Building a Flexible Surveillance Tool: Supporting Epidemiology Through Informatics

BACKGROUND:  

Public health electronic case reporting systems (especially those supported by state health departments) are typically focused on notifiable diseases limiting system utility for broader disease surveillance, case investigation, and outbreak response. Many outbreaks are caused by other pathogens and are outside the notifiable disease spectrum.  Local jurisdictions often respond to outbreaks whose etiology is neither identified nor reportable.  Tracking non-notifiable disease outbreaks in electronic notifiable disease registries is therefore not an option.  Without more flexible reporting systems, non-notifiable pathogen outbreak investigations have often been crude, lack efficiency and miss the benefits of automation.

METHODS:  

Using iterative requirements gathering, a team used epidemiology and informatics principles to design a flexible business intelligence (BI) solution for non-notifiable disease outbreak investigations.  Informed by epidemiology staff, design requirements were gathered for dashboard visualization.  Picturing the output was a significant design phase. An outbreak tool was designed to gather outbreak investigation data from case interviews, site visits, and environmental investigations.  Current and archived outbreak data were to be accessible to the end user for aggregation and visualization.  The surveillance tool was designed as a multifunctional outbreak investigation dashboard and an interactive customizable data entry tool.

RESULTS:  

In July 2012, the surveillance tool was pilot tested and implemented. The tool was used to document and track a foodborne illness outbreak at a local homeless shelter, where over 60 people became ill in July 2012. The tool properly captured case and control interviews, calculated attack rates and risk ratios for various food exposures, and documented symptomatic profiles for the outbreak.  Prior to implementing the tool, tracking interviews from an outbreak of unconfirmed illness would have been completed through paper forms and risk ratios calculated manually by staff.  Surveillance staff used the dashboards to plot an epidemic curve, provide statistical support for the putative outbreak agent and archived the investigation for future internal reference.

CONCLUSIONS:  

Innovative electronic tools for case investigation and outbreak response can drastically improve epidemiologic and surveillance capacity. A real-time combined data entry and reporting system for foodborne disease (and other) outbreaks can improve case identification and outbreak detection.  Application of informatics principles through BI to other public health domains has the capacity to improve efficiency and timeliness of public health practice.