Team Model, Big Data Bring Precision Medicine to Primary Care

Ricki Lewis, PhD

June 25, 2019

A pilot project tests a new model of primary care that incorporates collection and analysis of personal physiologic and genetic data into the electronic health record to inform personalized care plans with unique features of the whole patient.

Megan R. Mahoney, MD, and Steven M. Asch, MD, MPH, from the Division of Primary Care and Population Health, Stanford University, Palo Alto, California, describe the pilot project in an article published online in the May/June issue of Annals of Family Medicine.

The Humanwide demonstration project is part of Stanford Medicine's Precision Health vision. It focuses on disease prevention by identifying people at risk for cancer and cardiovascular disease. The use of wearable devices is a large component of the program.

Another Stanford team, led by Michael Snyder, PhD, recently identified 67 clinically actionable findings among 109 participants in an 8-year longitudinal assessment that included clinical testing, multiomics big data, and input from wearable devices. That study provided a proof of concept of the value of amassing patient data. The participants were selected for elevated type 2 diabetes risk.

"As in Dr Snyder's study, data were collected for each of 50 participants on a variety of factors known to influence health, but Humanwide also included visits with a collaborative care team that included a physician, nutritionist, behavioral health specialist, and clinical pharmacist. Working with a certified health coach, that team partnered with the patient to create a detailed, personalized care plan," Mahoney told Medscape Medical News.

Although 68% of the participants in Snyder's study discussed findings with their physicians, that was not the primary goal, Mahoney said. "Both studies demonstrated the promise of using health data and other information to more precisely predict and prevent disease for individual patients," she added.

The researchers recruited the diverse group of 50 patients from January to July 2018. Ages ranged from 24 to 86 years. More than half of the participants were obese, and 58% had a major cardiovascular risk factor, Mahoney explained. Three patients were undergoing cancer treatment, one was awaiting a kidney transplant, one had had a kidney transplant, and three patients had undergone cardiac bypass surgery.

Data collected from wearable devices and home scales reported directly to the electronic health record. These included data from pedometer readings, blood pressure measurements, and glucose monitoring.

Each participant had an "in-person wellness-oriented behavioral, social, and environmental health assessment," the authors state. Genetic tests revealed pharmacogenetic (drug-gene) interactions and more than 145 gene variants that were related to inherited cancers, cardiovascular disease, and other actionable conditions.

The data were presented in graphic form to clearly reveal the beginnings or progression of disease. The information enabled the team to stay ahead of symptoms and lower the risk for complications.

For 3 years prior to the year-long study, the researchers investigated the dynamics of the care teams that are a critical part of the plan, which they call "the Primary Care 2.0 model." In addition to the primary care provider, the team includes a registered nurse, a clinical pharmacist, a behavioral health specialist, a nutritionist, a certified health coach, and a genetics counselor.

The team "frees up the physician by 'sharing the care,' distributing responsibility for patients, and assigning a medical assistant as care coordinator. These teams meet regularly and communicate constantly about their patients," said Mahoney.

Following testing, patients received in-person and virtual wellness coaching and referrals to community resources.

Initial feedback from a trial run of Primary Care 2.0 at a Santa Clara clinic indicated enthusiasm for the teams from practitioners and patients. Implementation required no additional staffing, but team members needed training on using the mobile digital devices and on the health coaching and genetic testing aspects of the program.

Most patients joined the project for access to genetic testing, and providers were highly interested in this as well, the researchers write. However, as in the larger longitudinal study, participants became interested in other parts of the evaluation.

The authors offer examples of the value of wearable devices.

"There was a diagnosis of masked hypertension, which happens when a person's blood pressure readings are normal at the doctor's office but actually high enough in regular life to carry a heightened risk of cardiovascular problems. We worked with him to manage the condition and lower his risk," Mahoney said.

Pharmacogenetic testing indicated optimal dosages and enabled evaluation of drug combinations on the basis of the rate at which an individual metabolizes a drug. Findings for 1 in 4 patients led to changes in medication for treating chronic conditions. For example, one patient with leg cramps learned that the pain arose from the fact that the patient metabolized statins slowly. A lower dose prevented the problem.

Of 33 women screened for breast cancer risk, five had a high-risk mutation, and enhanced surveillance was begun. Routine screenings would not have detected the risk, Mahoney said.

"Humanwide was designed to bring a precision health approach to a clinical setting. Our early experience demonstrated the feasibility of a more comprehensive patient-centered, data-driven environment in primary care," she concluded.

The researchers have disclosed no relevant financial relationships.

Ann Fam Med. 2019;17:273. Full text

Follow Medscape on Facebook, Twitter, Instagram, and YouTube


Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.
Post as: