Share to Facebook Share to Twitter Share to LinkedIn Pinterest Email

By Teresa Carr
Einstein and Montefiore scholars meld research with data to improve patient care

All too often, cancer cells can break away from breast, prostate, and other tumors and travel through the bloodstream to different parts of the body. For some 330,000 Americans, that means living with cancer that has metastasized, or spread, to their bones.

Metastases can crop up anywhere in the skeleton, most commonly in the spine, where they can weaken bones and lead to fractures. Identifying and treating spinal tumors early—before they can cause severe pain and disability—helps people live longer and better lives. But that’s easier said than done.

“Early symptoms, such as back pain or fatigue, caused by bone metastases are so subtle and common that they’re easily overlooked,” says Parsa Mirhaji, M.D., Ph.D., director of clinical research informatics at Einstein and Montefiore and director of the Center for Health Data Innovations. He notes that the tumors themselves are rare and, in the earliest stages, hard to spot on images—even by highly skilled radiologists.

“A radiologist may review 50,000 MRIs of patients with back pain and only 10 will have a spine metastasis; it’s like looking for the proverbial needle in a haystack,” Dr. Mirhaji says. “And even when scans reveal a possible tumor, it can take weeks to follow up and get patients to the appropriate specialist for treatment—a critical delay that can allow tumors to progress.”

EXPLORE grant co-principal investigators, from left, Chinazo Cunningham, M.D., M.S.; Michael Rinke, M.D., Ph.D.; and Paul Marantz, M.D., M.P.H.

A Strategy for Complicated Cases

In order to solve complex medical problems, a talented group of young researchers at Einstein and Montefiore is harnessing the power of something called a learning health system (LHS). Combining research, data, and a culture dedicated to continuous improvement, the LHS educates physicians and develops ways to more efficiently care for people. Not only do patients benefit from state-of-the-art care, but their experiences are helping drive the future of medical research.

“In the simplest terms, an LHS is a system where research and data collection inform practice and, in turn, practice influences research,” says Michael Rinke, M.D., Ph.D., associate professor of pediatrics at Einstein and co-director of the Health Research Implementation Core at the Harold and Muriel Block Institute for Clinical and Translational Research at Einstein and Montefiore. “With an LHS, you’re constantly improving and innovating to achieve best practices, which you incorporate into the healthcare you deliver.”

An LHS “epitomizes evidence-based medicine and is the best way to get at complex medical issues like detecting spinal metastases,” says Vijay Yanamadala, M.D., M.B.A., M.S., assistant professor of neurological surgery at Einstein and director of the Center for Surgical Optimization at Montefiore. In fact, he says, the logic behind an LHS is not unique to healthcare.

“To devise a game plan, football coaches use what they learn from trying different configurations and plays during practice,” Dr. Yanamadala says. “After game day, the team goes over the films to see what worked and didn’t work, and it then builds on its analysis of those data to develop a better strategy.”

In the simplest terms, an LHS is a system where research and data collection inform practice and, in turn, practice influences research.

— Dr. Michael Rinke

But medicine has been reluctant to let data steer it toward better healthcare. “Multiple studies show that when a new and better way to treat patients comes out, it takes about 17 years to be widely adopted,” Dr. Rinke says. “The classic example we give in quality improvement is that we’ve all tried to change something about ourselves—exercise more, for instance, or get more sleep—and it’s really challenging. Changing things in a workplace can be just as hard.”

One of the main goals of LHS research is to greatly shorten that 17-year lag time, he says. And Einstein and Montefiore are at the forefront of this approach. “We have incredible researchers who analyze data—primarily information from patients’ electronic health records—to continuously generate information aimed at solving problems,” Dr. Rinke says. “But just as important, once a new treatment or tool has been rigorously tested, we have an equally strong operations group to integrate it into our systems to make sure that all patients benefit.”

Dr. Yanamadala, in collaboration with Dr. Mirhaji, is using LHS practices to detect and treat spinal tumors. Dr. Mirhaji is creating a machine-learning algorithm to do what no human can: analyze millions of pieces of data—images, lab results, doctors’ notes, and more—to uncover the patterns indicating that a patient’s cancer may have spread.

“Physicians are only human, so they can factor only about seven variables—things like blood pressure, pulse rate, and body temperature—into their decision-making process,” Dr. Mirhaji says. “But the algorithm can look at hundreds of different variables and draw conclusions regarding a patient’s condition from that information.”

Dr. Yanamadala, for his part, is interviewing patients and their providers and collecting other patient-care data to uncover roadblocks that cause delays. The goal is to create a center staffed by a variety of specialists, including neurosurgeons, radiation oncologists, and physiatrists. The new center will enable patients to receive prompt, appropriate treatment in one place for spinal metastases.

“This way of operating—constantly collecting and analyzing data and using what you learn to continually improve—is where medicine is headed,” Dr. Yanamadala says. “And Einstein and Montefiore are at the cutting edge, which is what drew me here.”

Parsa Mirhaji, M.D., Ph.D., processes data generated by the EXPLORE scholars. Physicians can factor about seven variables into their decision-making process, he says.

