In a new study by Harvard Medical School researchers, humans faced off against computers for the most accurate diagnosis
Orange County, CA - October 14th 2016 - Technological advancement has driven development of every economic sector and changed the quality of life on a global scale. Focusing on the healthcare industry, there has been an average of 9.8% increase of annual spending, 2.5% more than projected estimates. This investment has led to the development of various medical innovative technologies, saving countless lives since 1970’s.
While the surge in healthcare technology is far from precarious, certain technology is now interfering with the.decrease diagnostic errors made by healthcare professionals. provisional care of physicians. Websites like WebMD and HealthLine have been created and designed to offer symptom evaluation. Users will input their symptoms they are experiencing, and the site will list all the indexed diseases, disorders, and/or conditions that could account for the health idiosyncrasies.
While being proactive about one’s health is encouraged, it is making physician’s jobs more challenging. This is briefly touched upon in our post titled “Paranoid Parenting in the 21st Century”. This article discusses mobile applications like My Child and MyToddlr, designed to monitor child development through health questions, outside content, and finding local experts on the topic at hand. While these apps and websites are giving parents piece of mind and allowing connectivity, a new study suggests these sources have less accuracy compared to actual physicians.
In a study led by researchers at Harvard Medical School, it was found physicians’ diagnosies were exceedingly competent compared to commonly used symptom checkers. Physicians were able to accurately diagnose a patients underlying issue more than twice as often as 23 of these prevalent sites.
For the study, 234 internal medicine physicians were each given 45 clinical cases to initially diagnose, and supply two additional possible diagnoses. The cases involved both common and uncommon ailments with differing degrees of severity, but each illness was determined by at least 20 physicians.
Physicians were able to list the correct diagnoses with 72% accuracy, compared to the 34% rate of the cross-examined apps and websites. The biggest discrepancies recorded were when diagnosing severe and uncommon conditions, in which physicians once again outshined their digital counterparts.
While the physicians did trounce their competition, misdiagnoses were made in 15% of cases. Researchers plan to continue studies in the development of computer-based algorithms to decrease diagnostic errors made by healthcare professionals.
Senior investigator and associate professor of health care policy at Harvard Medical School has said, "Clinical diagnosis is currently as much art as it is science, but there is great promise for technology to help augment clinical diagnoses. That is the true value proposition of these tools."