Five overlooked steps to identify new epidemics, fast
LONDON — A Nature analysis argues that current epidemic surveillance systems are ill-equipped to detect emerging pathogens when they present with unusual or atypical symptoms, a gap that has repeatedly delayed public-health responses from COVID-19 to mpox.
The article identifies that outbreak tracking relies on three approaches — event-based monitoring of local information sources, genomic sequencing to confirm pathogen identity, and syndromic surveillance looking for characteristic symptoms — but these strategies do not account for pathogens that evade standard recognition.
Symptom presentation can change as a pathogen evolves and varies between individuals with different immunity levels. Co-infections with similar symptoms further obscure diagnoses. During the early 2020 wave of SARS-CoV-2, initial diagnoses relied on fever and respiratory distress; it took months for loss of smell to be accepted as a COVID-19 indicator, allowing individuals to bypass isolation protocols. The Omicron variant was associated with cold- and influenza-like symptoms more often than loss of smell or taste.
In the 2022 mpox outbreak, some individuals presented with genital and anal sores rather than the widespread rash associated with classic mpox. These people were initially misdiagnosed with sexually transmitted infections, potentially delaying recognition and public-health action. Similarly, Nipah virus has proven hard to diagnose in clinics in Bangladesh and India because its symptoms can shift between brain-tissue inflammation causing confusion and seizures and breathing difficulties.
Unless clinicians recognize that unusual symptoms warrant investigation, even the most advanced tracking systems cannot keep up. The analysis proposes 'symptom preparedness' as a loop in which clinical signals are detected, validated, translated into surveillance and diagnostic guidance, and rapidly fed back to front-line clinicians.
The first step is flagging clusters of unusual cases. Formal mechanisms are needed for clinicians to report intuition that they are dealing with something unusual, such as several people presenting with the same unrecognized set of symptoms. National public-health agencies and health ministries should provide reporting procedures through existing digital surveillance tools such as the District Health Information Software 2 (DHIS2). To prevent false-positive alerts, a second senior clinician or hospital epidemiologist must verify that the cluster is atypical. Verified reports would then trigger localized digital alerts to national public-health intelligence centres, where epidemiologists would cross-reference regional data before formal surveillance measures are activated.
The second step involves using AI to flag atypical cases. Machine-learning models trained on temporally and geographically labelled clinical data could help physicians detect cases whose symptoms differ from patterns normally observed in the same population. The models should not provide a diagnosis but could flag cases warranting further examination, augmenting rather than replacing clinical judgement. Implementation would require reliable internet connectivity, adequate data quality, testing and evaluation in the intended local setting, ongoing performance monitoring and privacy safeguards. In low-resource settings, health workers could submit anonymized clinical information to a country-wide centralized AI system through a phone app, SMS-based reporting system or existing electronic registries.
The third step is adopting tiered case definitions. Public-health authorities should transition away from using only standardized clinical criteria to classify disease cases. Instead, they should adopt dynamic, tiered case definitions that distinguish suspected, probable and confirmed cases. Committees of epidemiological specialists would stratify diagnostic criteria into tiers and update them dynamically — on a weekly or rolling basis — as real-time clinical data emerge. For example, if a dynamic diagnosis system had been available during the 2022 mpox outbreak, authorities could within days of recognizing atypical clusters have expanded the 'suspected case' definition to include localized genital sores, helping to identify misdiagnosed individuals sooner.
The framework requires investment in data infrastructure and clinical feedback loops to be effective. Without recognition of unusual symptoms, the article concludes, even the most sophisticated tracking systems will continue to miss emerging threats.
---
