Looking in all the wrong places: A rationale for signal detection for pandemics based on existing data sources

Abstract

Global surveillance systems did not detect the early stages of the COVID-19 pandemic. We argue this is because the national surveillance systems which report to centralized systems are not designed to detect the emergence of novel infectious diseases. Likewise, substantial resources devoted to hunting for deadly new viruses in obscure places did not predict COVID-19. We suggest an alternative approach to make better use of baseline human mortality and morbidity data to detect anomalies, building on existing frameworks for data collection and standardization and drawing on data from individual medical facilities. While most emerging diseases in humans originate in animals, focusing on animal surveillance may be an ignis fatuus, and detection should focus on human cases as early as possible after spillover. Animal-based surveillance for pandemic prevention is warranted for recurring outbreaks of known zoonotic pathogens when it can inform the detection of human cases. Further research is suggested in surveillance for pandemic preparedness utilizing human baseline data, using available routine health data, as well as other data sources generated outside the health sector which could detect anomalies. The methodology is potentially highly cost-effective and applicable to low- and middle-income countries. Data sources can be evaluated with historical data, where evidence of detection should be seen in the early stages of within-country spread of COVID-19.