Household and population size

There is some evidence that overcrowded living environments at the household and population level increase the prevalence of AMR and the abundance of antimicrobial resistance genes (ARGs). A study from Kenya showed that human carriage of ARGs from Escherichia coli in households that keep livestock was positively associated with increased household size. ^{Reference Muloi, Hassell and Wee33} A metagenomic study on raw sewage in Cameroon found that the abundance of clinically relevant ARGs was higher in a city of 3.5 million inhabitants than in a city of 1.0 million. ^{Reference Bougnom, McNally, Etoa and Piddock34} Specifically, 71 ARGs were common between the 2 cities, but an additional 61 ARGs were unique to the larger city, with only 4 ARGs unique to the smaller city. ^{Reference Bougnom, McNally, Etoa and Piddock34}

Pastoralist communities

Pastoralist communities may be at increased risk of AMR, given their close proximity to livestock and their use of prophylactic antimicrobials. A study in Nigeria showed that most (60%) pastoralists purchased antimicrobials from veterinary drug shops (as opposed to “animal drug hawkers”), and two-thirds of these pastoralists practiced arbitrary applications, rather than following the instructions given. ^{Reference Alhaji and Isola35} In addition, most pastoralists did not know of AMR (67%), and few (30%) thought that antimicrobial misuse in livestock could result in the emergence of resistant pathogens. ^{Reference Alhaji and Isola35} Low education, lack of funds for veterinary care, seasonal movement, and nonenforcement of antimicrobial distribution laws were all strong predictors of antimicrobial misuse. ^{Reference Alhaji and Isola35}

A study from Tanzania highlights the complexity of understanding patterns of antimicrobial use in livestock against the backdrop of variations in cultural and livelihood practices. ^{Reference Caudell, Quinlan and Subbiah36} Ethnic groups varied in how often they sought veterinary consultations (Chaaga = 96%, Arusha = 45%, Maasai = 36%), which was inversely related to the use of self-administered antimicrobials (Chaaga = 1%, Arusha = 21%, Maasai = 74%). Additionally, all pastoralists self-administering antimicrobials did not observe the recommended withdrawal period from consumption of milk and meat. Maasai, who purchased over-the-counter medicines from veterinary drug shops, were less likely to consult with a veterinarian for livestock care. Arusha, owning transportation or a cell phone or with higher levels of education, were more likely to self-administer antimicrobials. Maasai, who owned a radio or who had electricity, were more likely to observe posttreatment withdrawal from the consumption of milk and meat. ^{Reference Caudell, Quinlan and Subbiah36} In a follow-up study, ^{Reference Caudell, Mair and Subbiah37} AMR from human stool samples was highest for Maasai and Arusha households and lowest for Chagga households, with >40% of isolates resistant to ampicillin, tetracycline, trimethoprim, sulfamethoxazole, and streptomycin in Maasai and Arusha households, compared to 20%–30% resistance in Chagga households. ^{Reference Caudell, Mair and Subbiah37} However, antimicrobial use in humans and livestock was not associated with resistance to any antimicrobial. ^{Reference Caudell, Mair and Subbiah37}

Rural versus urban

Four studies explicitly assessed the impact of rural versus urban living on antibiotic use, AMR and ARG prevalence, and knowledge of AMR. The Tanzanian study noted above ^{Reference Caudell, Quinlan and Subbiah36} found a negative association between distance to urban centers and the likelihood of consulting a veterinarian for livestock care and purchase of antimicrobials in Arusha populations but not Chagga or Maasai populations. ^{Reference Caudell, Quinlan and Subbiah36} A Kenyan study of communities and hospitals in an urban (Nairobi) and rural (Siaya) setting showed that extended-spectrum cephalosporin-resistant Enterobacterales were more prevalent in the urban setting (52%) compared to rural setting (45%) and in urban hospitals (70%) compared to rural (63%) hospitals. ^{Reference Ita, Luvsansharav and Smith38} A large global burden of antibiotic resistance study including 1589 metagenomes from 26 countries showed that increased access to improved water and sanitation was associated with a lower abundance of ARGs; this association was stronger in urban compared to rural settings. ^{Reference Fuhrmeister, Harvey and Nadimpalli39} A Tanzanian study on antibiotic residues in foods reported that 42% of egg and Chinese cabbage consumers were unaware of the likelihood of antibiotic residues in their food, ^{Reference Mongi, Meshi and Ntwenya40} even though 73% knew that animal waste containing antibiotics was used for cultivating fruits and vegetables and 78% knew that antibiotic residues could be harmful to humans. Educational status and urban living were linked to awareness of antibiotic residues in foods. ^{Reference Mongi, Meshi and Ntwenya40}

