By Aaron Krol
November 11, 2015 | The New England Journal of Medicine has published an opinion piece by Drs. Hardeep Singh and Mark Graber, calling for healthcare reforms to reduce the rate of diagnostic errors. These errors, Singh and Graber write, “may be one of the most common and harmful of patient-safety problems,” affecting an estimated one in ten diagnoses and often delaying effective care, or even leading to unnecessary interventions.
To combat diagnostic errors, the authors recommend new health information technology (HIT) approaches, cultural and organizational changes in the way patients are diagnosed, and increased funding for research into the issue.
The occasion of the opinion piece is the publication, this September, of the Institute of Medicine’s (IOM) report, Improving Diagnosis in Health Care. This report follows on 1999’s To Err Is Human: Building a Safer Health System, which first brought to wide public attention the impact of common medical errors on American healthcare. Today, To Err Is Human is remembered for its troubling conclusion that medical error could be responsible for as many as 4% of all deaths in the U.S., and that errors in treatment are a systemic problem of the health system, not isolated problems of individual physicians.
Improving Diagnosis in Health Care expands the scope of the earlier report to cover diagnoses, and includes the stark assertion that a typical American can expect to be misdiagnosed at least once in his or her lifetime. “We are optimistic that the report will spark a renaissance of interest in improving diagnosis and reducing patient harm from diagnostic error,” Singh and Graber write in their perspective.
Graber is a member of the IOM committee that compiled the report, and also serves on the board of the Society to Improve Diagnosis in Medicine. Singh, who works at Baylor College of Medicine and the Veteran Affairs Center in Houston, is not a member of the committee, but did offer testimony to the IOM.
The rapid rise of HIT, and especially of electronic health records, might appear to offer new opportunities to reduce diagnostic error. HIT tools can remind physicians to give appropriate tests to high-risk patients, or automatically flag signs and symptoms in the health record that conflict with a diagnosis. However, Singh and Graber caution against emphasizing technological solutions over cultural ones, given physician dissatisfaction with electronic records and the risk that HIT tools will not fit comfortably in the workflows of the outpatient visits where most diagnoses are made.
“One of the promises of transitioning our health records from paper to electronic systems is that we can do a better job of understanding how care is delivered,” says Dr. Robert El-Kareh, a professor at the University of California, San Diego who studies the issue of diagnostic error and HIT. El-Kareh was not involved in the IOM report or the NEJM paper, but he supports Singh and Graber’s conclusions, including that HIT can only aid in diagnosis to the extent that it accommodates clinicians’ working processes.
“It has become clear that engaging cognitive scientists and human factors engineers in the design of HIT is very important,” El-Kareh says. “We’ve transitioned from practicing with limited data available to practicing with huge amounts of data to sift through… By designing better computer interfaces and ways to display clinical data that align with human cognition, we can tame the information overload.”
He adds that there are two roles for HIT in reducing diagnostic error that are best-supported by evidence: ensuring that patients receive the right screening tests, and reminding physicians to follow up appropriately on test results. “Both of these types of systems will help avoid delays in diagnosis,” he says. “HIT holds promise in helping us do real-time surveillance of likely diagnostic errors. Once we’ve identified high-risk scenarios, it’s up to the people in the system to learn from them and develop ways to avoid them.”
A System-Wide Response
For Singh and Graber, the highest priority in addressing diagnostic error has to be cultural reform. When healthcare providers in the early 2000s took on other forms of medical error, they often had to reduce the independence of physicians, adding checklists to reinforce rote preventive measures in treatment and empowering nurses and other care workers to correct errors before they happen.
Similarly, Singh and Graber write, “The challenge here is cultural, because diagnosis has traditionally been the physician’s responsibility; physicians may not seek help, despite uncertainty or diagnostic difficulties.”
The authors advocate for changes to medical education that foster better communication between members of a care team when making diagnoses. More immediately, they encourage physicians to consult with one another on more challenging diagnoses, and propose new roles for nurses in tracking clinical measures and identifying those with diagnostic significance.
El-Kareh agrees that gaps in communication between health workers are important contributors to diagnostic error. “There are well-established ways to send messages between team members, but their impact will be limited if patients, nurses and others don’t want to be perceived as challenging a physician’s assessment or decision,” he says. “In-person communication is still crucial. Technology can help fill in gaps in communication when necessary, but face-to-face communication is best, when feasible.”
Singh and Graber hope that the IOM report will provide momentum for a system-wide effort to address diagnostic error, with the understanding that not all forms of error are well-studied or easily tracked. In their NEJM perspective, they propose that the federal government build on the IOM’s findings, by uniting multiple agencies to fund research and pilot projects that will test different methods of ensuring reliable and timely diagnoses.
“Now could be an opportune moment to create a public-private partnership to propel progress,” they write, suggesting that the VA, Centers for Disease Control and Prevention, National Institutes of Health, Agency for Healthcare Research and Quality, and the Office of the National Coordinator for Health Information Technology all pool efforts with each other and with nonprofit and professional groups. Many such organizations are already united under the Coalition to Improve Diagnosis.
Calling attention to the prevalence of diagnostic errors is an important first step: because diagnosis is inherently uncertain, and errors usually appear isolated, it can be difficult to see misdiagnoses as a systemic issue. But our experience with other forms of medical error show that recognizing the extent of the problem is only one step in correcting practices that are dangerous to patients. Fifteen years after the publication of To Err Is Human, many forms of preventable medical harm, like hospital-acquired infections and medication errors, still occur on a daily basis. Diagnostic errors, which can sometimes take months or years to come to light, could prove even more difficult to overcome.
“Diagnostic errors are too common and can cause a lot of patient harm while increasing the overall cost of health care,” says El-Kareh. “More investment in research to figure out how to reduce these errors is crucial.”