PatientsLikeMe, FDA explore how patient-generated data could help event reporting

By Jonah Comstock
Share

Credit: FDA.gov

Online patient community PatientsLikeMe, which connects patients with different symptoms and conditions and collects patients' voluntarily submitted data for research, has teamed up with the FDA on a research study that could point the way toward a new channel for patient-generated health data for the agency. The study was recently published in the Journal of Medical Internet Research.

Right now PatientsLikeMe collects data for research purposes, but codes it using the the Medical Dictionary for Regulatory Activities (MedDRA), the FDA's coding framework for Adverse Event Reporting (AER). That's done in order to make the database more useful and searchable for researchers.

In this study, however, PatientsLikeMe wanted to see if that coding was close enough to make the data actually useful to the FDA.

For some time the FDA has been exploring different kinds of real-world evidence and other feedback mechanisms that keep pace with the growing paradigm of connected health. The agency has also been developing an Office of Patient Affairs that would help it listen more effectively to the voices of patients.

In the study, which included a dataset of 3,234 codes, researchers found that the FDA reveiwer agreed with the coding 97 percent of the time. In the remaining 3 percent, the disagreement was largely owing to PatientsLikeMe assigning general codes when more specific ones were available, something the company did in order to make the site more searchable for patients. The reviewer compared the codes to the MedDRA Term Selection: Points to Consider (MTS:PTC) guidelines.

"This review demonstrates that PGHD consisting of signs, symptoms, and ADE [adverse drug events] data entered by patients in curated structured fields can be reliably coded to the MedDRA terminology and that the coding of these data by PLM is generally aligned with MTS:PTC principles," researchers wrote. "Understanding the coding purpose and approach is informative for the optimization of data retrieval strategy. These findings suggest that efficient electronic searching and aggregation of PGHD might be possible when consistent, systematic curation processes are applied to PGHD as they are reported by patients. This standardization makes PGHD more electronically accessible and therefore elevates the visibility and importance of events patients find most significant."