Addressing the Misdiagnosis Epidemic with Human and Artificial Intelligence
Several thought leaders and stakeholders have joined forces to create GoodDx.org, a searchable database that has the potential to reduce the human suffering affecting millions of Americans.
By John Halamka, M.D., President, Mayo Clinic Platform, and Paul Cerrato, senior research analyst and communications specialist, Mayo Clinic Platform.
Our new book, Redefining the Boundaries of Medicine, discusses the troubling statistics plaguing the U.S. health care system. A report from the National Academy of Medicine points out that about 5% of adult outpatients in the United States experience a diagnostic error annually. The same report found that diagnostic mishaps contribute to about one in 10 patient deaths, cause as much as 17% of adverse effects seen in hospitalized patients, and affect approximately 12 million adult outpatients a year, which translates into one in 20 Americans. About 71,400 of the 850,000 patients who die in U.S. hospitals every year had a significant health condition that went undetected. Several organizations have been diligently working to solve the problem, including a group called GoodDx.org.
While there are numerous reasons for the misdiagnosis problem, one of the most important is that there is no formal system to routinely provide clinicians with positive and negative feedback about their diagnostic decisions. In almost every field of human performance, including the arts, sports, and weather forecasting, professionals review their own prior performance to improve future performance. But in medicine, using feedback to improve diagnoses mostly remains the exception to the rule.
At Mayo Clinic, there is a process to get better. Physicians often conduct peer reviews, morbidity and mortality conferences and tumor boards, and participate in Specialty councils, which build consensus for practice change or setting standards for care.
GoodDx hopes to make big changes. GoodDx is an online library of diagnostic performance resources that can provide feedback to thought leaders, vendors, providers, public safety officers, and patient advocates trying to address the issue by offering a free searchable database. For example, suppose you were a health care executive or physician leader attempting to improve diagnostic performance in GI disease. In that case, you might search using the term gastroenterology and discover a report on the development and validation of algorithms designed to identify delays in diagnosing GI cancer. Similarly, the database provides resources to help identify pediatric patients whose high-risk diagnoses are often overlooked in the Emergency Department. The search engine enables users to find resources for a long list of medical specialties and search using various data sources, including billing, claims, EHR, malpractice, patient-reported data, and registries.
GoodDX.org is funded by the Gordon and Betty Moore Foundation and has enlisted the help of several prominent organizations, including the University of California, San Francisco, Mayo Clinic Platform, AHRQ, CRICO, HCA Healthcare, Intermountain Healthcare, Premier, The Doctor’s Company, UC San Diego, Veterans Administration, Vizient, Johns Hopkins, Kaiser Permanente, University of Michigan, and University of Minnesota.
We discussed possible solutions to the misdiagnosis dilemma in Reinventing Clinical Decision Support, in which we recognized that completely eliminating diagnostic errors is virtually impossible. Nonetheless, there are numerous solutions that can lessen this burden on patients and the health care ecosystem. Among the guidelines listed in The Institute of Medicine Report, Improving Diagnosis in Health Care, the top recommendation is: “Facilitate More Effective Teamwork in the Diagnostic Process Among Health Care Professionals, Patients, and Their Families.” Although there are undoubtedly technological tools that can improve teamwork, at its core is better human communication, and improving the communication process requires that team members overcome numerous psychosocial impediments.
Combining better communication skills with the technological tools outlined in the GoodDx database is the perfect marriage of human and artificial intelligence.
Recent Posts
By John Halamka and Paul Cerrato — Regulators who put too many restrictions on the technology can slow down innovation, but if they set up guardrails that aren’t strict enough, they needlessly endanger patients’ lives.
By John Halamka and Paul Cerrato — The latest research suggests that these digital tools will eventually become useful medical assistants, but we still need to watch for unpredictable fabrications and misstatements.
By Paul Cerrato and John Halamka — In our efforts to encourage prevention and wellness, we often advise patients to exercise, eat better, and manage their stress. Let’s not overlook the importance of a strong social network.