By Brian Dolan
Four years ago a keynote speaker at the American Medical Informatics Association meeting in Washington D.C. made a provocative prediction about the near-future importance of patient generated health data:
"I think it's not that bold to say that in five years the kind of things I was showing you are going to put us in a position where the vast majority of information that a doctor might use to determine how a patient is doing will be things that are collected outside of that doctor's office," Gregory Abowd, Distinguished Professor, School of Interactive Computing Georgia Tech, told AMIA 2011 attendees. "The technologies of ubiquitous computing are very well positioned to make this a reality... you need to recognize that this transition is going to happen, and -- it's OK to handle the problem of electronic medical record deployments in family physicians across the US and the health information exchange channels that reside there -- but do not forget the point that much of the data that might be relevant to the medical record of an individual is going to be accumulated in their homes and elsewhere."
Abowd's research into text message-based patient engagement and connected home systems helped lead him to these predictions. While it may take more than two more years, Abowd's 2011 keynote address is poised to be prophetic. Since then there has been a growing recognition of the importance of patient generated health data -- even leading to serious discussions about including requirements for it in future Meaningful Use stages.
Last year the always pioneering Veterans Health Administration threw its weight behind this trend's increasingly inevitable rise.
“The future is PG[H]D – patient-generated [health] data,” Dr. Susan Woods, director of patient experience for connected health at the Veterans Health Administration said last summer. “The voice of the patients and the caregivers has never been louder.”
Clinicians who resist growing demand to cede some of their authority and to accept health data from patients would be wise not to ignore changing expectations, for the sake of better care, Woods suggested. “Transparency breeds trust. Trust between patients and clinicians breeds outcomes,” she said.
Patients with chronic conditions have long understood that what happens between doctor visits is often just as or even more important than what happens during them.
“I see my doctors as consultants,” Anna McCollister-Slipp, Co-Founder of Galileo Analytics told attendees at the mHealth Summit last year. “They don’t manage my disease, I do. I make dosing choices 24-7, where I’m constantly evaluating what the data from my blood glucose monitor says. Whether or not I’m going to go to the gym or where or not I’m going to go to lunch. I make decisions about how much insulin to take and diabetes is taking an extreme example, just because insulin is lethal if you get it wrong, but everybody in many respects needs to be doing the same kind of thing and ultimately will be as we continue to live longer.”
Thanks to changing payment models and increased focus on accountable care, more providers are investigating the use of population health management platforms that help them stay in touch with their patients between visits. While most of these currently employ relatively simple, health risk assessment-like surveys via texts or secure portals, the opportunity for adding data from personal health devices and, in some cases, even lifestyle tracking devices is of increasing interest, too.
Health researchers have recently begun to incorporate off-the-shelf, smartphone-enabled personal health devices in large scale studies. UCSF's Health eHeart initiative is one high profile example along with the Wired for Health trial at Scripps Translational Science Institute in partnership with Qualcomm. Many expect that the results of initiatives like these will not only produce meaningful clinical data, but could also indirectly help convince providers that the data generated by personal health devices are worth considering as part of their own conversations with patients.
Rob McCray, President and CEO of the Wireless-Life Sciences Alliance, told MobiHealthNews a few months ago that patient generated data is becoming an essential part of healthcare.
“In my opinion, the core issue is that HIT systems of the future must adapt to enable the use of data that is generated by patients when they are not in the clinic,” he wrote. “Patient engagement in their own health and wellbeing is central to plans to decrease the cost of health care and improve population health. HIT systems that do not support these developments should be obsolete and certainly should not be supported by taxpayers.”
Challenges for patient generated health data
By Aditi Pai
Although patient generated data creates a fuller picture of a patient's day-to-day life, there are various challenges for healthcare providers to get over such as maintaining privacy and security, dealing with expenses, and convincing physicians that the device data is useful, trustworthy, and manageable.
Privacy Concerns
For patients, there's a risk that more organizations than just a health provider will ultimately see that data. A survey from San Diego’s California Institute for Telecommunications and Information Technology (Calit2), funded by the Robert Wood Johnson Foundation, found that 90 percent of respondents said it was important that their data remained anonymous before agreeing to make their self-tracked, personal health data available to researchers. Additionally, 13 percent of respondents specifically mentioned an aversion to commercial or profit-making use of their data.
Organizations that could potentially benefit from this data include employers and insurance companies. Of course, there's also the off chance that data will also be leaked by accident, as it was in 2011 when news broke that Fitbit user profiles were available publicly, and with the profiles was information about sexual activity, weight and other sensitive metrics.
Costs
Even if patients trust where their data is going, the costs for sending this data to healthcare providers are steep. Most devices connect to smartphones, so communities in which smartphones are not commonplace -- and health trackers are even less so -- may have a more difficult time providing information to physicians.
