The University of North Carolina at Chapel Hill is hiring a Postdoctoral Research Associate to develop methods for defining rare disease cohort controls and assessing and reducing re-identification risk in rare disease data. The role includes designing and evaluating control populations and disclosure controls, measuring their effects on statistical utility, and engaging patient advocacy groups, data stewards, and IRBs.
The Postdoctoral Research Associate will investigate methods for defining rare disease cohort controls and assessing and reducing re-identification risk in rare disease data, including EHR , claims, registry, and linked real-world data
Specifically, this will include developing a framework for organizing privacy risk factors unique to rare conditions, estimating sample and population uniqueness using multiple data resources and reference populations, and modeling added risk from public sources such as advocacy registries, case reports, and social media
The Postdoctoral Research Associate will design and evaluate the creation of control populations for small, rare disease populations where traditional propensity methods may not be appropriate
They will also design disclosure controls, including ontology-based aggregation, phenotype coarsening, suppression, and access-based approaches such as enclaves and federated analysis, and measure their effects on statistical utility
The work will involve engagement with patient advocacy groups, data stewards, and IRBs, and requires human subjects research training
The goals of the work will be to develop practical guidance and methods that protect rare disease patients while preserving the scientific value of their data for research
Qualification
Required
Ph.D. in biostatistics, statistics, epidemiology, biomedical informatics, computer science, or a related quantitative field
Experience with statistical disclosure control, privacy-preserving data methods, or re-identification risk assessment, shown through coursework, research, or publications
Experience in R or Python for statistical modeling, simulation, and analysis of large, complex health datasets
Experience working with sensitive health data such as EHR , claims, or registry data
completion of (or willingness to complete) human subjects research training
Ability to work with a geographically distributed team
Preferred
Research experience in rare diseases, or familiarity with the challenges of small-population and low-count data
Knowledge of biomedical ontologies and coding systems such as Mondo, HPO , ICD , SNOMED CT, or OMOP
Familiarity with de-identification standards and regulations, including HIPAA Safe Harbor and Expert Determination, and with disclosure control tools such as sdcMicro or ARX
Experience with secure data environments, including data enclaves, federated analysis, or cloud-based research platforms
Experience engaging patient advocacy groups, IRBs, or community stakeholders in research design or data governance
A strong publication record
excellent written communication skills, including the ability to turn technical findings into practical guidance for non-technical audiences
Benefits
Comprehensive medical and vision coverage
Paid leave
Benefits and services that support professional development and a healthy work/life balance
The Carolina Population Center (CPC) is a community of outstanding scholars and professionals associated to create new knowledge about population size, structure, and processes of change.
