Investigating risks for human colonisation with extended spectrum beta-lactamase producing E. coli and K. pneumoniae in Malawian households: a one health longitudinal cohort study

Derek Cocker, Kondwani Chidziwisano, Madalitso Mphasa, Taonga Mwapasa, Joseph M. Lewis, Barry Rowlingson, Melodie Sammarro, Winnie Bakali, Chifundo Salifu, Allan Zuza, Mary Charles, Tamandani Mandula, Victor Maiden, Stevie Amos, Shevin T Jacob, Henry Kajumbula, Lawrence Mugisha, David Musoke, Rachael Byrne, Thomas EdwardsRebecca Lester, Nicola Elviss, Adam Roberts, Andrew C Singer, Christopher Jewell, Tracy Morse, Nicholas A Feasey

Research output: Working paperWorking Paper/Preprint


Background Low- and middle-income countries (LMICs) have high morbidity and mortality from drug-resistant infections, especially from enteric bacteria such as Escherichia coli. LMICs have varying infrastructure and services in the community to separate people from human and animal waste, creating risks for ESBL-Enterobacterales (ESBL-E) transmission. Limited data exist from Southern Africa on the prevalence of ESBL-E the community.

Methods and findings In this longitudinal cohort study we took a one-health approach to investigating prevalence and distribution of ESBL-E in urban, peri-urban and rural Malawian households between May 2018 and October 2020. We described human health, antibiotic usage (ABU), health seeking behaviour, structural and behavioural environmental health practices, and animal husbandry at these households. In parallel, human and animal stool and diverse environmental samples were collected and cultured to identify presence of ESBL E. coli and ESBL K. pneumoniae. Univariable and multivariable analysis was performed to determine associations with human ESBL-E colonisation.

We recruited 300 households, totalling 841 visits, and a paucity of environmental health infrastructure and materials for safe sanitation was noted across all sites. In total, 11,975 samples were cultured and ESBL-E were isolated from 41.8% (n=1190) of human stool and 29.8% (n=290) of animal stool samples. Animal species with particularly high rates of ESBL-E colonisation included pigs (56.8%, n=21) poultry (32.5%, n=148) and dogs (58.8% n= 30). ESBL-E were isolated from 66.2% (n=339) of river water samples and 46.0% (n=138) of drain samples. Urban areas had greater ESBL-E contamination of food, household surfaces, floors and the external environment, alongside the highest rates of ESBL-E colonisation in humans (47.1%, n=384) and animals (55.1%, n=65). Multivariable models illustrated that human ESBL E. coli colonisation was associated with the wet season (aOR = 1.66, 95%CrI: 1.38-2.00), living in urban areas (aOR = 2.01, 95%CrI: 1.26-3.24), advanced age (aOR = 1.14, 95%CrI: 1.05-1.24) and in households where animals were observed interacting with food (aOR = 1.62, 95%CrI: 1.17-2.28) or kept inside (aOR = 1.58, 95%CrI: 1.00-2.43). Human ESBL K. pneumoniae colonisation was also associated with the wet season (aOR = 2.23, 95%CrI: 1.63-2.76.

Conclusion We identified extremely high levels of ESBL-E colonisation in humans and animals and contamination of the environment in Southern Malawi. Urbanisation and season are key risks for ESBL-E colonisation, perhaps reflecting environmental contamination as toilets overflow in high population density areas in heavy rains in the wet season. Without adequate efforts to improve environmental health, ESBL transmission is likely to persist in this setting.
Original languageEnglish
Place of PublicationNew York
Number of pages33
Publication statusPublished - 21 Aug 2022


  • drug-resistant infections
  • Malawi
  • longitudinal cohort study


Dive into the research topics of 'Investigating risks for human colonisation with extended spectrum beta-lactamase producing E. coli and K. pneumoniae in Malawian households: a one health longitudinal cohort study'. Together they form a unique fingerprint.

Cite this