Last week I wrote something about how the Maryland Health Care Commission (MHCC) had awarded telehealth grants to three organizations to assess the impact of remote patient monitoring on reducing hospital encounters, improving patient care and decreasing healthcare costs.
As that article mentioned, three years ago the State of Maryland launched a bold experiment to improve care. With a waiver from the Centers for Medicare & Medicaid Services, the state converted its hospital payment system from traditional fee-for-service to a global system, in which hospital total revenue for all payers is set at the beginning of the year. The MHCC recognizes that if the state is to succeed with the All-Payer Model, providers need to consider adopting non-traditional approaches to care delivery.
Coincidentally, last week I also saw a great presentation by Jonathan Weiner, DrPH, director of the Johns Hopkins University Center for Population Health Information Technology (CPHIT) in Baltimore. He spoke about the center’s first five years and why he is so excited about the future of population health, particularly in Maryland.
Weiner said it is an exciting time for population health because EHRs provide a new data source. They have been around for 40 years, but only in the last few years has their use become widespread. Although 95 percent of hospitals and doctors are using health IT, only 5 to 8 percent of EHRs are used for population health so far, he estimated, so the potential is huge.
Among the highlights of Weiner’s talk was his perspective on the role population health IT can play in Maryland’s experiment in changing how provider organizations are paid.
“We are excited that the State of Maryland approached us, working with CRISP [the state health information exchange] and other colleagues to develop a framework for understanding population health,” Weiner said, “not because it is a pretty academic slide from the Bloomberg School of Public Health, but because it may be a new framework for how the hospitals and doctors of Maryland will be paid and held accountable. The folks at CMS said if you want to continue your waiver, and get the extra billion dollars you are getting now, you must include population health metrics.”
Over time, the hospitals will be held accountable for the health of populations if they want to stay in existence, Weiner explained. “We have for the first time in history the attention of all the hospital directors when it comes to population health. It is very exciting. The stars are in alignment to make Maryland something quite unique. One of the reasons that CRISP, our HIE, is as successful as it is, with 100 percent of all hospitals collaborating, is because the hospitals have no choice if they want to get paid.”
The proposed framework includes health system factors, social determinant factors, and various outcomes not at the patient level, but at the level of the community, which Weiner stressed was key. “We have a wonderful hospital across the street, but this is not a healthy community,” he said. “In the future, a proportion of every hospital’s revenue — and one day it could be a very significant portion — will be held accountable for what is happening in this community, and that will turn things on their head.”
For example, as competitors, all the Baltimore-area hospitals, Johns Hopkins, MedStar and Mercy, are beginning collaborate to make sure people in their catchment area are getting care without regard to where they have been hospitalized. This is a way to bring together data, he said, to look for diabetes visits to the emergency room. “That should not happen,” Weiner stressed. “Asthma visits to the emergency room. that should not happen.
Johns Hopkins will start looking not just at the people who walk into the emergency room, but at the people in particular Zip codes and census tracks. To do that, you need data across all the EHRs, from public health, and all the insurance claims for all people insured. In Maryland, they have most of that. “Not everything we would like to have in a population health database is fully available,” Weiner said, “but that is our responsibility over time to develop. If I sound excited, it is because I am,” he said. “I have never seen anything as close as where we are today. This is a 15- to 20-year journey, but we have never been this close before.”
Before getting into how CPHIT is working with the State of Maryland to begin to change how hospitals deal with the communities around them, Weiner introduced Hadi Kharrazi, M.D., Ph.D., research director of the CPHIT, who gave several examples of the type of partnerships the center is forming. In one case, Johns Hopkins researchers have teamed up with the Veterans Health Administration’s (VHA) on perhaps the largest study of body mass index (BMI) to date. They can see the BMI of 30,000 veterans in a single day. Using the VHA data warehouse, they have developed a metadata repository for obesity and analyzed the effect of different population denominators on existing obesity rates among VHA members. They also have created a series of predictive models to forecast the obesity trajectory in a given population denominator, which could be eventually tailored as feedback to clinicians, policy makers and others. “These are all objective, EHR-driven measures, not surveys,” Kharrazi stressed.
Kharrazi talked about the VHA’s GIS-clustered Population-based Risk Prediction Modeling (GIS-PRPM) for obesity. It can generate a series of risk prediction models to forecast obesity trajectory for a given patient and/or a larger population over a defined period of time. The outputs can be used in various levels of care from a clinical point-of-care to a larger strategic planning for the VHA’s population health activities. “They can predict the BMI of certain neighborhoods,” Kharrazi added. It has interactive tools so that higher-ups at the VA can drill down and bring in other data sources to see what is contributing to it.”