Mount Sinai to Apply AI to Improve Kidney Disease Detection, Management | Healthcare Informatics Magazine | Health IT | Information Technology Skip to content Skip to navigation

Mount Sinai to Apply AI to Improve Kidney Disease Detection, Management

June 1, 2018
by David Raths
| Reprints
RenalytixAI’s machine learning will be deployed against Mount Sinai’s massive patient data warehouse biorepository

Mount Sinai Health System in New York is collaborating with startup RenalytixAI Plc, to use artificial intelligence solutions to improve kidney disease detection, management and treatment for patients with diabetes and other at-risk patient populations.  

The partnership will leverage Mount Sinai’s data warehouse containing over 3 million patient health records and 43,000 patient records in the biobank repository, and using de-identified clinical data, will create an advanced learning system to monitor and flag patients at risk for kidney disease and costly unplanned “crashes” into dialysis. 

Approximately 1 million patients cared for in the Mount Sinai Health System are either diagnosed with Type II diabetes or are of African ancestry, two of the major at-risk population segments for kidney disease.

The first product launch is anticipated in the second quarter of 2019 targeting preventable dialysis and chronic kidney disease costs. Additional U.S. based healthcare systems are expected to participate in clinical utility data development and product launch.  

“Our ability to apply the power of artificial intelligence against such a deep repository of clinical data in combination with prognostic biomarkers has the potential to change the game for all of our patients with diabetes and other populations at risk for kidney disease,” said Barbara Murphy, M.D., Dean for Clinical Integration and Population Health Management and Chair for the Department of Internal Medicine at the Icahn School of Medicine at Mount Sinai, in a prepared statement. She also chairs the RenalytixAI Scientific Advisory Board. 

Established in 2018, RenalytixAI expects to pursue expanded clinical utility trials through collaborations with leading academic medical centers, pharmaceutical and patient advocacy organizations in the United States and Europe this year, followed by submission of product applications for review by the U.S. Food and Drug Administration. 


2018 Seattle Health IT Summit

Renowned leaders in U.S. and North American healthcare gather throughout the year to present important information and share insights at the Healthcare Informatics Health IT Summits.

October 22 - 23, 2018 | Seattle


Partnership to Fine-Tune Care Model for High-Needs Patients

August 21, 2018
by David Raths, Contributing Editor
| Reprints
Goal is to spread the proven Advanced Preventive Care model

The Peterson Center on Healthcare has entered into a partnership with the Camden Coalition of Healthcare Providers and Health Quality Partners (HQP) to improve care and lower costs for high-need patients.

The new project, made possible through a $605,000 grant from the Peterson Center, will work to accelerate adoption of the Advanced Preventive Care (APC) model, which the partners say has been shown to decrease mortality (by 22 percent), reduce hospitalizations (by 25 percent), and lower healthcare costs among chronically ill, older adults (by 10 percent).

The APC model prevents avoidable complications among chronically ill, older adults by filling the gap between office-based primary care and later stage interventions. Skilled nurses closely assess, monitor and engage participants and their families to proactively mitigate health risks and preventable complications. While several health systems have implemented the model, more work is needed to help other health systems adopt complex models such as APC.

 “Improving care and lowering costs for high-need patients should be among our highest priorities in healthcare,” said Jay Want, M.D., executive director of the Peterson Center., in a prepared statement.  “By spreading the proven Advanced Preventive Care model, this exciting partnership has great potential to help the growing population of high-need patients nationwide who require well-coordinated care, but too often are not getting it.”

Through the project, the Camden Coalition’s National Center for Complex Health and Social Needs and HQP will engage a multidisciplinary group of experts (from disciplines including healthcare, implementation science, user-centered design, and evaluation) to design more efficient implementation methodologies and infrastructures for health systems to incorporate APC into their patient care. The Camden Coalition and HQP will also develop an evaluation framework to test the newly designed systems for replication and share key learnings with the field that may help to further scale similar efforts.


More From Healthcare Informatics


Connecting Medical Care and Social Service Needs: Chicago Health Leaders Take Charge

August 21, 2018
by Rajiv Leventhal, Managing Editor
| Reprints
A new project will investigate the most effective means for connecting medical care and social service needs

In the new and constantly evolving healthcare landscape, patient care organization leaders are now coming to a near universal acceptance that paying more attention to social determinants of health data has become vital to improving care and lowering costs. Indeed, across the country, there are countless projects that are examining, in various ways, how health outcomes are influenced by factors such as social circumstances, behavioral patterns and environmental exposures.

On example of this is in Chicago, where a group of healthcare leaders have been working on creating a seamless and purposeful link for sharing health and other information between social services agencies and patient-centered medical homes (PCMHs) to improve the health of those most at-risk and address social determinants of health.

