The Office of the National Coordinator for Health Information Technology (ONC) released a report in March 2021 summarizing perspectives of researchers on the use of application programming interfaces (APIs) for clinical research. Currently, extracting data required for research from the electronic health record (EHR) or data warehouse, or curating data to conform to a common data model (CDM) across multiple sites, are complex and resource-intensive activities. The ONC interviewed researchers to evaluate whether standardized APIs may reduce dependence on custom data extracts and complex data curation. While increased use of standardized APIs may be useful to streamline research efforts across organizations and support multi-site trials, there are several barriers to broader use of APIs in clinical research studies at the present time.
Standardized APIs are designed to gather structured data elements for clinical decision support, care coordination, and other use cases. The stakeholders interviewed by ONC reported that current standardized APIs only extract patient-level data from certified EHR systems and not from other health information technology (IT) products. Text-based information is not readily available. Also, EHR systems can be implemented differently across institutions using the same EHR, and users may enter data inconsistently even within the same organizations. Research studies often require unique data sets or study-specific data manipulation, and so reliance on standardized APIs may be inadequate for certain research purposes. Some researchers reported that in many cases, data outside the EHR is needed. In particular, research in precision medicine, genomics, and rare diseases frequently demand detailed, difficult-to-locate information.
The report described recent projects using standardized APIs in research. The Oregon Health & Science University is working on a planning tool for use with patients who have chronic kidney disease, cardiovascular disease, diabetes, or opioid use disorder, as part of a research study funded by the Agency for Health Research and Quality and other entities. The tool would extract data from the EHR to support clinical decisions for care planning for individuals with complex conditions. Vanderbilt University developed a data exchange platform known as REDCap, which offers a secure web application for building and managing online surveys and databases. Weill Cornell Medicine has developed a project using APIs to access REDCap within the Epic EHR and automatically populate forms with data from Epic. Other projects described include the Cumulus project at Boston Children’s Hospital, and a project at MedStar Health Research Institute to evaluate open source health IT tools.
The ONC report notes that the Cures Act Final Rule requires certified health IT developers to enable the Fast Healthcare Interoperable Resources (FHIR) Bulk Data Access API, which provides access to patient-level data across a patient population. However, the FHIR Bulk Data Access API was still in development stage at most organizations as of December 2020.
Finally, the report describes several concerns relating to privacy and security. Researchers felt that consumers are unaware of the extent to which their health data can be de-identified and reused without their consent, and that current mechanisms to segment unnecessary clinical data from data elements needed for research are inadequate.