The New Faces of Medical Research

Dr. Yanamadala is one of four scholars who’ve recently joined the new Einstein and Montefiore LHS center—known as the Excellence in Promoting LHS Operations and Research at Einstein and Montefiore, or EXPLORE. It was established in November 2018 with a $3.3 million federal grant—one of only 11 LHS Centers of Excellence in the nation and the only one in the state of New York.

“This is the first federal grant to train investigators in the LHS model,” says Dr. Rinke, who is also a co–principal investigator for the EXPLORE grant. Traditionally, promising physician-scientists such as Dr. Yanamadala approach medical institutions with research questions they’d like to pursue. EXPLORE turns that model on its head, Dr. Rinke says. It first identifies high-priority questions—such as how to detect bone metastases early—and then seeks applicants with the experience and the intense curiosity to answer those questions.

“A major advantage to that approach is that we can have all the resources in place for the scholars to hit the ground running,” says Paul Marantz, M.D., M.P.H., associate dean for clinical research education and a co-principal investigator on the grant. He notes that each scholar will have both clinical and health-systems mentors, plus access to expertise in sophisticated data processing.

The abundance of medical data now stored digitally “amounts to a treasure trove for an LHS center like EXPLORE,” Dr. Marantz says. “At the heart of the LHS model is the notion that we can use all those data—not only to figure out the right things to do clinically, but to influence complex health systems to adapt and be willing to implement new evidence.” To that end, he says, EXPLORE brings together professionals from all parts of the healthcare system—from physicians and nurses to data scientists and pharmacists. “Their collaboration is key in a learning health system,” he emphasizes.

“There’s great value in having people from different disciplines and different levels of the health system leadership sitting around the same table,” says Chinazo Cunningham, M.D., M.S., a co–principal investigator for the EXPLORE grant. “That takes discussions to another level where we’re able to see beyond our respective roles and consider how best to make changes in a complex system. And by publishing what we’ve learned about improving clinical practice, we’re helping patients at other hospitals and health systems benefit as well.”

The EXPLORE program “reflects the changing culture of medical research,” says Shalom Kalnicki, M.D., chair of radiation oncology at Einstein and Montefiore and one of Dr. Yanamadala’s mentors. “Research done at wet-lab benches using test tubes and animal models is vital,” he says. “But research increasingly is being done by computer modeling, with results that can be fed into an LHS. I feel that these young investigators we’re training and mentoring in collaborative research represent the future of medicine.”

EXPLORE scholar Vijay Yanamadala, M.D., M.B.A., M.S., uses learning health system practices to detect and treat spinal metastases.

This way of operating— constantly collecting and analyzing data and using what you learn to continually improve—is where medicine is headed.

— Dr. Vijay Yanamadala

Four Scholars and Their Complex Challenges

The four Einstein and Montefiore EXPLORE scholars (three more will be selected next year) take courses and undergo training in three areas essential for a career as an LHS investigator: health-information technology, quality improvement, and clinical research. Applying their new skills and their previous experience, the scholars work with mentors to solve thorny healthcare problems.

As noted above, Dr. Yanamadala’s EXPLORE project leverages the power of an LHS to research ways to detect and treat spinal metastases. “Vijay is a skilled spine neurosurgeon, but he’s dedicated to improving care by reducing the need for surgery,” Dr. Kalnicki says. “From the bottom of his physician heart, he feels compelled to find ways to improve patients’ quality of life, to really make a difference.”

The projects of the other three EXPLORE program scholars are described below.

EXPLORE scholar Kaitlyn Philips, D.O., M.S., uses electronic health records to predict which patients are most at risk of developing sepsis.

Being able to predict which patients are at risk and to prompt doctors to quickly take preventive action can save lives.

— Dr. Kaitlyn Philips

Using Electronic Health Records to Spot Sepsis

Sepsis—the body’s overwhelming and potentially fatal response to infection—is the most expensive inpatient problem in American hospitals, costing an estimated $27 billion yearly. Kaitlyn Philips, D.O., M.S., assistant professor of pediatrics at Einstein and an attending physician at the Children’s Hospital at Montefiore (CHAM), is pursuing an EXPLORE project aimed at improving the care of sepsis patients.

Dr. Philips says that her experience treating sepsis in children, who often display more-subtle symptoms than adults, helps in her effort to recognize sepsis at an early, treatable stage. “I’m able to act as a liaison between the pediatric and adult worlds in sepsis research,” she says.

One-third of patients who die in hospitals have sepsis. But according to Dr. Philips, the challenge is diagnosing sepsis early enough to save lives—before the heart weakens and organs start shutting down. For her project, Dr. Philips is sifting through the electronic health records of patients whose organs have failed due to sepsis.

Sepsis can progress rapidly, and it’s not known why one patient develops it and another, similar patient doesn’t. “Once we can identify early signs and symptoms that these patients have in common we can modify electronic alerts and make them more usable to help physicians identify patients at high risk for sepsis and treat them,” Dr. Philips says. “Being able to predict which patients are at risk and to prompt doctors to quickly take preventive action can save lives.”

