Revisiting Non-specific Syndromic Surveillance

Infectious disease surveillance is of great importance for the prevention major outbreaks. Syndromic aims at developing algorithms which can detect outbreaks as early possible by monitoring data sources allow to capture occurrences a certain disease. Recent research mainly focuses on specific, known diseases, putting focus definition pattern under surveillance. Until now, only little effort has been devoted what we call non-specific syndromic surveillance, i.e., use all available detecting any kind outbreaks, including infectious diseases are unknown beforehand. In this work, revisit published approaches and present set simple statistical modeling techniques serve benchmarks more elaborate machine learning approaches. Our experimental comparison established synthetic real in injected shows that these already achieve very competitive results often outperform algorithms.