors have horrible handwriting. The subsequent illegible prescriptions left pharmacists bewildered, and sometimes, the wrong medication was dispensed.

This problem led to a key recommendation in 2006 from the Institute of Medicine’s Preventing Medication Errors report that all healthcare organizations should implement systems for electronic prescribing, and that all pharmacies should be able to receive them. ¹

Today, physicians don’t need to write much of anything. All notes and orders are entered electronically. Even my admittedly illegible signature has been captured electronically and is applied when I enter a password. For the most part, this is a good thing for patient safety. The calls from the pharmacist, which used to occur on a routine basis to request clarification, have been driven nearly to extinction.

But occasionally, the computerized prescription order entry creates its own problems. Recently, by accident I wrote a series of prescriptions for the wrong patient, because I had two patient notes open on my screen at the same time. The electronization of medicine has led to a generation of new problems, and the creation of new kinds of medication errors.

Still, electronic prescribing has virtually eliminated doctor handwriting as a source of medication errors. In that case, and others, like the replacement of paper records, it provides the solution. Electronic health records are beginning to permit health information exchange across the delivery system. But information technology alone is not the answer to many of the most important problems with patient safety and healthcare quality. Technological innovations need to be balanced with a greater appreciation of the human element.

Healthcare is a sociotechnical system that involves constant interaction between people and technology, as well as how those people interact with the complex infrastructure of health systems. ^{2 , 3} A common problem today is the lack of recognition that this is the case. Healthcare managers tend toward a “build it and they will come” attitude to new technology. This underestimates the time, commitment, testing, and adjustments required for people to adapt and become expert on, for example, a new electronic health record. An important element of sociotechnical systems is that optimizing the technical alone has a detrimental effect on performance.

In the science of safety, we have reached the point where we need human solutions to take full advantage of the advances we have made. We know some things that work, particularly in areas like infection prevention, and teamwork. The next step is to work on implementation, to ensure that procedures known to be effective are applied reliably to every patient who needs them. This entails engaging practitioners as well as patients and families.

One of the biggest challenges facing patient safety is how to engage healthcare workers to do things that we already know to be effective. Examples include consistent hand hygiene, use of the safe surgery checklist, and listening to the patient long enough to get the information needed to help the patient.

It is settled science that consistent handwashing is essential to reducing healthcare acquired infections, and that alcohol-based hand rub is generally superior to soap and water for the cleaning hands. ⁴ Despite this, it is a persistent challenge for hospitals and healthcare organizations to achieve adequate compliance with hand hygiene guidelines. At Johns Hopkins, it took over a decade of coaxing, nudging, and feedback to obtain levels above 90% in all departments.

The initial publication of the WHO safe surgery checklist indicated that implementation of the 19-item checklist in eight countries was associated with significant reductions in mortality and inpatient complications. ⁵ In 2014, Urbach et al. reported the surprising finding that implementation of the surgical safety checklist in over 100,000 cases in Ontario, Canada, was not associated with significant reductions in operative mortality or complications. ⁶ However, in an accompanying editorial, Leape suggested the likely reason for this failure was that the checklist was not actually used. ⁷ He recommended local adaptation, training staff on the new workflow, and involving the entire healthcare team.

Wrong patient errors are a well-established problem in healthcare, with dramatic consequences in wrong site, wrong procedure and wrong person surgery, and more frequent medication administration errors. ¹ Bar-code technology of patients and medications has been shown to reduce medication administration errors and adverse drug events. ⁸ However, this innovation gave rise to a new crop of workarounds aimed to compensate for the extra time needed to implement the technology. ⁹ We once observed a large collection of patient wristbands looped onto a medication cart – when asked, the nurse explained that they had been removed from the respective patients in order to increase efficiency. In cases like this and others, workarounds have been associated with medication administration errors, leading to the recommendation that more post implementation evaluation is needed if it is to achieve the intended benefits. ¹⁰

In this issue, several papers address topics related to the adoption and integration of information technology-based solutions into clinical practice to improve quality and safety, and the relative contributions of IT-based vs. human factors.

Yin et al. were confronted with the perennial problem of getting clinicians to clean their hands consistently to prevent healthcare acquired infection. ¹¹ They recognized the utility of strategically placed just-in-time reminders to cue clinicians to do the right thing. Although sophisticated interventions have been developed like electronic hand hygiene monitoring systems, they opted for a much lower tech solution of marking zones and demonstrated its effectiveness. ^{12 , 13}

Miller et al. review the literature on the effectiveness of mindfulness-based interventions – a distinctly non-technical solution – to reduce stress, burnout, and increase self-compassion in healthcare workers who have become second victims. ¹⁴

Adams et al. analyzed a large database to investigate how healthcare systems handled patient safety events related to Health Information Technology. ¹⁵ They found that nearly two-thirds of events had not been resolved at the time of reporting. Perhaps more important, they noted that the majority of the recommendations were for training and education of providers, types of solutions known to have limited impact.

Guttman et al. describe how a strategy based on error management theory – an approach for embedding the preoccupation with failure and commitment to resilience that are central to high reliability theory – can be adopted by healthcare organizations. ¹⁶ The initial focus is on recognizing errors in others in oneself, the latter in particular being a highly reflective activity. However, technology-based solutions, including simulation, can be helpful tools in applying the paradigm.

It may be time to take a breath and consider the human solutions that are needed around the advances being introduced into our health systems. Some of these are in health information technology, while others involve novel treatments and procedures. Some are even designed specifically to improve patient safety.

This would require a better understanding of how the humans who work in healthcare interact with the different elements of the healthcare system, including health information technology and tools, but also with the task to be performed, organizational policies and culture, and the environment. Human factors and ergonomics experts are trained expressly for the purpose of gaining understanding in these areas, but to date, relatively few of these individuals are employed by healthcare organizations.