Need to Address Diagnostic Errors is Urgent, IOM Says

Register now

Diagnostic errors in healthcare are at epidemic proportions and causing serious safety concerns, yet the problem has received limited attention from the medical community, according to a core finding in an Institute of Medicine report on diagnostic errors released on Tuesday.

“Urgent change is warranted to address this challenge,” states the IOM report. “For decades, diagnostic errors—inaccurate or delayed diagno­ses—have represented a blind spot in the delivery of quality healthcare. Diagnostic errors persist throughout all settings of care and continue to harm an unacceptable number of patients.”

Errors in diagnosis affect one in 20 patients annually, an estimated 12 million Americans, according the report, which concludes that “most people will experience at least one diagnostic error in their lifetime, sometimes with devastating conse­quences.”

However, Paul Epner, executive vice president of the Society to Improve Diagnosis in Medicine (SIDM), argues that this figure is based on estimates from ambulatory care settings alone, not including emergency rooms and hospitals, in which case the number would be significantly higher.

In addition, Epner says it is difficult to accurately quantify the number of diagnostic errors that occur annually because very little research has been conducted in this area and there is no common format for reporting these errors. “So much of it goes unknown and unreported,” adds Ross Koppel, a professor in the Department of Sociology at the University of Pennsylvania and principal investigator at the Perleman School of Medicine. “It’s grossly underestimated.”

SIDM petitioned IOM to conduct the study, which Epner hopes will be a catalyst to improve the safety of diagnoses across the healthcare continuum.

“This is a cross-cutting problem. No one organization or person can solve the problem of diagnostic errors because no one organization or person owns the responsibility for the problem,” observes Epner. “There are things we need to do collaboratively as well as within each respective organization.”

Calling it a “moral, professional, and public health imperative,” the IOM report states that “improving diagnoses will require collaboration and a widespread commitment to change among healthcare professionals, healthcare organizations, patients and their families, researchers, and policy makers.”

Towards that end, the IOM report provides recommendations to improve diagnoses and reduce errors, including several related specifically to health IT.

Also See: Inherent Risks of Health IT, EHRs Putting Patients in Peril

Under the goal of “ensuring that health information technologies support patients and healthcare professionals in the diagnostic process,” the report recommends:

*HIT vendors and the Office of the National Coordinator should work with users to ensure that health IT used in the diagnostic process demonstrates usability, incorporates human factors knowledge, integrates measurement capability, fits well within clinical workflow, provides clinical decision support, and facilitates the timely flow of information among patients and health professionals involved in the diagnostic process;

*ONC should require health IT vendors to meet standards for interoperability among different health IT systems to support the effective, efficient and structured flow of patient information across care settings to facilitate the diagnostic process by 2018, and;

*The Department of Health and Human Service should require health IT vendors to: routinely submit their products for independent evaluation and notify users about potential adverse effects on the diagnostic process related to the use of their products; and permit and support the free exchange of information about real-time user experiences with health IT design and implementation that adversely affect the diagnostic process.

EHRs Part of Problem

Undoubtedly, electronic health records are contributing to diagnostic errors, comments Epner. Dean Sittig, professor at the University of Texas School of Biomedical Informatics and a specialist on the safety of EHRs, agrees.

“Health information technology has a role both in causing the problems and hopefully in solving the problems,” Sittig claims. “We’ve come a long way with health IT but there’s still a considerable amount of work to be done. For example, the usability and interoperability of these systems are two huge areas for improvement.”

“How many more hundreds of thousands of people do we have to kill from lack of interoperability?” asks Koppel. “IOM’s recommendations focus on having ONC do what needs to be done rather than rolling over and playing dead. They have to be the agency that ensures that the vendors do the right thing rather than just ensure that the vendors make more sales.”

But, Andrew Gettinger, M.D., ONC’s chief medical information officer and acting director of the Office of Clinical Quality and Safety, takes exception and argues that the agency is already playing a major role in this area. “Those recommendations from IOM are exactly what we’re working on,” says Gettinger. “We released earlier this year a draft Interoperability Roadmap. It has specific milestones that we hope to achieve,” adding that “our push to interoperability, when it’s realized, will help decrease diagnostic errors.”

While Gettinger believes that health IT improves healthcare and patient safety, he acknowledges that the technology has created some problems—including adverse outcomes due to poor implementation or improper use—but they are “small in magnitude relative to the benefit.”

One of IOM’s recommendations is to “facilitate more effective teamwork in the diagnostic process among healthcare professionals, patients and their families,” and that “providers should ensure patient access to electronic health records, including clinical notes and diagnostic testing results, to facilitate patient engagement in the diagnostic process and patient review of health records for accuracy.”

Sittig sees this emphasis on teamwork as critical to proper diagnosis. “Right now, our electronic health records are not particularly good at facilitating teamwork,” he says. Sittig and Hardeep Singh, a patient safety researcher at the Michael E. DeBakey VA Medical Center and Baylor College of Medicine in Houston, point to the misdiagnosis last year of Thomas Duncan, America’s first Ebola patient, who was initially treated and released from Texas Health Presbyterian Hospital’s emergency department in Dallas—only to return to the hospital’s ED where he was admitted for treatment of Ebola and later died.

Despite the fact that information concerning Duncan’s travel from Africa was gathered by a nurse as part of the influenza screening process and documented in the patient record, it was not verbally communicated to an attending physician. Sittig and Singh say the nurse should have informed the physician of the information instead of relying on the physician to find the information in the EHR.

“In the Ebola case in Texas, the nurse recorded it in some nursing documentation but that’s not routinely reviewed by physicians,” says Sittig. “So, they’re both doing their work using the same computer application but they are operating separately.”

Sitting and Singh co-wrote a paper published last year in the journal Diagnosis regarding the Ebola case, in which they conclude: “EHR-based clinical workflows often fail to optimize information sharing amongst various team members, leading to lapses in recognizing specific clinical findings that could aid in rapid and accurate diagnosis.”

EHRs Part of Solution

Sittig and Singh are working on EHR-based “electronic triggers” or algorithms that look for common patterns of events that might be indicative of potential diagnostic errors, including unusual patterns of return visits after an initial care visit or lack of follow-up of abnormal clinical findings.

Seventy-two primary care providers participated in a 15-month study of such an algorithm for identifying patients at risk for delays in diagnostic evaluation of colon and prostate cancer. “Electronic trigger-based interventions seem to be effective in reducing time to diagnostic evaluation of colorectal and prostate cancer as well as improving the proportion of patients who receive follow-up,” found the study published recently in the Journal of Clinical Oncology. “Similar interventions could improve timeliness of diagnosis of other serious conditions.”

“One of the goals of an IOM report is to raise awareness of a problem and then to try to spur some action,” says Sittig. “One of the actions we’d like to see is a lot more research in this area.”

For reprint and licensing requests for this article, click here.