On the healthcare provider's side, there's the issue of piling on more costs in an effort to integrate this data into the patient's records. Association for the Advancement of Medical Instrumentation VP of Standards Development and Emerging Technologies Joe Lewelling said at the mHealth Summit near Washington DC last year that the immediate costs to bring interoperability to medical systems are a significant barrier.
“I think we can all see that in the end we’re going to save money,” Lewelling said. “But there are going to be a lot of transactional costs before we reach that point. And there are going to be transactional costs in terms of patient safety because any time you change technology, you have to change behaviors.”
He added that while some patients can manage their own healthcare, he believes there are others out there who can’t and for them the question might be, “how do we assist them?” and “how do we educate them?”
Companies similar to Tyto, which offers users a handheld device that can help patients examine their mouth, throat, eye, heart, lung, and skin, might be one answer to Lewelling's question. These exams can be guided by a physician online or guided by the device offline.
Trusting the Data
Although, there's also the issue of getting physicians and other care providers to trust the data that patients provide.
Rick Valencia, Qualcomm VP and general manager of Qualcomm Life, said that the biggest concern from doctors is trusting that patients took their own health readings correctly or aren't accidentally submitting readings from someone else.
"Health systems and doctors have been trained for the one on one interaction," he said. "So just shifting and saying now we're going to do that remotely and we'll send it all to you? Doctors aren't going to embrace that until they know exactly where it's coming from."
Physicians may be more willing to trust a digital health tool's readings than a patient's manual entries. When pharmaceutical company Sanofi and its partner device maker Agamatrix launched IBGStar in May 2012, a smartphone connected glucometer, the companies included a feature that differentiates glucose entries. Readings from the iBGStar device are that are automatically loaded into the app are “locked” in so users cannot edit them and are indicated by a lock symbol on the corner of the user's scorecard. Manually entered readings are marked with an “x” to indicate that they are editable and were manually entered. This feature suggests that Sanofi thought healthcare providers would be interested to know which readings came from the device and which were not edited.
Two years later, care providers are still having trouble with integrating a patient's self-reported data in to a patient's medical record. This year, physicians at PinnacleHealth have started to recommend that patients track vital signs like blood pressure, heart rate, or activity and the health system has included a space in the portal for patients to enter that data and track it against goals.
Although currently patients mostly enter the data manually, the health system is working on integrating with devices so patients can enter data automatically from smartphone-connected fitness trackers or health devices. Still, PinnacleHealth CIO Steven Roth said at HIMSS in Orlando, Florida that they aren’t currently moving the patient-generated data into the EHR, so the doctors have to log into the portal to access the information because doctors are hesitant about making manually entered patient data part of the patient’s legal medical record.
Chris Wasden, Innovation Practice Leader at PricewaterhouseCoopers, echoed those sentiments in an interview with MobiHealthNews at HIMSS last month, in response to a question about data from consumer health trackers.
"A clinician doesn't want a copy of this information," he said. "There's a risk issue -- they're afraid if they have this information and didn't act on it they could be liable. But also most information we collect is low-risk and not very actionable. So there's a whole bunch of reasons why clinicians don't want this information. It doesn't mean it's not valuable, but it doesn't necessarily need to be integrated into the EHR. I think there's a whole world of health data that's just for you as a consumer that never needs to go into the EHR."
Managing the Data
Nonetheless, most PGHD advocates do want to see the data in the EHRs, including Valencia. The problem, he says, is refining the data into a form physicians can actually use.
"At this stage of the game, [patient data] needs to get into the EMR," Valencia said. "The last thing [doctors] need is another technology platform. The real key is integration into the EMR at the moment. Step two is how do you make all this data meaningful for a doctor? And there's a number of ways to approach it. I think the standard right now the one we see primarily, is providing the doctor with the tools to determine for themselves what's meaningful and what's not and focusing on the exceptions."
This could mean involving someone else in the process -- whether it be a nurse or a sophisticated algorithm -- to monitor the data and only bring the anomalies to a doctor's attention. Philip's IntelliBridge Enterprise eICU technology is an example of a software that analyzes several patient vitals monitors and alerts a nurse when one reading falls out of normal levels.
"I think with our capabilities, data doesn't necessarily have to go to the physician," Chuck Parker, president of the Continua Health Alliance, told MobiHealthNews. "You can have a nurse or a medical practitioner look at the data, identify those things that are concerning, and have the first line of interacting with the patient."
