Questions to Consider When Collecting Patient-Generated Data

Prior to the start of the annual meeting, an afternoon of presentations focused on “Applications of Mobile Health Technologies.”

Eric M. Ruderman, MD, of the Feinberg School of Medicine at Northwestern University, encouraged participants to consider many essential questions during his talk, “Efficient Collection of Patient-Generated Data.”

The overarching question, he said, is to decide which data to collect and for what goal, such as research, clinical care, or target a particular issue. Focusing on patient-generated data in the clinical settings, Dr. Ruderman said practitioners must make sure that data patient-reported outcome measures are:

• feasible (able to be collected and used effectively)
• meaningful
• minimal (presenting a low collection burden)

Furthermore, Dr. Ruderman said, practitioners should consider how much data they want to collect, taking into account the complexity of tools, how precise the tools are and whether precision will actually be used in clinical care, and patients’ willingness to complete instruments, including “questionnaire fatigue.” Ultimately, for data to contribute to clinical care, the information must be collected in real time and must be concise, meaningful, and actionable, “otherwise you’re spinning your wheels,” he said.

Once these larger questions are addressed, many important considerations remain. For example, Dr. Ruderman specifically reviewed three times when patients can provide data.

When collecting data during patient visits, practitioners must weigh when to collect data, whether to collect it from every patient at every visit, how data collection with affect workflow, and who will be responsible for data collection and entry. “How much time will data collection realistically take?” Dr. Ruderman asked, and when will it be done—during the appointment when the patient and physician are face to face, or on a tablet when the patient is in a waiting room?

Alternatively, when practices ask patients to provide such information before a visit, logistical questions arise, such as whether to use paper, web-based systems, or apps. One of the more difficult aspects of pre-visit data collection is deciding who ensures that it occurs, sends reminders to patients, and delivers the data to the physician.

Some practices are starting to collect between-visit data, using apps or web-based solutions. Again, Dr. Ruderman asked participants to consider who will be responsible for entering the data, tracking changes longitudinally, and making sure the data actually adds to the next visit.

Finally, practitioners must think about how to integrate patient-generated data with objective data such as joint counts, physical exams, and laboratory results. Ideally, he said, patient-generated data, objective data, visit notes, and changes over time should all be on the same system.

Although there are many questions to consider and challenges to address—including patient acceptance, patient capabilities, clinical capacity, and infrastructure—Dr. Ruderman said patient-generated data has many potential benefits, including better outcomes, patient engagement, and improved patient understanding of the importance of adherence.

The next steps in the efficient collection of patient-generated data will be demonstrating value; integrating tools into the electronic health record; and crossing practice domains among care, research, and registries.