Draft framework for evaluating syndromic surveillance systems.

Interest in public health surveillance to detect outbreaks from terrorism is driving the exploration of nontraditional data sources and development of new performance priorities for surveillance systems. A draft framework for evaluating syndromic surveillance systems will help researchers and public health practitioners working on nontraditional surveillance to review their work in a systematic way and communicate their efforts. The framework will also guide public health practitioners in their efforts to compare and contrast aspects of syndromic surveillance systems and decide whether and how to develop and maintain such systems. In addition, a common framework will allow the identification and prioritization of research and evaluation needs. The evaluation framework is comprised of five components: a thorough description of the system (e.g., purpose, stakeholders, how the system works); system performance experience (e.g., usefulness, acceptability to stakeholders, generalizability to other settings, operating stability, costs); capacity for outbreak detection (e.g., flexibility to adapt to changing risks and data inputs, sensitivity to detect outbreaks, predictive value of system alarms for true outbreaks, timeliness of detection); assessment of data quality (e.g., representativeness of the population covered by the system, completeness of data capture, reliability of data captured over time); and conclusions and recommendations. The draft framework is intended to evolve into guidance to support public health practice for terrorism preparedness and outbreak detection.