EXPLORE scholar Kevin Fiori, M.D., M.S., M.P.H., uses social-needs data on more than 50,000 Montefiore patients to identify people most in need of help from community health workers.

Montefiore already has social-needs data on over 50,000 patients—an impressive amount of information.

— Dr. Kevin Fiori

Connecting At-Risk Patients Via Big Data

After a divorce, Montefiore outpatient Fatoumata Camara lost her apartment, and she and her four children wound up living in a homeless shelter for a year. Montefiore community health worker Janet Gonzalez was determined to find the family a place to live—and succeeded. “Now that we have a home, we’re sleeping better, eating better—I’ve even lost weight,” Ms. Camara says.

Research confirms that the health of people like Ms. Camara improves when healthcare systems link them with community resources. But what’s the best way to identify those outpatients, like Ms. Camara, most in need of a community health worker’s help?

Asking all Montefiore outpatients to fill out a social-needs survey isn’t practical. So for his EXPLORE project, Kevin Fiori, M.D., M.S., M.P.H., assistant professor of pediatrics and of family and social medicine at Einstein and an attending physician at Montefiore Medical Group Pediatric Practice, is using data to identify people most likely to need social services and then developing strategies to alert clinical teams to administer surveys and, when warranted, connect those patients to various community health workers in the clinic.

Dr. Fiori recently spent 15 years directing a nongovernmental organization in Togo that uses community health workers to address health disparities and improve care. His insights from Africa should help with his EXPLORE project.

“Montefiore already has social-needs data on over 50,000 patients—an impressive amount of information,” Dr. Fiori says. “If the algorithm reveals which of those patients could benefit from taking a social-needs survey, the next step is to find the best way to efficiently integrate into a busy clinic the screening and subsequent referrals to community health workers.”

A key aspect of the research is what’s known as “implementation science,” he says. “It involves repeatedly tweaking processes to work out all the details.” For example, who’s responsible for inputting data on patients coming in for appointments so that doctors can discover who should be screened? And what’s the best way to alert physicians to administer a survey?

“Once we have optimized the screening process in our pilot clinic, we will be able to incorporate that process throughout the Montefiore system,” Dr. Fiori says. He adds that he hopes the findings from his EXPLORE research will eventually benefit patients in the larger world.

EXPLORE scholar Justina Groeger, M.D., works with a team of clinicians, performance-improvement experts, and researchers to develop a protocol for prescribing the right amount of opioids to surgical patients.

It’s exciting to pull together a multidisciplinary team to work on a project like this.  

— Dr. Justina Groeger

Using Decision Tools to Prescribe Pain Medicine

Opioid pain medications can effectively relieve pain after surgery. But opioids are often started without a clear plan for transitioning patients to safer forms of pain relief. As a result, as many as 1.6 million Americans who have surgery each year wind up taking opioids over the long term, increasing the risk of serious side effects such as opioid-use disorder and overdose.

For her EXPLORE project, Justina Groeger, M.D., assistant professor of medicine at Einstein and an internist at Montefiore, is trying to figure out how to prescribe the right amount of opioids to surgical patients.

“The severity of pain is highly dependent on the individual patient,” Dr. Groeger says, “so a one-size-fits-all approach doesn’t work when prescribing pain medication.” She is researching how to tailor opioid prescribing, starting with the amount of pain medication a patient was taking before leaving the hospital and tapering the dose over several days at home.

Using the LHS approach, Dr. Groeger is working with a team of clinicians, performance-improvement experts, and researchers to develop a protocol, test it in small groups of patients, and refine it based on feedback. “It’s exciting to pull together a multidisciplinary team to work on a project like this,” Dr. Groeger says.

Initially, Dr. Groeger is studying people who’ve had knee replacements. But once the prescribing protocol is solid, she plans to integrate it into Montefiore’s electronic-record system. “Ideally, the provider who is discharging a patient will be prompted with a recommended dose and tapering schedule for opioid pain medication as well as customized instructions for patients,” she says.

Dr. Groeger’s project also involves developing a clinical decision tool to help physicians tailor their pain medication prescribing to patients’ individual needs. Ultimately, she envisions using the clinical decision tool for other purposes as well, such as customizing doses of insulin for patients with diabetes.

Mining Data for Better Health

“Having an LHS means recognizing that every patient’s experience is an opportunity to learn as much as possible to help the next person,” Dr. Rinke says. “As data from electronic records accumulate and our ability to analyze those data improves, we’ll be relying on a pipeline of rigorously trained researchers to harness data and take us into the future. Thanks to the EXPLORE grant, we’re fortunate to have this crack squad of researchers focused on finding the best possible healthcare strategies and delivering them to patients.”

Share to Facebook Share to Twitter Share to LinkedIn Share to Pinterest Email

More From Einstein

Prize for Cancer Research
The Latest on COVID-19 Research
Stand By Me: Ph.D. Mentoring
Welcoming the Class of 2024
Finding a Home at Einstein
The Role of Autophagy in Aging
Diversity in Medicine Scholarships

Content

Research Notes

Share

Share to Facebook Share to Twitter Share to LinkedIn Email

Past Issues

Download Magazine

Search

Subscribe