Public Health Research - Information Affinity Domain (PHIAD) - overview
Public Health Information Affinity Domain
IBM delivered a successful Nationwide Health Information Network (NHIN) prototype to the U.S. Department of Health and Human Services (HHS), Office of the National Coordinator for Health Information Technology (ONCIT) relying heavily on HL7 standards and Integrating the Healthcare Enterprise (IHE) interoperability specifications.
Now referred to as a Healthcare Information Exchange (HIE), this infrastructure transformation could also be used to improve the often poor bi-directional communication with public health agencies and our ability to build better tools for early disease detection and prevention. IBM Research partnered with public health agencies to enhance reporting of their notifiable food borne illnesses within and across the partner governments. Our Public Health Information Affinity Domain (PHIAD) research prototype provides a web-based end user application and central IHE XDS repositories. This prototype provides transformation of data to the HL7 CDA R2 IHE XD-Lab document, policy controls for sharing documents with the CMC, and new tools for document-based analysis, visualization, and reporting (AVR). We worked closely with the IHE Lab domain to incorporate new public health data requirements into the existing IHE XD-Lab profile specification which is now in Final Text.
Population health is an important universal government function. Population health covers many types of programs, each collecting large amounts of population data to identify trends in their particular field. We focus here on infectious disease surveillance and outbreak detection, particularly with respect to food borne illnesses. Food safety involves regular testing at food production sites as well as detecting and investigating outbreaks in the human population that result from a single source contaminant, i.e. an outbreak. Governments receive clinical laboratory results when patients test positive for organisms such as Salmonella, Shigella, or Escherichia coli. Public health laboratories receive these isolates and complete epidemiological testing for serotyping and potentially pulse field gel electrophoresis analysis to confirm single source contamination. The governments ability to quickly aggregate and analyze incoming data is often stymied by inconsistent formats of data and the expensive resources needed to preprocess data prior to being able to analyze, visualize, and report on the current situation. The population is better served by a robust standardized reporting technology that simplifies data collection and automates detection processing so that epidemiologists can devote their time to investigation and prevention. Adopting Healthcare Information Exchanges (HIE) with public health goals in mind is a potential entry point for building a nationwide infrastructure for traditional clinical Healthcare Information Exchanges.
The HL7 Clinical Document Architecture provides several benefits for resource poor governments seeking solutions to improving their population health programs. First, documents closely mirror traditional paper-based reporting workflows still commonly in use. Second, simple style sheets provide for much improved report appearance and documents are easily printed. Third, documents are not locked into a particular system or proprietary vendor solution. Future system implementations can import standardized content from the machine readable document section. Fourth, simple sharing policies can implement nullFlavor masking enabling politically feasible data sharing.
Since public health data gathering is hierarchical in nature, PHIAD was developed as a hierarchical solution. Data is automatically propagated from one layer to another by using the same communication mechanism and the same technology (IHE-XDS). Sharing and privacy policies define who has access to the data, which data is to be shared, and when to share the data. The MECIDS participants share limited case data within their collaboration to identify and investigate cross-border outbreaks. Collaborative system build ensures consistent LOINC and SNOMED-CT coding of data such as specimen type, test code, and organism identification. CDA nullFlavor masking is a critical piece to enabling document sharing politically.
PHIAD leverages three open source technologies. The PHIAD web application is built atop the open source software for XDS provided by the Open Health Tools (OHT) project. OHT provides plugin implementation to support the client side of IHE IT Infrastructure Technical Framework Cross Enterprise Document Sharing (XDS) Profile. Second, PHIAD uses the Eclipse Business Intelligent Reporting Tool (BIRT) to produce table, bar chart, and pie chart reports. Additionally, PHIAD uses the Eclipse Spatio-temporal Epidemiological Modeler STEM tool to help epidemiologists and public health officials create and use spatial and temporal models of emerging infectious diseases. These models could aid in understanding, and potentially preventing, the spread of such diseases. Within each national PHIAD system, reports of Salmonella and Shigella notifiable conditions are a combination of clinical findings and epidemiological findings. Clinical findings include patient information, early organism identification, and often susceptibility interpretations. The isolates are then forwarded to a public health laboratory for identification confirmation, serotyping, susceptibility interpretations, and potentially genetic analysis to compare the strains for a genetic match in the case of an outbreak.