When Ari Robicsek, M.D., was a young physician, a more experienced colleague passed along an important tip: Keep personal notes on every patient you treat. One day, those notes will help inform how you treat future patients.

The lesson fits with the enduring belief among the medical community that the best way a physician can learn is from personal experience. This paradigm has been the basis for medical training for decades, but it’s a paradigm some are beginning to challenge.

Take, for example, a patient who presents with fever and cough. Physicians rarely know upon physical exam if the symptoms are being caused by bacteria, which can be treated with antibiotics, or a virus, which cannot. As a result, physicians vary widely on how they treat these symptoms, Robicsek, vice president of clinical analytics at NorthShore University HealthSystem in Evanston, Ill., said at an event hosted by Chicago-based health care start-up incubator Matter earlier this month.

Robicsek and colleagues studied prescribing habits among primary care physicians treating patients who presented with fever and a cough at NorthShore from 2006 to 2011. Their findings supported wide variation in treatment of these patients, with the rate prescribing among physicians varying from 18 to 83 percent.

However, the researchers realized that when the physicians were informed of pandemic flu in the community, prescribing among all physicians went down significantly. This got Robicsek thinking: What if physicians could be alerted of “what’s going around” in real-time? Could this encourage more appropriate prescribing habits?

Digital health records and other technological advancements now allow data on outbreaks to be analyzed in real-time, allowing physicians to learn not only from their own experiences, but those of thousands of other physicians and millions of patients.

Robicsek led efforts at NorthShore to mine patient data, build algorithms and integrate information on “what’s going around” directly into the health system’s EHR. His team began by focusing on four diseases, based on the likelihood real-time data could impact physician treatment decisions: influenza, Group A Streptococcus, whooping cough and asthma. The tool launched in January. Robicsek hopes it reduces the physicians’ reliance on “empiric decision making,” the term used by the medical community to describe what is essentially “guesswork,” he said.

In addition to the EHR integration, NorthShore plans to make the “What’s Going Around” tool available to the public through a mobile app currently in development.

Compiling and analyzing data from within individual EHRs allows physicians to see “beyond the keyhole of [their] own practice,” Robicsek said, which encourages more appropriate prescribing and, ultimately, better outcomes for patients.