What are the strengths and limitations of data on PatientsLikeMe?

What are the strengths of PatientsLikeMe data?

Every data source carries positives and negative aspects that help to answer the research question at hand. PatientsLikeMe data is based on patients who choose to participate in our site and is derived primarily from self-reports. We are working to include data from other sources in the future, e.g., genomics, electronic medical records, wearables, and other data sources. As with all of our current data, patients can choose to contribute as much or as little data as they like.

PatientsLikeMe is a unique data source because:

  • Patient-reported data collected between clinical visits
  • Real-world patients with multiple conditions, facing real-world issues in work, relationships, finance, healthcare access, and daily life
  • Patient reports of treatments including off-label, supplemental, alternative, behavioral, and environmental modification, with evaluations specific to their condition(s)
  • Increasingly, additional data sources incorporated into profiles (e.g., activity monitor data)

What are the main limitations of PatientsLikeMe data?

  • Our users tend to be slightly more likely to be female, are a few years younger, a few percentage points more likely to be white, and a few percentage points more likely to have a higher level of education than the general population.
  • We have more data and more patient activity in some conditions (e.g., MS) than in others (e.g., psoriasis).
  • We have relatively few seniors (aged 75+) using the site, which may relate to familiarity with or access to technology.
  • Conditions in which patients rapidly lose cognitive function may also make interaction difficult given the reliance on usernames, passwords and complex medical information.
  • In a few conditions, such as fibromyalgia or IPF, our communities are so much larger than the few systematic studies that it is very hard to tell whether our biases are meaningful. We also work to minimize these biases and in larger communities can control for them by over-sampling those subsets of the community where we are under-represented and under-sampling those where we are over-represented.
Powered by Zendesk