BACKGROUND: Ensuring population health in a community requires surveillance and timely public health response to health threats. To meet this need, clinical and laboratory organizations are required to report selected conditions to public health authorities. To automate reporting, jurisdictional reporting requirements must be communicated in a manner understood by electronic health record (EHR) systems. The Reportable Condition Knowledge Management System (RCKMS) collaborative project provides this jurisdictional knowledge to enable public health reporting. Our objective is to: a) describe the knowledge development process and explain ‘trigger codes’ and reporting logic, b) summarize findings about reporting logic for nationally-notifiable conditions, and c) describe challenges and opportunities.
METHODS: Since Fall 2015, the RCKMS team (including epidemiologists, analysts, vocabularists, and a knowledge engineer) has developed default reporting specifications based on a) CSTE Position Statements for 74 nationally notifiable conditions, and b) input from jurisdictional and CDC experts during 31 content vetting work sessions. Optional logic was added when needed by a subset of jurisdictions. The RCKMS team created value sets of codes from standard terminologies to use when implementing logic and creating the ‘trigger codes’ for clinical systems to filter EHR data. Value sets are managed using the National Library of Medicine Value Set Authority Center. Currently, reporting logic for Chlamydia, Gonorrhea, Pertussis, Salmonella, and Zika (with other conditions to follow) is being implemented in the RCKMS clinical decision support (CDS) tool, and the initially-identified reporting specifications are being refined and validated with the public health community.
RESULTS: Default reporting logic has been defined for 69 nationally-notifiable conditions, with optional logic requested for 39% of the conditions. Value sets for diagnoses, laboratory-related content, and trigger codes are under development (n>300). While diagnosis-related logic should be straight-forward to implement, a review of the logic for 61 conditions identified criteria that may add complexity, including epidemiologic criteria (41% of the conditions), symptoms (39%), requests for preliminary lab results (30%), lab orders (20%), lab value comparisons (20%), and immunization criteria (5%). The expert vetting process has uncovered useful patterns for structuring knowledge useful for structuring the reporting specifications for conditions that are reportable in jurisdictions but not addressed by a CSTE Position Statement (n= >100).
CONCLUSIONS: The knowledge development process used by the RCKMS team is successfully curating content to support electronic case reporting, but some reporting criteria will be a challenge to implement.