Architecture for Public Health Decision Support

The United States government has recently made an unprecedented investment aimed at increasing the adoption and meaningful use of health information technologies. Although nationwide adoption of EHRs could provide an opportunity for better communication of public health guidelines and recommendations, significant barriers exist. First, there is no thorough understanding of how public health guidelines can be effectively incorporated into EHR-enabled clinical workflows. Prior work has found that "human-interpretable" public health recommendations are too non-specific to instantiate into computerized alerts. Second, there is no standardized format/medium for communicating this content in a human-readable and machine-interpretable format to be consumed by multiple EHR systems. Finally, there is currently no widely used, openly-available federated architecture for distribution of standardized public health messages. The end result is a limited ability by public health to effectively communicate its messages in an increasingly IT-enabled clinical environment.

We have developed a public health decision support architecture that will enable the creation and distribution of public health guidelines that will be vendor-neutral, extensible, rapidly updateable, acceptable to clinicians, and privacy protective. This architecture makes use of a standards-based format derived from the Continuity of Care Document to represent public health guidelines in an executable format. These executable guidelines called (popCCDs) represent the parameters for a case definition, thereby identifying a population of interest rather than an individual person. These guidelines will be interpreted by an open-source decision support engine that performs a simple “matching” evaluation between patient CCDs and the popCCD guideline. The "Plan of Care" section of the popCCD will be used to convey the recommended action plan associated with matching CCDs. We are currently working with the eClinicalWorks EHR vendor to implement the popCCD standard. We have also implemented the architecture independently as a web service with future plans of adding it to the Public Health Grid.