A group of experts convened by the Duke-Margolis Center for Health Policy urged healthcare stakeholders and mobile health app companies to develop the mechanisms and standards needed to use mhealth data in medical research.
Mobile health data could generate "real-world evidence" for use in clinical trials, according to the Duke-Margolis report. "But there is not yet a clear path for how to efficiently elicit, validate, and transmit relevant data from mobile devices," the stakeholders write.
If such a path existed, the report notes, the data generated from mobile devices could greatly increase the understanding of health in the everyday lives of patients. "Today, most health data are collected in healthcare settings and fail to adequately capture critical information from patients about their daily lives," the experts point out.
The report mentions the following four areas of potential data generation:
Patient/consumer-reported data collected via mhealth apps
Task-based measures of a person's mental or physical abilities that could be captured via remote sensors and/or mobile apps that use sensors in a smartphone
Active sensor data that would require a person's active participation, such as stepping on a digital scale
Passive sensor data that do not interrupt a person's normal activities, such as activity level, heart rate, and sleep patterns
Although few mhealth data are currently being used in research, the report cites a Johns Hopkins study that uses iPhone and AppleWatch apps to enable patients to manage their epilepsy by tracking medications, seizures, and possible triggers or side effects. Early results from the nearly 600 participants are providing researchers with data on what was happening before a seizure struck.
The report suggests that researchers partner with mhealth app developers to validate the use of mhealth data, such as the daily activity level measured by a Fitbit activity tracker. Some studies have shown that mhealth data are often inaccurate. For example, two studies found that the Fitbit's measurements of heart rate were off by 20 to 30 beats/min. An "instant blood pressure checker" from another vendor was also shown to be grossly inaccurate.
The Duke-Margolis report makes five recommendations to bring the use of mhealth data in research closer to fruition.
First, the experts recommend the creation of "a learning mhealth research community" encompassing a wide range of stakeholders. This learning community would have four main tasks:
Develop methods to support consumer engagement with mhealth apps for research
Find ways to engage clinicians in this activity
Create methods and tools for using mhealth data in research
Address the variability in mhealth data quality and accuracy
Second, the report suggests that stakeholders help mhealth companies save development time and increase marketability with a research-capable app design. This would require the creation of an implementation guide to help mhealth firms comply with interoperability standards and regulatory requirements. However, the report also says the purpose of the guide is to encourage standardization through the use of application programming interfaces (APIs), software development kits, and open source applications.
Third, the report calls for "efficient access to well-characterized, standardized, and robust patient/consumer-generated health data." Noting that it is difficult to analyze data unless it is standardized, the experts suggest that vendors contribute "schemas and APIs" to frameworks and libraries such as Open mhealth and Xcertia.
Fourth, the report recommends that researchers use mhealth technologies "to communicate with study participants to provide meaningful and understandable feedback of study progress and research results." This might include communicating with participants via mobile apps to acknowledge that a person's data have been received, provide information on a study's progress, and explain study results.
The fifth recommendation is that researchers use mhealth apps to promote easier participation in research through awareness and adoption of standardized approaches for informed consent and patient privacy. Mhealth multimedia capabilities offer advantages in obtaining informed consent and could shorten the time required to recruit participants for clinical trials, the report says.
Among the other problems that must be addressed before mobile apps can be used in research, the report says, involve data linkages and interoperability across platforms with electronic health records, cyber security, patient/consumer consent for data usage and sharing, and the usability of health information technology.
Medscape Medical News © 2017
Cite this: Mobile Apps Could Be Used in Research Someday, Experts Say - Medscape - Sep 20, 2017.