Patient perspectives on the linkage of health data for research: Insights from an online patient community questionnaire

Int J Med Inform. 2019 Jul:127:9-17. doi: 10.1016/j.ijmedinf.2019.04.003. Epub 2019 Apr 6.

Abstract

Objective: To examine the patient perspective on the risks and benefits of linking existing data sources for research.

Materials and methods: Between December 2015 and February 2016, we fielded a questionnaire in PatientsLikeMe, an online patient community representing over 2500 health conditions. The questionnaire was developed using subject matter expertise and patient feedback from a concept elicitation phase (N = 57 patients). The final questionnaire consisted of 37 items.

Results: Of n = 5741 who opened the email invitation, n = 3516 respondents completed the questionnaire (61.2%). Of these, 73.8% were women, 86.4% were Caucasian, 14.5% were 65 or older, and 44.9% had completed college or post-graduate education. Questionnaire respondents indicated that the most important benefits of sharing data were "helping my doctor make better decisions about my health" (94%) and "helping make new therapies available faster" (94%). The most important data sharing risk identified was health data being "stolen by hackers" (87%). Of 693 patients who were not comfortable with researchers accessing their de-identified data, most reported that their comfort levels would increase if they were able to learn how their data was protected (84%). In general, responders felt more comfortable when unique identifiers such as social security number (90%) and insurance ID (82%) were removed from the data for linkage and research use.

Discussion: The majority of patients in a US-based online community are comfortable with researchers accessing their de-identified data for research purposes.

Conclusions: Developing methods to link databases minimizing the exposure of unique identifiers may improve patient comfort levels with linking data for research purposes.

Keywords: Data linkage; Ethics; Patient-centered research; Privacy.

Publication types

  • Research Support, Non-U.S. Gov't

MeSH terms

  • Adolescent
  • Adult
  • Aged
  • Databases, Factual
  • Feedback
  • Female
  • Health Records, Personal*
  • Humans
  • Information Dissemination
  • Male
  • Middle Aged
  • Surveys and Questionnaires*
  • Young Adult