How to harness the power of patient-generated data

Most doctors, if asked to define the term “patient-generated data,” probably would associate it with answers given in patient interviews or the clipboard forms new patients fill out in the waiting room. But in the computer age, patient-generated data includes far more than that. The sources of this information now range from pre-visit questionnaires and health risk assessments to functional status surveys and remote monitoring devices.

The government wants physicians and hospitals to start using this kind of information more broadly. For example, the Centers for Medicare & Medicaid Services (CMS) recommends that physicians administer health risk assessments (HRAs) to gather information for annual Medicare wellness visits.

The Stage 3 rules of Meaningful Use, still in effect for participants in the Medicaid electronic health record (EHR) incentive program, require that 5% of patients seen contribute data to their EHR. The successor to the Medicare EHR incentive program-Advancing Care Information, part of the Merit-based Incentive Payment System that is scheduled to take effect next year-has a similar though less challenging mandate.

The Office of the National Coordinator for Health IT (ONC) has also been promoting patient-generated data, especially functional status information. In the interoperability road map that ONC released last year, the agency stated that patient-generated data should be incorporated into longitudinal health records to help individuals improve their health

Patient-generated data is also essential to the value-based care delivery model that the healthcare industry is pivoting toward. In the future, physicians will have to monitor and keep in touch with patients between visits to manage population health, and they will have to be aware of every patient’s health risk factors.

“There’s no way you can move toward value-based care without incorporating patient-generated health data,” says Danny Sands, MD, an internist affiliated with Beth Israel Deaconess Medical Center (BIDMC) in Boston, Massachusetts, and an assistant professor of medicine at Harvard Medical School.

Here are some of the different kinds of patient-generated data and what some physicians are doing with it.

Patient-generated data, explained

A report from a technical expert panel for ONC defines patient-generated health information as “health-related data created, recorded, gathered or inferred by or from patients, family, personal caregivers or designees to help address a health concern. This data could be an observation, a result, a device finding, a confirmation or a change/correction/addition of data in the patient’s existing health record.”

This definition covers health-related information that a patient brings to visits, such as copies of his or her medical and family health history, notes the report. Sometimes doctors may request that a patient or a caregiver keep track of certain information and report on it at the next appointment.

For example, Jeffrey Pearson, DO, a San Marcos, California-based family physician, asks some patients to track their blood pressure or their blood sugar. Pearson may have a hypertensive patient use a blood pressure cuff to take random readings during the week after a visit to see whether his high reading during that encounter was the result of “white-coat” syndrome. The patient is asked to send a secure email message to Pearson’s patient portal, which automatically documents it in his EHR.

Pre-visit questionnaires

Among the more formal methods of tracking patient health between visits is with a pre-visit questionnaire. A growing number of practices are adopting these instruments, often to improve billing, says John Mafi, MD, MPH, an assistant professor of medicine at the UCLA David Geffen School of Medicine in Los Angeles, California.

The best-known pre-visit questionnaire is Instant Medical History, a 20-year-old survey that has been integrated with EHRs such as Cerner, Allscripts, NextGen and Greenway. Allen Wenner, MD, a family physician in Columbia, South Carolina, and the founder of Primetime Medical Software, which markets Instant Medical History, says that his system has grown to encompass more than 100,000 variables related to patients’ medical complaints. 

Instant Medical History is an interactive program that uses branching logic to ask patients questions about their illness. Patients can fill out the surveys online or on tablets in the office. The results are summarized in bullet points, with only positive answers highlighted.

Wenner says doctors can scan the summaries in under 10 seconds, absorbing information that can help them diagnose problems more quickly. Having to ask patients fewer questions, he says, can save doctors an average of four minutes, or more than 25% of a 15-minute visit.

In addition, he says, the data can help physicians refer patients more appropriately and sometimes faster. For example, he has sent patients with early signs of appendicitis to surgeons without seeing them after reviewing their survey results.

Sands, cochair of the panel that authored the ONC report, agrees that pre-visit questionnaires such as Instant Medical History can be useful. But patients must be willing to take the time to complete these surveys, he notes, and doctors must suggest it to them. 

