NLM logo

National Information Center on Health Services Research and Health Care Technology (NICHSR)

HSRProj (Health Services Research Projects in Progress)

Information about ongoing health services research and public health projects

Advancing privacy preserving record linkage methods in the context of real-world data networks and health information exchange (HIE)
Investigator (PI): Grannis, Shaun
Performing Organization (PO): (Current): Indiana University, School of Medicine, Department of Family Medicine / (317) 278-0300
Supporting Agency (SA): Patient-Centered Outcomes Research Institute (PCORI)
Initial Year: 2018
Final Year: 2022
Record Source/Award ID: PCORI/ME-2017C1-6425
Funding: Total Award Amount: $1,027,578
Award Type: Contract
Award Information: PCORI: More information and project results (when completed)
Abstract: The problem: Health care data are fragmented. Each time a patient visits a hospital, health system, clinic, pharmacy, long-term care facility, or public health agency, new information is created. This new information can help health care professionals better diagnose and treat their patients. This new information can also help scientists make new patient-centered outcomes research/comparative effectiveness research (PCOR/CER) discoveries. Unfortunately, this information is stored in many data repositories without a single unique identifier. Without a single unique identifier, clinicians and scientists cannot easily create a connected, complete record of each patient's information. Without a complete record of patient information, physicians cannot fully understand their patients, patient safety risks increase, public health reporting is weakened, and the ability to use patient information for PCOR/CER is severely harmed. However, joining data to create a comprehensive record can expose personally identifiable information and may place patient privacy in jeopardy. Therefore, we need techniques that accurately and securely link data to deliver safe and effective health care and to realize the nation's cost and quality improvement goals, while at the same time ensuring patient privacy is preserved. Outcomes we hope to achieve: Based on our significant prior innovations for advancing the state of the art in identified record linkage, we will 1) develop and apply successful identified linkage tools to the de-identified linkage problem, and we will also 2) build new tools for linking de-identified data. Our approaches also include techniques to "clean" the data before using de-identified linkage tools. We will then evaluate the ability for the tools and data cleaning techniques to improve matching accuracy, while protecting patient privacy. This work will contribute practical guidance and prototype technology enabling tools (algorithms and data preparations techniques) that practically improves the accuracy and protections afforded by privacy-preserving linkage performance. By achieving our proposed aims, we will advance the state of the art for privacy-preserving record linkage (PPRL) by making critical scientific contributions to the field, which will help research networks around the country more likely to be sustainable and enable better and lower cost research by lowering the barriers to data sharing by providing greater assurances for patient privacy. Also, this work will provide much needed evidence to prove "what works and what doesn?t" with respect to PPRL. How patients will be engaged: While our outcomes will be highly technical in nature, to have maximum effect we must communicate our aims and approach clearly with patient, clinician, researcher, information technology, and policy stakeholders. We have a patient representative as part of our research team and will convene multiple focus groups to hear their ideas and concerns. By implementing a strong patient engagement plan to obtain and respond to stakeholder input, we hope that patients and other PPRL stakeholders have a deeper shared understanding of the value of these highly technical, though increasingly necessary tools and techniques.
MeSH Terms:
  • Algorithms
  • Comparative Effectiveness Research
  • Focus Groups
  • * Health Information Exchange
  • Health Services Research
  • Humans
  • Medical Informatics /*methods
  • * Medical Records
  • Outcome Assessment, Health Care
  • Patient Safety
  • Patient-Centered Care
  • * Privacy
  • Program Development
  • Reproducibility of Results
Country: United States
State: Indiana
Zip Code: 46202
UI: 20184402
Project Status: Ongoing