Preventing and controlling outbreaks and AMR: From counting cases to monitoring risks

Abstract

Disease control measures during large infectious disease outbreaks rely mainly on counting disease cases, which can result in under- or delayed reporting. I explored surveillance complements and alternative surveillance data addressing limitations of facility-based disease surveillance, to more effectively inform measures.

Control measures require timely and accurate identification of disease cases in order to be effective. Controlling the 2014-2016 Ebola outbreak in West Africa largely relied on the isolation of confirmed cases. Confirming an Ebola infection could take several days, creating barriers to testing and isolation. A single on-site test combined with clinical and epidemiological data could speed up admission, tailored care, and remove test/isolation barriers. Similarly, during the 2016 yellow fever outbreak in DR Congo, case definitions using recent epidemiological data could speed up detection, which at the time required jaundice, appearing late in disease progression, delaying patient management, isolation, and vector control.

When outbreak control measures are introduced only when a threshold in the number of cases reported in disease surveillance is exceeded, measures are inherently delayed – depending on the disease, that is either problematic or acceptable. To respond timely to outbreaks of infectious diseases with strong growth potential - cholera in non-endemic areas, COVID-19 - the risk of an outbreak can be a better indicator to inform interventions. To identify areas and populations that should be prioritised for oral cholera in DR Congo, we analysed the historical geographical distribution of cholera cases and deaths, from which we deduced where outbreaks and highest case fatality can be expected during a future outbreak. Similarly, data tracking risk behaviour over time can timely inform control measures or estimate the impact of different interventions on epidemic growth. We could use the trend in the average number of reported risk contacts from COVID-19 contact tracing to model changes in daily COVID-19 incidence, and estimate what combination of interventions might be sufficient to keep virus circulation low.

For interventions to control antimicrobial resistance in low- and middle-income countries, where bacteria acquire resistance or resistant bacteria are transmitted at community level, classical hospital surveillance is particularly inadequate, delayed or non-existent. Unlike industrialised countries, access to hospitals and diagnostics is limited and the disease burden of antibiotic resistance is not mainly attributable to hospital-acquired infections. Monitoring modifiable risk factors as antibiotic use or hygienic conditions at community level better inform and evaluate interventions targeting these risks. As measuring outpatient antibiotic use is complicated by a significant proportion of antibiotic self-medication, over-the-counter medicine vendors or private clinics need to be included in population-level antibiotic use surveillance.

Linking risk surveillance, such as spread during past outbreaks, human behaviour, mobility, immunity or uptake of control measures, to facility-based disease surveillance can allow earlier and more targeted prevention/control measures to be deployed.