Jonathan H. Chen was an intern at Stanford Hospital a few years back, admitting patients with unusual medical syndromes or rare diseases.

He wasn’t always sure how to immediately treat these patients.

“I found myself clueless at times,” said Chen. “I thought to myself, I should review the chart of a similar patient who had an experienced clinician care for him so that I can learn from their care plan.”

That triggered Chen’s eureka moment. 

“Why look at just one person’s chart?” he thought. “Why not look at the last thousand charts to see how all doctors take care of their patients in similar cases?”

Doing so, he would have the potential to crowd-source the collective wisdom of physicians all in one central location.

Already having a PhD in computer science and spending a few years as a software developer before medical school, Chen — a wunderkind who started college when he was 13 — knew he had the rare set of skills to marry medicine and technology.

“I thought about how the Amazon product-recommender algorithm works and thought, `Can we do this for medical decision-making?’” said the 34-year-old Chen, a VA Medical Informatics Fellow at Stanford Health Policy.

So instead of, other people who bought this book also liked this book, how about: Other doctors who ordered this CT scan also ordered this medication.

“What if there was that kind of algorithm available to me at the point of care?” he asked. “It doesn’t tell me the right or wrong answer, but I bet this would be really informative and help me make better decisions for my patients.”

The National Institutes of Health agrees. Chen was recently awarded a five-year NIH grant as the principal investigator behind OrderRex, a digital platform that data-mines electronic medical records to learn clinical practice patterns and outcomes to inform concrete medical decisions.

Chen is designing and coding OrderRex with the help of his chief mentor, Russ Altman, a professor of bioengineering, genetics and medicine and director of Stanford’s Biomedical Informatics Training Program. Stanford Health Policy professors of medicine, Mary Goldstein and Steven Asch, round out his core team of grant mentors. Grant collaborators Nigam Shah, Lester Mackey, and Mike Baiocchi are providing additional critical expertise.

“I think OrderRex is a first step towards an entirely new way to provide decision support to physicians,” said Altman. “We will not only have a large database of patients from which we can collect similar patients to create virtual cohorts, but we will also have a database of the decisions that their physicians have made in different clinical situations.”

Altman added: “Each of these capabilities would be transformative — but together they would really change what is possible for a provider sitting with a patient, making decisions about diagnosis and therapy.”

The NIH’s Big Data-to-Knowledge grant will allow Chen to develop and test the platform. Stanford Medicine’s Center for Clinical Informatics provided Chen a year’s worth of Stanford Hospital records, including every medical order for every patient. The more medical data he loads, the more patterns begin to form.

Chen has been using a derivative of Amazon’s algorithm to make his platform scalable with millions of patient records. The broad vision is to eventually integrate this tool with hospital computer networks to assist physicians with their decisions.

“Imagine, technology allowing medical decisions to be informed by the collective experience of thousands of other physicians right at the point-of-care,” Chen said.

There are naysayers who worry such a product will further alienate physicians from their patients and allow doctors to jump to crowd-sourced conclusions about treatment. Chen emphasizes OrderRex would only serve as a tool, which does not substitute for human contact, calculations and conclusions.

“Tools like this are simply to augment the medical decision-making process and hopefully — and I know this is a big goal — improve the quality and efficiencies of health care.”

Altman says the lacking-human-touch argument is imprecise and potentially unethical.

“Of course, providers will always be real people and of course they should be empathetic, listen to the patient, examine the patient, and think about what’s best in the big picture,” he said. “But if there are technological tools that they can use to improve their decision-making, it is probably unethical to replace data-driven decision-making with `touch’ and ‘intuition’ — which often perpetuates the status quo and contributes to variability in practice and variability in outcomes.”

Stanford Medicine is already leading the revolution in precision health and big data to overcome human error and misdiagnosis.

In a 2014 Health Affairs article, Stanford pediatrician Christopher A. Longhurst along with Nigam Shah, MBBS, PhD, assistant professor of biomedical research and assistant director of the Stanford Center for Biomedical Informatics Research, and Robert Harrington, MD, professor and chair of medicine, outlined a vision for drawing medical guidance from day-to-day medical practice in hospitals and doctors’ offices. They called it the Green Button.

The idea is to give doctors access —a green button — to patient data from a vast collection of electronic medical records. They wrote that the instant access to EMRs isn’t a substitute for a clinical trials, but better than resorting to the physician’s own bias-prone memory of one or two previous encounters with similar patients.

Chen is working with those professors, but notes the Green Button concept is to look for “patients like mine” and ask questions about different treatment options that may yield different results. His approach looks for “doctors like me,” and anticipates what the doctor wants before they ask for it.

“The conceit of my approach is that all practicing doctors are already trying to make our best-guess decision to improve our patients' outcomes,” he said. “Rather than trying to directly predict how to change a patient outcome, I look to the records of physician decision-making that already represent a wealth of expertise we are not leveraging in a systematic way.”

Could that wealth of expertise one day make Chen a wealthy man, perhaps the Jeff Bezos of the medical informatics world?

“I wouldn’t complain if I was,” Chen said with a grin. “But if I just wanted to make money, I wouldn’t have gone to medical school,” He gave up a lucrative living as a software developer.

He does recognize, however, that for OrderRex to have a big impact, commercial applications such as licensing the product as an add-on to EMR systems are likely.

“So, having a broad impact that will serve the mission of improving quality and efficiency — that is the ultimate goal.”