Helping clinicians make their best decisions
When ill patients walk into their offices, it can be difficult for a physician or nurse practitioner—despite their years of training and experience—to determine what’s wrong with them and prescribe appropriate treatment. Even determining whether the person is experiencing a cold or simply seasonal allergies can be difficult, but as health care races toward the age of precision medicine in which treatment is tailored to a person’s specific genetics, having all of that information memorized will become literally impossible.
That’s where technology comes in.
“We might be asking too much of a physician to hold all of this information in his or her head,” said Cason Schmit, JD, research assistant professor at the Texas A&M School of Public Health. “Instead, we need appropriately disruptive technologies to help them.”
Just as the GPS in our cars can guides us to the best route, another type of technology, called a clinical decision support tool, can help health care professionals make the best diagnosis and treatment decisions.
Programs already exist to take natural human biases into account to help physicians make better decisions. But what if someday a tool existed to integrate this with the world’s millions of electronic health records (EHRs) and hundreds of thousands of medical research studies to provide diagnostic and treatment insights? It could alert the physician to recent changes in the evidence base and even suggest possible options based on thousands of pieces of data.
“A learning health system is an entirely different way of thinking about health care and research, in which each and every patient encounter is a learning experience and is added to the evidence base,” Schmit said. “Certain groups are underrepresented in clinical trials, so we need to learn from real clinical outcomes as well.”
Much of this learning can be done by the computer. “You don’t need a person to compare treatment outcomes,” Schmit said, “just good EHRs and a well-written algorithm.”
Of course, there are privacy and regulatory concerns that will need to be addressed before such a system could be a reality. Still, it will be important going forward to take full advantage of what precision medicine promises.
These technologies are synergistic as well. “There are too many genotypes for providers to memorize and recall on demand for appropriate decision making,” Schmit said. “Precision medicine requires the clinical decision support tools to get the right information, at the right time, in front of the right patient. A learning health system would permit rapid integration of new evidence into clinical decision support for continually improved precision medicine.”
Schmit acknowledges that having a perfect system may still be a long way off. “It’s still very much fairy tale-esque,” he said, “but it’s when you combine EHRs with other health information technologies—e-prescribing, electronic reporting, health information exchanges, and especially these advanced clinical decision support tools—that the true benefits are very apparent.”