Other observers are skeptical. Lisa Iezzoni, MD, director of the Mongan Institute for Health Policy at Massachusetts General Hospital in Boston, thinks some patients lie on questionnaires and that visual cues can help doctors get at the truth. Jeffrey Kagan, MD, an internist in Newington, Connecticut, and member of the Medical Economics editorial advisory board, says he believes patients might misunderstand some questions in pre-visit surveys. He is more comfortable with information gathered during face-to-face interviews, he says.

Health risk assessments

As mentioned earlier, CMS urges physicians to use HRAs in conjunction with the Medicare Annual Wellness Visit. The purpose of this visit is “to encourage individuals to take an active role in accurately assessing and managing their health, and consequently improve their well-being and quality of life,” according to a report on HRAs sponsored by the Centers for Disease Control and Prevention (CDC).

Administering an HRA and using it as the basis for feedback and advice to patients can help accomplish this goal, the report says.

HRAs include questions about many kinds of health risk factors, including health behaviors, notes the report’s lead author Ron Z. Goetzel, PhD, a senior scientist at the Johns Hopkins Bloomberg School of Public Health and vice president at Truven Health Analytics. For example, an HRA might ask patients how physically active they are, what kind of food they eat, how much stress they feel, and whether they smoke, he says. 

HRAs, which usually take 15 to 30 minutes to fill out, typically are completed online, Goetzel says. Most of the surveys are general, covering 10 to 15 health categories. But some are specific to a particular condition. For example, the HRA of the American Heart Association focuses on seven risk factors relevant to coronary artery disease.

How can HRAs help doctors care for individual patients? “A patient who walks in may be smoking and drinking, may be overweight, having problems at work, feeling lousy and not physically active,” Goetzel says. That information, which the patient might not feel comfortable sharing with the doctor  face to face, is now available in their HRA. “So the physician can say, ‘Here are some of the risk factors we saw in your report, let’s talk about them.’”

Jeffrey Lederman, MD, an internist in Long Branch, New Jersey, and his partner regularly use HRAs. Part of the national MDVIP network, which encourages its concierge practices to use these surveys, Lederman’s practice asks patients to fill out the HRA two weeks before their annual physical exam. They can complete the survey either online, using the MDVIP portal, or in the office. Although that portal is not yet integrated with Lederman’s EHR, he says his staff can upload the PDF of the results so he can see it when he opens a patient’s chart. 

“An HRA gives a great summary of the patient’s own self-awareness about their health care, and it can be very helpful to physicians,” he says. If a patient is in a medium risk category because of hypertension, diabetes and being overweight, he says, he can look up their physical exercise score and see that they’re not exercising enough. Then he can discuss an action plan with the patient.

In some cases, HRAs have alerted Lederman to issues he wouldn’t have thought of asking about. For example, responses on an HRA prompted him to order a sleep study for one patient, who turned out to have sleep apnea. 

Kagan has begun using a home-grown HRA in conjunction with Medicare wellness visits, he says. His partner built an EHR template for an annual wellness visit that includes  the HRA and questions about activities of daily living and the patient’s physical environment. 

Functional status surveys

Functional status surveys, which measure how patients feel about their health and their ability to cope with activities of daily living, have been around for decades. They can be used to measure outcomes and to manage the care of patients who are recovering from procedures. But they are still little-used outside of academic medical centers.

One reason is that many doctors don’t believe that the data from these surveys are as valid as objective clinical measures. Iezzoni rejects that view. 

“I think there’s always an inherent validity to what the patient says,” he says. “They’re the ones who are living in their body, they’re the ones who have to perform activities of daily living or get someone to help them with it. So even if a physician disagrees with a patient’s assessment from a patient-reported survey, it’s important for the physician to understand what the patient’s perceptions are.”

Kagan, who cares for a lot of Medicare patients and does work in nursing homes, uses a functional status survey with his hospice patients. The survey includes basic questions about their ability to talk, walk, hold their head up and so forth, he says. This kind of instrument could also be valuable to doctors in measuring the functional status of other elderly people, he adds.

At Dartmouth Hitchcock, a New Lebanon, New Hampshire health system that includes an academic medical center, functional status surveys have been used in the spine center for 20 years and, more recently, in the orthopedic surgery department. The spine center uses a 36-question general survey, plus shorter questionnaires on conditions such as back pain.

