Quantifying uncertainty in ESBL-producing Enterobacterales transmission from animals and the environment to humans in Singapore: a structured expert elicitation study
Yewei Xie, Nicholas Graves, Zhi Zhen Lim, Yin Mo, Isabelle Maya Srivastava, Yot Teerawattananon, Hwee Lin Wee, Alec MortonAbstract
Objectives
This study aimed to estimate daily transmission rates of extended-spectrum β-lactamase (ESBL)-producing Enterobacterales from hospital surfaces, companion animals and community environmental reservoirs to humans in Singapore and to quantify the uncertainty in these expert-derived estimates.
Methods
A structured expert elicitation was conducted with a panel of eight experts: three for the human sector, three for the environmental sector, and two for the animal sector. Expert judgements were aggregated using both equal weighting and Cooke’s classical model (performance weighting) for hospital and companion-animal pathways, but only equal weighting was used for the community-environment pathway.
Results
Under performance weighting, the daily transmission rate was 1.76 per person-day [90% credible interval (CrI) 0.10–8.61] for the hospital-surface pathway and 0.03 (90% CrI 0.00–0.11) for the companion-animal pathway. The corresponding equal-weighted estimates were 0.20 (90% CrI 0.01–2.77) and 2.12 (90% CrI 0.09–23.87), respectively. For the community-environment pathway (equal-weighted only), the rate was 0.81 (90% CrI 0.10–4.12).
Conclusions
The expert judgements suggest that environmental and hospital pathways may represent important but uncertain contributors to ESBL-producing Enterobacterales spread in urban contexts, whereas estimates for animal-associated transmission depend strongly on the weighting method. This study demonstrates that structured expert elicitation can provide systematic probabilistic priors for transmission and cost-effectiveness models, informing future resource allocation and the development of prevention and control strategies.