The prospect of Meaningful Use including advanced PGHD requirements
By Jonah Comstock
Dr. Paul Tang
The Office of the National Coordinator for Health Information Technology's Meaningful Use guidelines have done impressive work in getting EHR adoption in the United States off the ground relatively quickly. The system of carrots and sticks for hospitals and other care providers is currently in Stage 2, with a recently delayed Stage 3 set to begin in 2017.
Because Meaningful Use has proven effective at moving the needle for technology adoption, stakeholders seeking greater adoption for patient-generated health data (PGHD) are looking to the Stage 3 guidelines for help. For instance, last December a group of stakeholders representing the American Telemedicine Association, the Association for Competitive Technology, the Center for Data Innovation, Continua Health Alliance, HIMSS, IHE USA, the Telecommunications Industry Association, and the Wireless-Life Sciences Alliance sent a letter to Paul Tang, chairman of the HIT Policy Committee working on Meaningful Use Stage 3 standards.
“We believe that the HITPC should appreciate the value of PGHD by setting clear expectations in MU Stage 3 for providers, patients, and other stakeholders,” they wrote at the time. “PGHD should become elemental to the efficient delivery of healthcare.”
That policy committee met earlier this month, on March 11, and presented an overview of the current direction of Meaningful Use Stage 3. In fact, the group did add a section on PGHD, but as written the guidelines won't require hospitals to actually accept device data into their EHRs. Instead, providers will be required to accept patient-generated data in the form of either secure messaging or electronically submitted structured questionnaires. Tang explained during the meeting that the omission of a recommendation pertaining to device data was because of a perceived lack of accepted industry standards.
"You've heard us talk about [PGHD] a lot of times," he said at the time. "It was our lifelong dream to have this available by Stage 3. You see, because of the immature standards or, certainly, the immature adoption of standards in the device space, we're not able to make a recommendation in that area. But we do want to include more and more information about what's happening outside our four walls, whether it's in an ambulatory care setting or a hospital setting, and getting it in front of the provider is one of those important things to do. So even though there is some development work [to be done], and there is some work [to be done] in the provider accepting and making use of this information, we thought it's really important as we bring patients onto their healthcare team."
Chuck Parker
Chuck Parker, President and Chair of the Board of Directors of the Continua Health Alliance and one of the signers of the letter, believes there are plenty of robust, industry-developed standards for PGHD.
There's a set of standards based on open, internationally-accepted pathways that are already in place," Parker told MobiHealthNews. "Our goal is to inform the government that this has already been done. There's a robust set of mature standards that are industry capable today and are available internationally."
Continua -- now a part of HIMSS -- is on its sixth set of industry guidelines for device interoperability, Parker said, and is already working with other groups within the government, like the FDA and NIST, on these standards. He says what makes it hard to convince the HIT Policy Committee is the lack of large-scale deployments to demonstrate the existing standards.
"There are many people out there saying the standards aren't ready yet and it's because there's no wide scale deployment and no large pilots we can point to today," he said. "It's hard to move companies away from proprietary standards because they feel like they have a lock-in. They can own that relationship if they have that lock-in. From a historical perspective, some of the companies have put up roadbloacks because they want to maintain the proprietary nature of their architecture. They feel that once they move to an open architecture they can be replaced, and that really hasn't been shown to be the truth."
Qualcomm VP and Qualcomm Life general manager Rick Valencia agreed that there was work to be done on the standards, particularly getting a critical mass of device makers to adopt them.
"I think we have a long ways to go," he told MobiHealthNews. "I think it's our responsibility as an industry to figure it out. I don't think we have yet. It is a pretty big shift and an expensive development effort for these device companies and we have to come up with simpler ways for them to do it. ... We are definitely on the way and I think Continua is a great organization with the absolute right objectives. The reality ... is a lot of the device managers are not embracing it because it's expensive and they don't know what the outcomes are going to be."
There is something of a catch-22 at play, Parker acknowledged -- lack of government support is one of the things holding back the large pilots that the government is waiting to see. But it's not an insurmountable problem. And for evidence that patient-generated data can be made to work, ONC needs to look no further than other countries, which Parker and Valencia agree are currently ahead of us in terms of telehealth and telemedicine.
"When you compare us to the other developed countries, we're behind," said Valencia. "They've been using what we refer to as a telehealth and telecare for some time. I believe they have much better trust in patient-generated data than we do."
Despite the most recent line from Tang, both Parker and Valencia have confidence that when the final Meaningful Use Stage 3 guidelines are finalized, they will include PGHD from devices.
"Knowing the individuals I work with at ONC, that answer is 'yes'," Parker said. "What it looks like I can't predict yet."
"I think it's likely, but I think there's a lot of work we need to do to make that happen," Valencia said. "Everyone would like that to happen, but we've got to prove that it would be trusted and useful, before it will be included."
