Industry vet signs on with new company with a fresh view

After 50 years in healthcare, spanning biomedical engineering, health policy, IBM’s Watson supercomputer effort and more, Martin Kohn, MD, has seen a lot. But, he decided to join Sentrian, the first remote patient intelligence company, serving as chief medical scientist.

Kohn told Clinical Innovation + Technology he was intrigued by Sentrian’s deviation from the traditional simplifying assumptions used in literature that created limited value. Most models for making decisions were based on population studies in journals but those studies weren’t very helpful in deciding how to manage patients with chronic conditions, he said.

Sentrian, however, took on the specific issue of helping reduce avoidable hospitalizations and ED visits using data collected from home monitoring systems and integrating that with longitudinal health information. “They’re going about it in the correct way in my mind.”

There have been many efforts to use home monitoring, he said, but researchers assumed for the purposes of monitoring that the patient had a single disease. In fact, about 40 percent of patients with chronic obstructive pulmonary disease also have congestive heart failure, for example. Researchers assumed they could monitor one parameter like body weight to track disease status. They further assumed that a single threshold change would be an accurate indication of change. While that did reduce some avoidable hospital readmissions, it led to many false positives, he said.

Sentrian makes none of those simplifying assumptions, according to Kohn. To effectively intervene, clinicians need to identify a patient’s deterioration with enough time to allow for an intervention. To do so, “we need to take a much broader approach, looking at a variety of physiological metrics.” The company aims to find patterns that predict a patient will become severely symptomatic up to eight days in advance to allow time to change the patient’s course.

This is exactly what needs to be done to improve outcomes while reducing costs, Kohn said. “One or the other is not nearly enough. That’s why I joined Sentrian. It’s the right combination of technology and clinical insight.”

Sentrian combines and overlaps tools to see comorbidities. The system employs machine learning feedback to improve itself so as they work with a set of patients, the system changes its own rules to improve performance. “The longer we work with a group of patients, the better the system becomes.”

The company will add more patients and expects to have several thousand patients on the system in the next few months. “The opportunities for learning and improving will be even greater. Our initial experience gives us great confidence that the concept is valid.”

The return on investment threshold is very low, said Kohn. “If you look at a group of patients with chronic disease who have three acute care encounters each year, you don’t have to make much of a dent.” Sentrian charges based on monitored members per month at about $100 each. The cost of devices and the analytic platform is about $1200 per patient per year. Compared with a typical hospitalization cost of about $10,000—“you don’t have to make much of a dent in those hospitalizations to have a positive ROI.”