How AI is helping unlock data to find clues for timely treatment
Whether it's transmitted digitally through an HIE or arrives in the form of a fax, healthcare data is more voluminous than ever. A clinician might have to spend considerable time and effort, for instance, poring over the data to know whether a medication a patient was on ten years ago was still necessary at the same dosage now.
Machine automation can help. Medal is among the growing market of companies, including edgeMED, Inovalon, Telegenisys and others, that work to extract, process and contextualize the vast amount of information that exists about patients – and present it to physicians in a timely and relevant way.
Medal founder Lonnie Rae Kurlander says she has a mantra: "Simplicity is the ultimate sophistication." She says that unwieldy mass of patient information can be tamed with artificial intelligence, benefiting physicians and patients alike by reducing workloads and speeding time to relevant information.
Andy McMurry, Medal's co-founder and chief information officer, says machine learning can do the reading and assign context to tell the doctor about the patient's most relevant condition.
"I was serving a patient in the ICU who was in a coma because we couldn't get his patient info in a timely manner," says Kurlander. "If I'm able to quickly find the hints, the clues, maybe the prior diagnosis is already there. Maybe there were similar incidents before that could have helped us."
Kurlander sees Medal's approach as a collaborative tool more than a road to replacing physicians. She invokes Elon Musk's success in meeting car manufacturing goals at Tesla by starting a second production line in the factory parking lot staffed entirely by humans: "He said his biggest mistake was over automating."
Computer augmentation isn't ever supposed to replace a physician she argues – instead, it has the great potential to "put power in hands of power users of health care," she says. She notes that in ancient games like chess or go, even a novice user "playing with (computer) assistance will win every time."
McMurry has a mantra of his own: "A wealth of information creates poverty of attention."
Physicians, he says, spend too much time as "fancy store clerks" who enter and sift through massive amounts of data. The benefit of it all is lost through the sheer volume of it; AI-driven software like Medal can help assign context and meaning to it for practitioners, bringing them back to the "joy of delivering medicine."
Doctors feel swamped with the data side of the new healthcare landscape. Leveraging data to provide deeper insight into patient outcomes has been shown to significantly reduce time spent on routine data entry.
Using machine learning and AI to help "read through the chaff," as McMurry says, takes over for physicians in areas that they shouldn't be and aren't even really supposed to be doing. This leads to a greater range of potential reached by professionals in healthcare, not to say improved outcomes and money saved.
Finally, using AI to simplify how doctors interact with patients will prepare healthcare organizations for the future.
Benjamin Harris is a Maine-based freelance writer and and former new media producer for HIMSS Media.