Key points are not available for this paper at this time.
Electronic Health Records (EHRs) offer a rich data source for clinical researchers to assess a wide variety of treatments and outcomes. Researchers can use Honest Data Brokers (HDBs) to gain access to EHR data for research. Unfortunately, HDBs’ data analysts can inadvertently overlook nuances in clinical workflows when generating request code for data capture and cause data extraction errors that negatively impact EHR data quality. This paper presents findings from interviews with clinical researchers who have had experience with requesting EHR data from the Regenstrief Institute Data Core (RDC), an HDB with data analysts who provide authorised access to EHR data of nearly 25 million patients in the state of Indiana. Our participants wanted greater transparency when they had questions about the quality of the datasets they received. Participants wanted their data analysts to double check the data in their system, explain how they extracted the data, let them visit the data in their system, and/or view/edit the code used for data extraction. We offer a set of recommendations for HDBs on how to provide greater transparency to clinical researchers about the processes used to generate the EHR data they receive and discuss future directions for research.
Donaldson et al. (Sun,) studied this question.