Patient-generated Data: Integrating Information for Improved Health

More and more care delivery organizations are seeking ways to incorporate patient-generated health data into their EMRs. There’s a growing appreciation for the value of that patient-reported information–research indicates improved outcomes particularly when the information is used to track chronic conditions. Plus, they realize that most health and care happens outside of the precious few minutes patients get with their physicians each year.

The range of devices and apps available to consumers is making it easier for people to get more engaged with their healthcare. From sensors to apps to smartwatches, there are myriad options.

Aside from availability of these tools driving more integration of patient-generated data, the federal government’s push for value-based reimbursement will put quality and outcomes at a premium, says Joseph Kvedar, MD, director of Partners HealthCare’s Center for Connected Health in Boston. “In our experience, those types of models really favor the adoption of virtual care.”

All these tools for generating more data need a value proposition, however, before they can have an impact, he says. “It’s not as simple as throwing out an app and people will use it. The challenge is around interacting with and engaging patients and consumers.”

But it’s no easy task keeping up with the pace of change. Virtual visits using video technology was brand new three to five years ago, says Kvedar. He predicts a second wave based on the infrastructure, risk contracts and predictive modeling software providers are working on now. Once providers are using such tools, “they’ll be able to see value in bringing in patient-generated health data and integrate them into care delivery programs to improve quality and outcomes as well as lower costs.”

Partners uses a repository where patient-generated data are stored and electronic interfaces collect these data. The information is displayed in Partners’ EHR and patient portal. Kvedar calls it an unusual approach.

More common is a portal where patients can upload their steps or blood pressure readings to their own personal heath record. Patients can connect their devices and share their datastreams with just a few clicks. Partners’ approach is much more deliberate, says Kvedar. “We know the data will add value particularly in an era of risk-based payment.”

One problem with these datastreams is the fear among physicians that they will cause an overwhelming number of alerts and reminders. “They say, ‘I don’t have any use for that.’ These are folks who are saying they don’t have any more to learn about their patients other than what they see in the office,” says Kvedar. “To say we don’t need it is true folly. They’re really saying they’d like to have it curated so it’s information rather than data.” This is exactly the kind of information that just about every other industry has been using to tailor their offerings to their customers, he points out.

Pilots push more data

Ochsner Health System’s Center for Innovation is another organization utilizing patient-generated data, launching its efforts with a project that got digital scales into the homes of heart failure patients. The scales were Wifi-connected into the system’s EMR for 30 days to monitor patients’ weight to prevent readmissions.

“Weight can be a very good indicator of problems at home and can indicate readmission,” says Jonathan Wilt, assistant vice president of the center in New Orleans. A team of pharmacists reviewed the weights every day and initiated interventions when appropriate.

Patients loved the program and had no problems using the scale which was pre-hooked up with mobile Wifi. “They only had to take home this kit which was preconfigured to come into our system. They just plug it in and we do all the work.”

The project was very successful, he says, reducing readmissions among this group of patients by 45 percent. “It’s amazing to think that just stepping on the scale every day would make that much of an impact.”

That kind of success drives more efforts. While there have been programs aiming to reduce heart failure readmissions for years, Wilt says they knew that if the weight pilot was successful, “it would be immediately successful.” Building on those results, they are in the early stages of a new pilot for patients with uncontrolled high blood pressure. These patients sign up to submit their blood pressure reading every day and, again, a team of pharmacists review the readings to check for problems that require intervention.

The goal is “more data and more touch points,” says Wilt. Some patients only see their physician once or twice a year, and it’s difficult for those clinicians to manage a condition with just a couple of readings. So, physician buy-in was pretty easy, he reports. With journal articles now out there showing that patient-generated data increases the number of patients in control of their blood pressure, “we didn’t need much convincing that it would work out for our organization.”

Aligned with Apple

Ochsner’s blood pressure program will use the Apple HealthKit which “opens the door for more different brands of devices that can integrate with our EMR,” he says. A lot of patients have uncontrolled high blood pressure so there is the potential for a wide variety of devices. “We wanted options and any Healthkit-enabled blood pressure cuff will work.”

Ochsner was quick to align with the Apple HealthKit offering when it was released last fall because interfacing with every device manufacturer for integration is challenging. “The app did that work for us,” says Wilt. “A consolidated pipeline into the EMR was a no-brainer for us” and HealthKits lets his team concentrate on maintaining connections. “Having consistent, reliable data sources without a lot of effort on IT’s part is really big.”

An evolving market

The University of Pittsburgh Medical Center also has been integrating patient-generated data around medical history, medications and clinical conditions through its patient portal, says Vivek Reddy, MD, CMIO of the organization’s Health Services Division.

Patients can add and validate information prior to visits and data are prepopulated into the right parts of the record. “That’s been a huge success and it does a lot to make sure the data we have are accurate and accurately reflect what’s going on with the patient. It’s a good two-way discussion point.”

The organization also is working on collecting data outcomes to ensure appropriate interventions about pain control, depression and other factors influencing patients’ health. They push questionnaires to patients and those results go right into the EMR. “We can use that in the future to figure out what interventions are actually working through a data sampling.”

The next foray is wearable devices, says Reddy. “There are challenges in terms of being able to actively collect that data and import them into an EMR easily. There are not a ton of standards out there.” The market is crowded with lots of apps that aren’t certified. With no real governance, it’s hard to know “where to put our chips,” he says.

Like Wilt, Reddy is conscious of any extra work put on clinicians. Use of questionnaires drove more traffic to their patient portal along with more queries from patients. “We went from limited interactions to now medical questions and extra things that have to be sorted out. Providers enjoy that it’s digital and very easy but we’re very, very cognizant that we balance how much providers have to wade through especially if it’s not something actionable.”

Ochsner plans to continue using patient-generated data to support chronic disease management, says Wilt. Current technology allows for tracking of the behaviors, such as activity, weight and alcohol consumption, that strongly impact these conditions.

“Nobody would say it’s super easy to manage these diseases. We need to supplement our data sources. We need more touch point with these patients to try to change the big, influencing factors in why they have a chronic disease.” Patient-entered data is very beneficial, he says. “If the care team sees the data coming in, they can intervene more frequently.” They also can correct some of the behaviors that influence chronic disease right from the beginning.

Wilt cites a study that found healthcare influences about 10 percent of overall health. “If that’s true, we need to touch on other influences like behavior and social circumstances. The more data you get, the more likely you’re going to be able to do that.”

Outside of individual patients, patient-generated data contributes to overall patterns in treatments and outcomes. Data don’t have “to be actionable instantly,” says Wilt. “Trends can be just as powerful as individual data points. Blood pressure goes up and down throughout the day so taking it once every six months is not a very accurate indicator of overall trends.” IT’s job is to summarize these data and offer them back to clinicians in a meaningful way.

Reddy says he hopes for “strength in coordination and cooperation.” Once devices can crosslink datasets, “there will be interesting triangulation moments that give us a bigger picture.” That will also lead to some standardization, he says. And as long as there are enough identifiers and the patient wants to share, clinicians can get to the information. “I would love to see some of the big players in the health data collection or patient-generation health data space develop layers that allow people to easily jump in and say I want to give this to my doctor and we figure out a way to pull that down.” He says that’s a realistic goal for the next 18 months.

“We’re excited about it. This is going to be a big space for us.”