Look at Patient Safety Through an Equity Lens, Expert Urges

Health equity should be part of every aspect of a healthcare facility’s patient safety planning, Tejal Gandhi, MD, MPH, said Monday at the annual meeting of the Society to Improve Diagnosis in Medicine.

“There’s no such thing as high-quality, safe care that is inequitable,” said Gandhi, who is chief safety and transformation officer at Press Ganey Associates, a health analytics firm in South Bend, Indiana. “And as you think about equity, it really is a core aim of quality.”

“We certainly know that there are persistent inequities in healthcare and health outcomes, and those inequities can be based on race, sex, language, and gender,” she continued, adding that recently, “there’s been much more awareness about structural racism and other systematic discrimination that really has led to these inequities in the healthcare system, and much more broadly in our society.” This has been true even before the pandemic, but “COVID certainly brought this front and center — we’ve seen large inequities in rates of infection, hospitalization, and mortality in the U.S., across the continuum of care.”

However, issues like equity and bias aren’t always easy to discuss, Gandhi said. “The bias issue is one that needs to be addressed both for safety and equity,” she said. “We have to highlight the need for psychological safety, for difficult conversations and avoiding excessive focus on shame and blame, and then have a strong focus on leadership and culture.” Gandhi added that during the last year and a half, she has been leading a large equity effort around the country, “and I’ve been in lots of conversations with health systems who are just beginning to really think about what their strategies around equity ought to be, and I’ve had many moments that have felt like déjà vu,” she said. “It reminds me of conversations I had 20 years ago when we were just starting out in patient safety.”

Data need to be stratified by racial and ethnic groups so that inequities can be uncovered, “and we’ve certainly started to do that with various types of quality and safety measures,” Gandhi said. For example, research has found that maternal mortality rates for non-Hispanic Black women are 2.5 to 3 times higher than for non-Hispanic White women and Hispanic women.

Standardization of a health system’s procedures can work very well from a patient safety standpoint, “but we have to make sure that whatever we’re implementing is not unintentionally widening inequities,” said Gandhi, noting that one study from the Robert Wood Johnson Foundation found that giving everyone the same intervention may actually widen gaps in treatment. “With equity, you have to potentially have a customized solution to really achieve what you’re trying to achieve. There’s a tension between standardization and customization that we need to be really aware of.”

What steps can organizations take to improve equity in patient safety? Recognizing the role of bias in diagnostic errors is one part of the solution, Gandhi said. She and her colleague Hardeep Singh, MD, wrote about the eight types of diagnostic errors that they are seeing during the pandemic, and how bias may contribute to some of them. Examples of error types and the role of bias in each included:

• Classic diagnostic error, such as a missed or delayed diagnosis in a patient with COVID-19 who exhibits classic respiratory symptoms. “There’s certainly the potential for inequities in these kinds of errors because there may be lack of access to quality facilities, bias in who’s getting tested, and, potentially, who is actually getting listened to that could all contribute to these kinds of classic errors,” she said.
• Anomalous error, such as a missed or delayed COVID-19 diagnosis in a patient presenting with non-respiratory symptoms and who is not tested for COVID. Bias in who gets tested or listened to could play a role here as well, said Gandhi.
• Collateral error, which is a delayed diagnosis of a non-COVID condition due to a patient’s delay in coming in for care because of COVID concerns. “Certainly inequity could contribute to that collateral error occurring because of patients perceiving that the healthcare system was not trustworthy for them to come in,” she added.
• Unintended error, which is defined as a missed or delayed diagnosis because of less direct interactions, including the rapid increase of telemedicine. “We see lots of articles about lack of access to telemedicine, language barriers, and differences in digital literacy that can contribute to telemedicine potentially being less accessible or used less well for certain populations, and that could contribute to more of these unintended errors,” Gandhi said.

How can the situation be improved? “Safety event reporting is a really important place to start,” she said. “Most organizations have thousands of these that are coming in, but it’s not routine to have segmentation being done where you look at safety in terms of race or ethnicity patterns within those reports.” One study found that there were racial differences in reported safety events; the demographic characteristics of patients with reports differed significantly from overall hospital demographics. “Understanding what’s getting reported on what types of patients is going to be really important, especially if you’re relying on safety reporting as a primary source of obtaining information about safety-related issues.”

Root cause analysis (RCA) is another thing to consider, Gandhi added. “We have ways to train safety personnel to really add inquiry about equity into the root cause analysis process, embedding the equity lens into questions being asked, embedding it into the types of action-items being designed and implemented, and also monitoring for those unintended consequences … RCA is another kind of low-hanging fruit in terms of where you can start looking for inequities, and thinking about designing to try to reduce those gaps.”