Houston Methodist has partnered with an artificial intelligence company to provide physicians and staff with predictive tools to better identify at-risk patients and improve long-term outcomes.

The partnership with Health Data Analytics Institute (HDAI) leverages the company's large database of Medicare data to generate personalized information about Houston Methodist patients, including the likelihood of developing medical conditions, requiring readmittance or needing hospice care.

"Creating a holistic profile for each patient that is current, timely and available to every care team member system-wide allows us to improve transitions, care coordination and handoffs, and deliver a better overall patient experience," Roberta Schwartz, executive vice president of Houston Methodist Hospital, said in a HDAI news release.

The partnership should reduce rates of mortality, readmissions and lengths of stay, said Dr. Alan Lumsden, chairman of cardiovascular surgery at Houston Methodist, one of the two hospital departments — orthopedics is the other — that has used the HDAI program for the past two years.

Houston Methodist Hospital has begun rolling the HDAI program out to different departments, confident it will improve on old health care risk tools that too often are "fragmented, insular and unreliable," according to Schwartz. The launch represents a further unleashing of the long-awaited data-driven transformation of health care.

The program, known as Health Vision, provides a granular analysis of health care outcomes and costs as well as customizable lists of patients who would benefit from specific interventions.

Reducing big data into actionable insights

Dr. Lumsden said Health Vision quickly identified one important problem: the breakdowns that occur after the patient leaves the hospital.

Houston Methodist's mortality rate ranked No. 2 in the nation during the patient's stay in the hospital, Health Vision found, but the hospital dropped to No. 32 in the 14-day period after the patient was discharged.

Turns out, there could be as much as a tenfold mortality-rate difference in the skilled nursing facilities where patients sometimes went following discharge.

"That was an incredibly valuable piece of information," said Dr. Lumsden. "We spent a lot of time and effort focusing on the wrong thing: The problem isn't hospital processes. It's the after-discharge processes that we really need to focus on."

As a result, said Dr. Lumsden, Houston Methodist "virtually reengineered" the clinical care pathway — someone from the office makes sure to check in with high-risk patients two days after discharge to confirm they're taking their medications, eating and functioning properly and will be coming in for follow-up appointments.

"The new data alerted us to the need to reallocate our resources and change the skilled nursing facilities patients go to," said Dr. Lumsden.

Harnessing the power of AI

The partnership produces personalized risk profiles by matching real-time data from Houston Methodist patients with that of similar patients drawn from HDAI's archive of electronic health records of more than 100 million patients over 20 years. The result: predictable logarithms for each patient's anticipated mortality, length of stay, risk of developing a chronic condition or adverse event and readmission possibility.

Dr. Lumsden and Schwartz stressed the unique ability of AI to reduce massive amounts of cumbersome and sprawling data into actionable insights.

Houston Methodist's focus is the highest risk quintile, the top 20% of at-risk patients, because they generate most of the risk of readmissions and mortality. Dr. Lumsden said that allows the hospital to "focus our resources on where they'll have the most impact."

He also noted that Health Vision unburdens nurses, who spend more than half their time on computers. The idea is to maximize bedside time for bedside nurses, and minimize the amount of data they need to manually enter into computers.

Though it's hard to say they're directly related to the partnership, Dr. Lumsden noted that heart surgery outcomes have significantly improved since the department started using Health Vision. He said the department is collecting such data for future publication.

Dr. Lumsden acknowledged that AI's greatest challenge in health care is acceptance — overcoming passive resistance on the part of doctors and changing the direction of an industry not known for moving quickly.

"AI is not going to replace doctors," said Dr. Lumsden. "Doctors who use AI are going to replace doctors who don't use AI. That's kind of a core tenet, that doctors really need to engage and embrace this. It's moving forward, with or without them."