Update from the Phenotypes, Data Standards, and Data Quality Core of the NIH Health Care Systems Research Collaboratory
Rachel Richesson, PhD, Associate Professor of Informatics, Duke University School of Nursing
Clinical data; Computable phenotypes; Electronic health records; Assessing data quality
The goals of the Phenotypes, Data Standards & Data Quality Core of the NIH Collaboratory are to share experiences using electronic health records (EHRs) to support clinical research; identify generalizable approaches, best practices, and tools; and explore and advocate for cultural and policy changes related to using EHRs to identify populations for research.
Challenges include lack of standardized data representation in EHRs, reproducibility, changes in coding systems, no standard representation for phenotype definitions, and issues around data quality related to errors within EHR systems.
In a learning healthcare system, research informs practice and practice informs research. Ideally research and clinical definitions should be semantically equivalent; they should identify equivalent populations.
When creating the phenotype definitions and standard codes for them, how do we reconcile differences between different EHR systems?
We need adequate data and methods to detect likely and genuine variation between populations at different trial sites and/or intervention groups.
It is recommended to do a formal assessment of accuracy, completeness, and consistency for key data elements.
For More Information
Sharing and reuse of computable phenotypes is discussed in Richesson et al. (2016) in eGEMs (open access): http://bit.ly/2bo52ST.
#LearningHealthcareSystem, #ComputablePhenotypes, #ElectronicHealthRecord, #pctGR
@PCTGrandRounds, @Collaboratory1, @PCORnetwork