Evaluation of the Health Level 7 Consolidated Clinical Document Architecture for Notifiable Condition Reporting

BACKGROUND:  Leveraging electronic health record data could greatly improve the completeness and accuracy of notifiable condition reporting. Meaningful Use Stage 2 mandates use of the Health Level 7 (HL7) Consolidated Clinical Document Architecture (C-CDA) for several objectives, including exchange of clinical information. Despite expressed interest in using the C-CDA for notifiable condition reporting, its ability to satisfy reporting requirements has not been fully tested. We evaluated the C-CDA specification for its ability to support notifiable infectious disease case reporting to the New York City Department of Health and Mental Hygiene (DOHMH).

METHODS:  We reviewed data elements collected through Reporting Central, a custom web-based portal for provider reporting to DOHMH, for four key surveillance programs (Tuberculosis Prevention, Immunization, STD Prevention and Control, Communicable Diseases). To assist with prioritization and classification, data from 56,775 reports received between 9/1/12 and 8/31/13 for 53 selected conditions were analyzed for completeness. Surveillance staff then prioritized and classified all data elements as ‘required’, ‘highly preferred’ or ‘nice to have.’ We then attempted to map required and highly preferred data elements to the C-CDA, noting whether the C-CDA specified an XPath location for each data element. To increase generalizability, we also mapped CSTE Common Core Data Elements to the C-CDA.

RESULTS:  Reviewing data completeness for provider reports received from Reporting Central was helpful in prioritizing the data elements. Most fields classified as required (e.g., patient and facility names, patient date of birth) could be mapped from the C-CDA. However, some required fields and many highly preferred fields (e.g., specimen source, symptom onset date) were not clearly possible to map from the C-CDA.  Although most CSTE Common Core Data Elements could be mapped, some (e.g., employer address) could not be mapped; these were often data elements not collected from providers by DOHMH.

CONCLUSIONS:  This study found that the C-CDA can meet many case reporting needs for key disease surveillance programs, but identifies important gaps between data elements required or highly preferred for disease reporting and those currently available from the C-CDA. Public health should either identify proxies for data elements needed for reporting, or recommend adding them to the C-CDA specification. Future research will attempt to determine whether data required for public health surveillance are actually present in a sampling of real C-CDA documents for patients with known reportable conditions exported from healthcare facilities.