When Sohail Mirza, MD, was recruited to chair the orthopedics department in 2008, he implemented a series of short functional status surveys as part of a streamlining of the department’s overall workflow. These surveys-distributed online or in the waiting room- were integrated into the group’s EHR. For the past five years, they have been used routinely in knee replacement surgery and in hip operations. About half of the patients complete them online, and the rest do it on tablets in the waiting room.

Now medical director of Dartmouth-Hitchcock’s Center for Surgical Innovation, Mirza says that the orthopedics department has each new patient fill out a generic survey, which takes about 10 minutes. When patients undergo an elective procedure, they’re asked to complete a shorter survey with more detailed questions. The results are available in the EHR for physicians to see when they’re interviewing patients.

One purpose of the baseline functional status surveys is to help surgeons discuss the appropriateness of surgery with patients. “After five years, we have found that these baseline scores are very predictive of where patients end up three months, six months and a year after surgery,” Mirza says. “Take knee surgery. Some patients are doing so well that they’re not going to gain a whole lot more in terms of function by having a knee replacement.” If a patient’s score indicates that the patient is unlikely to benefit from a knee operation, the surgeon advises him or her of that.

The physical function scores on surveys taken after surgery help doctors gauge the speed of a patient’s recovery and decide what kind of therapy they may need, Mirza says. The score at 30 days after discharge indicates whether the patient is on track with other people like them, and how much more rehab they are likely to need.

Sharing visit notes

In recent years, the idea of sharing visit notes with patients has gained traction among physicians, partly because of the Open Notes studies done by Harvard Medical School researchers. These studies indicated that visit note sharing increases patient engagement and medication adherence.

Now some of the same researchers are working on a pilot to find out what happens when patients are encouraged to comment on and add to their doctors’ notes. In addition, an offshoot of the Open Notes project called Our Notes is investigating the concept of having patients contribute to their notes before, during and after visits. 

UCLA’s Mafi leads the Our Notes project. The aim of his research, he says, is to find out how to get patients more engaged in their own care. Pre-visit data entry is already occurring in some practices, and doctors who share their notes with patients may receive comments from them, he points out.

But experts have told Mafi’s team that having patients co-document encounters with physicians is a challenge because of workflow issues. “This is rare and I don’t think we’re ready for it yet, but it could be part of the future,” he says.

Sands of BIDMC is among the rare doctors who does this in his practice. He and the patient both look at the EHR during a visit, and he shows the note to the patient as he writes it. But for this arrangement to work, he says, the computer must be properly positioned and the doctor must know when to interact with it and when to talk with the patient while not typing in the computer.

In addition, Sands has started copying his notes for patients. He hasn’t had any problems with this, although it has altered how he documents visits in some ways. When a patient wants to add to or correct something in the note, he says, he inserts an addendum. (Legal considerations prevent changes to a note after it’s been signed, he says.) So far, few patients have asked to change anything.

Pearson has been sharing notes with his patients for 30 years. “I tell them to read the note, and if they see anything that I misheard or that they disagree with, to tell me so I can fix it or make an addendum,” he says. 

Whenever possible, he tries to finish the note in the exam room with the patient present so he can confirm the facts. Otherwise, patients can look at the note later and correct it if necessary, he says. While he’s usually accurate, sometimes patients will pick up errors.

Kagan also shares notes with patients, and sometimes they ask for corrections. In his experience, “most of the corrections have to do with social kinds of things.” For instance, patients might not want to be characterized as smokers or drinkers, or they’ll insist they don’t have hypertension, because they take antihypertensive drugs that keep their blood pressure down.

Most physicians already use some form of patient-generated data in their practices. But workload, liability and financial issues must be resolved before its use is widely expanded.

In addition, the lack of integration between EHRs and applications used to collect patient-generated data must be addressed before doctors can use more of this type of information in their everyday work. Particularly with functional status surveys, some doctors must be convinced that patients’ contributions to the record are as valid and as useful as clinical data.

Despite these obstacles, it’s likely that there will be an increased focus on patient-generated data as the industry moves to population health management, which requires continuous communication with patients. Notes Danny Sands, “With ACOs stressing value-based care, people are going to realize that they have to think outside the visit.”