Applying Human Factors to Make it Easier for Clinicians to Do Their Jobs

“The majority of my research portfolio are implementation-related projects, which nicely aligns with AHRQ’s mission of practical, applied research. It’s not research for research’s sake. We’re improving patient safety and quality at the point of care. Whether it’s building an opioid tapering app or trying to improve misdiagnosis of cardiovascular disease in women, these are projects bringing real change. And AHRQ is the agency supporting that work.”

AHRQ grantee Kristen E. Miller, Dr.P.H., M.S.P.H., M.S.L, admits she didn’t know much about human factors and systems engineering when she arrived at Texas A&M University to pursue graduate public health training in occupational health and safety.

But rubbing shoulders with industrial engineers studying safety in the oil and gas industries helped her quickly make the connection between human factors and patient safety. “This discipline is always anticipating failure, looking for things to go wrong, and applying a systems lens, attributes that fit naturally into improving patient safety,” said Dr. Miller. She is the senior scientific director of the National Center for Human Factors in Healthcare at MedStar Health, a 10-hospital system and the largest healthcare provider in the Maryland and Washington, D.C. area.

Applying a human factors perspective to patient safety, while a complex undertaking, gives Dr. Miller’s work a clear goal. “I like to think it’s my job to make it easier for clinicians to do their jobs,” she said. “Patients are at the center of all of that, too,” she adds, with their feedback on how they interpret risk and other variables that could lead to a missed, incorrect, or delayed diagnosis.

Dr. Miller’s current AHRQ-funded project is tackling a confounding problem: why women have higher rates of diagnostic error, treatment, and miscommunication for cardiovascular disease (CVD). Compared with men, women are nearly twice as likely to receive the wrong diagnosis following a heart attack and 30 percent more likely to have stroke symptoms misdiagnosed in the emergency room.

Dr. Miller’s project is addressing this issue on multiple fronts, leveraging AHRQ’s Patient Safety Learning Laboratory model. Working with cross-disciplinary input by researchers, clinicians, and patients, the team is evaluating physical environments, social and economic determinants, clinical management, health information technology,  and other factors that can increase a woman’s risk of a CVD diagnostic error. Solutions—some created with patient input—currently are being identified, and then will be tested in simulated environments and pilot-tested in ambulatory care practice settings.

The potential for a CVD diagnostic error can begin as soon as women describe their symptoms, according to Dr. Miller. “Patient-facing education materials still describe heart attack onset symptoms in the way we would expect for men,” such as crushing chest pain. Some updated materials may include symptoms more commonly described by women, such as pain in the neck and shoulders, nausea, or dizziness.

But even an expanded list of symptoms can create a certain level of ambiguity, Dr. Miller notes. For example, one professional group recently labeled symptoms more commonly affecting women as “atypical,” she said. “No, those symptoms are not atypical, they are just how women present symptoms.”

As part of the project, Dr. Miller’s team is studying the role of the ambulatory care environment in elevating the risk of a CVD diagnostic error. “We continue to hear about trust and communication [and] that patients feel their doctor didn’t hear them. So how can we help patients be better advocates for themselves? How do you design the layout of a patient room to facilitate shared decision-making and communications?” The project also will identify and share examples of positive uses of physical space that improve patients’ visits.

Data that clinicians use to help them predict CVD risk itself can be a contributing factor to diagnostic error for women. That’s because the widely used Framingham risk score—an algorithm used to estimate a person’s 10-year cardiovascular risk—is based on a long-term ongoing study of mostly older, white men, Dr. Miller notes. “How does that account for cardiovascular risk factors in young, Black female patients?” she asks. A broader, systems-based approach to addressing CVD risk factors also might include assessing whether patients have access to health-promoting options, such as local walking paths for exercise and convenient access to pharmacies and health facilities.

While a human factors engineering approach cannot identify an overall formula for reducing CVD diagnostic error in women, Dr. Miller is confident that examining multiple factors in combination will yield solutions that health providers can adapt to their needs. The project is scheduled to conclude in September 2023.

In a recently completed AHRQ ACTION (Accelerating Change and Transformation in Organizations and Networks) grant, Dr. Miller, her team at MedStar Health, and technical developers Vessel Partners used human factors engineering methods to develop clinical decision support systems to help patients safely taper from long-term opioid use.

“Opioid tapering is a highly sensitive subject,” Dr. Miller said, recalling interviews with clinicians and patients. Clinicians don’t like to bring up the subject, even when tapering is indicated, as it can raise patients’ fears of withdrawal and lead them to stop seeking care. "Patients were fearful they’d get a new, lower-strength prescription, and worried there would be a lack of communication about their pain prior to a follow-up visit,” she said.

Designed with input from patients and clinicians, the team developed an app for clinicians that identifies patients at high risk of harm from opioids and provides evidence-based guidelines to support opioid tapering. The app includes a calculator that displays the recommended drop in opioid-dose reduction, which can help patients understand the gradual approach with tapering. Patients receive weekly text messages that ask about their pain level, how it varies over time, and offers access to non-drug alternatives, such as physical therapy and yoga.

The app was tested at 15 primary care practices, but widespread use was hampered by clinicians’ time constraints imposed by COVID-19 protocols. “The infrastructure is built, and it’s ready to be deployed by EHR vendors,” Dr. Miller said.

In addition to her position at MedStar Health, Dr. Miller is an Associate Professor of Emergency Medicine at Georgetown University School of Medicine. She is a Certified Professional in Patient Safety, credentialed by the National Patient Safety Foundation.

Related AHRQ Resources

• Clinical Decision Support
• Opioids and Substance Use Disorders

Principal Investigator: Kristen E. Miller, Dr.P.H., M.S.P.H., M.S.L.
Institution: MedStar Health, Washington, D.C.
Grantee Since: 2018
Type of Grant: Various

Return to AHRQ Grantee Profiles