Data Helps Medical Patients Find the Right Care

Data Helps Medical Patients Find the Right Care

Whether it’s a severe migraine, a bad sprain or a powerfully sore throat, many people find themselves wondering—Do I need to go to the ER?

Unfortunately, it’s a high-stakes question not only for the health of medical patients, but also for their wallet. The average cost of an ER visit is on the rise, according to one study,1 with an average visit costing patients $1,917 in 2016. To make matters worse, the high cost of these bills are typically not revealed until after the visit, and some insurance companies won’t pay2 for a visit that was deemed “avoidable.”

Data Helps Medical Patients Find the Right Care

This definition of avoidable is widely open to interpretation: researchers report the number of avoidable ER visits could range from 5 percent to as much as 90 percent.3 For those medical patients showing up at the ER hoping to receive the care they need,  they are facing far higher bills than necessary for the care they received.

“We should figure out why the patient is accessing the ER and then find ways to address it,” said David Shih, M.D., executive vice president of strategy and co-founder of New York-area urgent care provider, CityMD.

A growing class of innovative healthcare companies believe medical care shouldn’t be such a daunting mystery at life’s crucial moments—and they believe data is the key to the puzzle.

Read more about AI Healthcare Trends and Breakthroughs

Taking a Closer Look at ER Data

According to a CityMD poll4, 89 percent of Americans chose the wrong level of care when asked about eight different patient health scenarios—including twenty-three percent of respondents who incorrectly believed that a bruised ankle warranted an ER visit.

To combat these misunderstandings, CityMD is working with insurance providers, hospital and primary care systems, and government-sanctioned regional health data organizations to list the most common symptoms or personal circumstances leading to ER visits. For now, CityMD is using this data to make insights available to insurance companies and healthcare providers to reveal the best pathways for medical patients in need of care on an individualized basis.

“If there is a way to communicate with patients right at the beginning and guide them where to go, that’s one of the ways to tackle this problem,” Shih said.

Data Helps Medical Patients Find the Right Care

Upon routine scans of available data, CityMD identified that the groups that are among the most frequent visitors to the ER are asthma and migraine patients. CityMD can then work with the insurance company or a health system to link those patients to a care provider who can better manage the recurring conditions.

The company also reviews final diagnoses and the circumstances of these visits—did the patient show up at a nearby ER only because their doctor was closed after business hours, or they didn’t have a car? Knowing why someone chose a certain care option is the first step in helping him or her choose the right option in the future. Eventually, CityMD hopes to implement its own triage system, directing medical patients who are unsure of where to go for immediate care to the right place. To that end, CityMD is exploring technology that will help direct patients to the right level of care.

Buoy Health Drives Diagnostics of Medical Patients with Data

One such technology that has these capabilities is Buoy Health, founded in 2014 by a Harvard Medical School graduate and other Harvard colleagues, Buoy has created a symptom-interview chatbot that leads consumers through a series of questions toward a diagnosis. The AI-powered algorithm processes the customer data and cross-references it with data from thousands of medical research papers over decades that link symptoms to diagnoses and outcomes, according to CEO and co-founder Andrew Le, M.D.

Once Buoy’s algorithm processes all this information, it serves up a recommendation to patients regarding the level of care they need, whether it be a normal office appointment, urgent care or emergency care. Buoy also “gets smarter” with each use, says Le. The algorithm learns from the connection between reported symptoms and eventual diagnoses and outcomes, as a portion of users are asked in follow-ups what happened with their cases.

Together with robust data from health system partners, the outlook for a more effective triage system looks bright. Based on a study of 500 medical patients, Buoy researchers found the algorithm produced the same diagnosis as a doctor 90 percent5 of the time.

“Our real vision is to become the front door of health care,” Le said. “Search engines were not designed for health care. We are. If we could replace them as that first step, we could really do some good in society.”

how is artificial intelligence solving medical data problem

This content is produced by WIRED Brand Lab in collaboration with Western Digital Corporation. 


1. Health Cost Institute, 

2. Vox

3. Health IT Analytics

4. CityMD

5. TechCrunch