The Medical Home Network (MHN) is an organization that manages a Medicaid ACO (accountable care organization), comprised of nine federally qualified health centers (FQHCs) and three hospital systems serving approximately 80,000 Chicago area Medicaid beneficiaries. Earlier this year, MHN and the National Committee for Quality Assurance (NCQA) launched a joint research project to study the effects of connections between Medicaid patient-centered medical homes and community-based organizations (CBOs) that provide social services. For this project, a CBO is defined as any social service provider organization that serves an individual but is not responsible for providing “whole person care” in the same way that a PCMH or a hospital is held responsible, officials noted.

One of the driving factors for the research grant—managed by the Systems for Action National Program Office with support from the Robert Wood Johnson Foundation—according to officials, is that addressing social risk factors has been challenging for medical homes as most do not have adequate mechanisms for referring patients to services. Additionally, most CBOs do not have electronic health records (EHRs) and cannot easily communicate with medical home EHRs. Issues of protected health information and privacy have also blocked progress in this area, according to NCQA and MHN executives.

As such, through this research, the organizations set out to investigate the impact of connecting medical homes and CBOs using a web-based system. More specifically, NCQA, MHN and Cook County Health & Hospitals System (CCHHS) are partnering in this effort with medical homes and CBOs in Cook County, Illinois. As many as 200 medical homes and 25 hospitals will link to a variety of CBOs using a web-based communication and care management platform, known as MHNConnect.

“We want to see how connecting medical and social service care teams can help people with social risks,” Sarah Hudson Scholle, vice president of research and analysis, NCQA, said at the time of the project launch. “For example, if the medical care team and social service agencies coordinate to help people with immediate problems (substance abuse, food, housing, jobs), will that help them manage their health needs better or keep them out of the hospital?”

Cheryl Lulias, president and executive director of the Medical Home Network, notes that the need for this connectivity is an outgrowth of MHN’s core premise of creating a community care record across all venues, while being able to communicate and collaborate between the agencies caring for its patients, and the care teams at the primary care practices who are managing the population throughout the continuum. So the question became, Lulias, says, “How do we connect and communicate, and enable communication in a meaningful way?”

Prior to this project, MHN had built foundational connectivity, and was exchanging real-time alerts and a longitudinal record between 27 hospitals and about 200 medical homes in Chicago. “We connected the acute system, but that wasn’t enough,” says Lulias. “Then we moved to start to connect the sub-acute [system], the behavioral health [facilities] and the community agencies. There are a lot of great systems on the market that do referrals to social service agencies from the medical home primary care practice, but we want to enable conversations and share relevant information on the patient to enable seamless transitions and ongoing communication between care teams serving the patients, as well as provide a more coordinated expedience,” she says.

As an organization, MHN is no stranger to researching how social determinants of health affect patient outcomes. A prior study done by researchers at MHN, and others, found that many risk factors outside of the traditional medical model may be associated with higher utilization and costs. That research, published last year in the Journal of Community Medicine and Public Health Care, found that some addressable factors are associated with greater medical and pharmacy spending, such as needing help getting food, clothing or housing, reporting fair or poor health status, and experiencing transportation challenges. The six most common addressable factors were all associated with higher hospital readmissions; most of them were linked significantly to greater subsequent inpatient stays and ED visits, according to the study’s results.

“We’ve proved that these social issues are critical to predicting the rising risk patients and predicting prospective cost and utilization,” Lulias asserts. “We now know that someone with transportation issues is X percent more likely to have a readmission or to go the ER within three months.” As such, the next step of MHN’s work, she adds, is to analyze how connecting medical homes and CBOs affect the use of hospital and ED services, and affect health outcomes of at-risk populations.

This project is currently in phase one, in which 12 behavioral health agencies and a home health agency in Cook County have access to the Community Care Connect tool—which is a module within MHNConnect—to simply search for patients and better coordinate care for those that are seeking services at their facility, says Sana Syal, project manager, Medical Home Network.

“What we are building out now, and finalizing the requirements for, is opening the messaging center to share that information and coordinate care between care managers at primary care settings and a case manager on the other end, which is likely a behavioral health or home health agency,” Syal explains. She notes that while there are plenty of good resources that provide social service directories for care managers to be able to refer patients, oftentimes, those referrals happen on paper or by phone, which in turn creates a gap in truly knowing what happened to that patient. “Did the patient have a good experience when I referred him or her to the food pantry, for instance? We can track those referrals and close the loop so that we are coordinating care in the best way,” she says.

Syal adds that referrals “have been happening since there were patients,” but the interest in what has happened at that service or that agency, and knowing how to track that, has long been a gap in care. “And the [providers] wouldn’t know unless the patient comes back and tells them. Maybe the patient says that he or she wasn’t eligible for what the [provider] referred him or her to, or the patient was turned away, or the wait time was three months. We are looking to fill that gap by connecting these different entities together,” she says.

What’s more, in addition to doing quantitative analysis, by comparing pre- and post-implementation of the web-based platform, MHN and NCQA will also be conducting interviews of patients and their providers, offers Keri Christensen, director for research innovation at NCQA, and formerly at the Medical Home Network. This will involve interviewing a patient, a community-based organization staff member, and a medical home staff member, she says. “The patient will be at the center of the interview and we’ll also be interviewing the two care providers from the two different organizations that are caring for him or her. We want to understand how the connectivity has assisted in their workflows and how they have seen things change over time—both for the specific patient we are talking about and for the organization as a whole,” says Christensen.

Going forward, both MHN and NCQA leaders believe that this project will further prove the value in this new paradigm of connectivity to enable better coordination and better health. “It’s a simple concept, but not one replicated in many places today,” Lulias attests. “And it’s all part of the need for better coordination to drive better care. I hope that this connectivity becomes a ‘need,’ as opposed to a ‘nice to have’ when it comes to population health, and connecting the social with the medical,” she says.

Related Insights For: Population Health


San Diego Connects the Dots to Create Social Snapshot of Clients

| Reprints
Community Information Exchange shares client-level data and participating organizations use a common risk rating tool
Click To View Gallery

This week my colleague Mark Hagland is covering the Strategic Health Information Exchange Collaborative (SHIEC) Conference in Atlanta. At last year’s SHIEC conference, Mark covered a panel session featuring executives of 2-1-1 San Diego. In that talk, William York executive vice president, said the key to its work is “connecting the dots to create a social snapshot of a client’s situation, and matching that with a database of social service providers and referrals.”

I was reminded of Mark’s story from last year because last week I saw another fascinating presentation on 2-1-1 San Diego’s Community Information Exchange in a webinar co-hosted by the Center for Health Care Strategies and Nonprofit Finance Fund.

Lots of health systems are starting to work on a better understanding of social determinants of health and better connections with social service agencies. I think the effort in San Diego is really a leading model. Its Community Information Exchange (CIE) platform shares client-level data and has participating organizations use a common risk rating tool. The CIE also facilitates community case planning, and care team communications to better address the social determinants of health.

Camey Christenson, senior vice president at 2-1-1 San Diego, said that when people hear the phrase community information exchange, they focus on the technology platform, “but the important piece of the CIE is that it is not just a technology platform, but rather a collective movement of what we are trying to accomplish in our region.”  

She said the 2-1-1 organization was motivated to create the CIE by seeing regional system failures. “We receive 1,500 calls a day. We saw every day how the system of getting people resources was not working and was putting the onus on clients in crisis. Putting prescription pads in doctor’s offices telling them to call 2-1-1 wasn’t sufficient,” she said. “We needed to move to a proactive model to have helpers working more closely together to use data to break down data silos between sectors, especially including health and social services.”

The 45 network partners are connected by the technology platform, the bidirectional referrals and a shared language using a risk rating scale. “We have leveraged our role in the community as a resource hub, worked with our partners to build trust, and created a network of diverse, cross-sector partners who were willing to take the leap and redefine what a client is – that it extends just beyond their four walls,” Christenson said. They had to change their business processes —accepting and confirming referrals, and sharing client-level data, and having that shared language. “We had to make sure we understand and agree on definitions of social determinants of health and how we measure them, and that is why we created a risk rating scale using 14 different domains of a social determinants of health framework. That creates a shared language about where the client is in terms of risk for each domain, and shows changes over time with a longitudinal client record.”

Several organizations are contributing client-level data into the CIE, and clients are consenting to the use of their data for ongoing care coordination across sectors. “This is changing our field to move to more proactive, person-centered work, which is starting to have an impact on community health,” she said.

The 2-1-1 San Diego team worked closely with San Diego Health Connect, the regional HIE, to learn about data sharing and what agreements and authorizations are required. “It was a learning curve for us,” Christenson said. They also worked to leverage technology in different sectors. For instance, they focused on understanding the systems that county governments use and how to connect to them. Housing providers use a system called the Housing Management Information System (HMIS). “We directly connected that HMIS system to the CIE, which was really useful information for healthcare providers,” she said. “We also leverage our existing 2-1-1 database so we could create closed-loop electronic bidirectional referrals using our database.”

The 2-1-1 organization is working with local universities to study whether the interventions are having an impact on outcomes such as reduced hospital readmissions.  It is governed by a board of directors and has an advisory committee with work groups made up of nonprofit CEOs, and executives from health plans and local universities.

There are many examples of groups working together on systems that link social services agencies and health systems. In fact, later this week I am going to interview people working on a community health record in Alameda County, Calif. But it sounds like this effort in San Diego has already tackled a lot of the technology and governance challenges, and is really starting to reap some of the benefits.





See more on Population Health