Clinical-based settings

A study of patients in urban and rural hospitals in Kenya found a high prevalence of extended-spectrum cephalosporin-resistant Enterobacterales in nasal and fecal specimens and a very low prevalence of carbapenem-resistant Enterobacterales and methicillin-resistant Staphylococcus aureus, with a higher prevalence of resistant Enterobacterales in urban compared to rural hospitals. ^{Reference Ita, Luvsansharav and Smith38} A study of a teaching hospital in Kigali, Rwanda, revealed that gram-negative bacteria are highly resistant to antibiotics that are usually prescribed. Additionally, patients who had undergone surgery, taken antimicrobials in the past 30 days, or were transferred from another hospital had a greater chance of developing AMR. ^{Reference Sutherland, Mpirimbanyi and Nziyomaze41}

In Ghana, a study exploring antimicrobial prescribing behavior and the knowledge of teams treating dental patients found that antimicrobials were viewed as a fundamental aspect of dental care and were being prescribed both therapeutically and prophylactically, with amoxicillin, metronidazole, and amoxicillin/clavulanic acid being the most commonly prescribed. ^{Reference Sneddon, Thompson and Kpobi42} Some respondents shared concern about the cleanliness of facilities and the sterility of instruments used as a rationale for the extensive use of antimicrobials to prevent postoperative infections. They also wanted prescribers to follow the prescription advice of senior colleagues rather than advice from the national standard treatment guidelines. ^{Reference Sneddon, Thompson and Kpobi42} Most respondents had a good general knowledge of AMR, but this knowledge did not affect their decision-making around antimicrobial prescribing, which was influenced more by the perception of the patient’s ability to pay. ^{Reference Sneddon, Thompson and Kpobi42}

Water, sanitation, and hygiene (WASH)

The aforementioned metagenomic global burden of antibiotic resistance study ^{Reference Fuhrmeister, Harvey and Nadimpalli39} found that the abundance of ARGs was highest in Africa, followed by Southeast Asia and South and Central America. Increased access to improved water and sanitation was negatively associated with the abundance of ARGs for tetracycline and trimethoprim but not with the abundance of ARGs conferring resistance to beta-lactams, fluoroquinolones, aminoglycosides, or streptogramins. A study in Nairobi, Kenya, showed that children eating soil and the presence of communal handwashing stations within a block of households were associated with an increased load of antimicrobial-resistant E. coli. ^{Reference Omulo, Lofgren and Lockwood43} However, rainfall was the best predictor of a decreased load of antimicrobial-resistant E. coli, suggesting that transmission through unsanitary living conditions overwhelms the incremental effects of prior antibiotic use. ^{Reference Omulo, Lofgren and Lockwood43} A study of water quality in Guinea-Bissau (West Africa) documented that samples from the wet season had 67% more fecal coliforms than samples from the dry season and had a higher proportion of samples resistant to 3 or more antibiotics (wet = 22%; dry = 9%) but a lower percentage of isolates resistant to at least 1 antibiotic (wet = 59%; dry = 72%). ^{Reference Machado and Bordalo44}

Several studies in different parts of Africa examined wastewater samples. In South Africa, raw sewage water had higher levels of ARGs than treated wastewater discharge. ^{Reference Hamiwe, Kock, Magwira, Antiabong and Ehlers45} In Uganda, most (61%) Salmonella typhimurium and Salmonella enteridis isolates collected from wastewater treatment plants and slaughterhouses were susceptible to all antimicrobials tested. However, there were high levels of resistance to a variety of commonly used antibiotics, and shared Salmonella genotypes were found in human, livestock, and environmental sources, indicating that zoonotic and environmental transmission to humans is likely. ^{Reference Afema, Byarugaba, Shah, Atukwase, Nambi and Sischo46} Another study comparing the wastewater from hospitals in Benin and Burkina Faso revealed a higher relative abundance of ARGs in wastewater from hospitals in these countries compared to 6 hospitals in Finland. ^{Reference Markkanen, Haukka and Pärnänen47} In Zanzibar, plastic litter from 4 rural sites that had a history of cholera outbreaks, and poor sanitation facilities contained multiple multidrug-resistant enteric pathogens, including Klebsiella pneumonia, Enterobacter cloacae, Citrobacter freundii, and Vibrio cholerae. ^{Reference Rasool, Saavedra and Pamba48} In a Kenyan study, the availability of WASH resources like toilet handwashing stations, proper waste separation, clean toilet facilities, appropriate distance between hospital beds, and drinking water storage might explain the differences in patterns of AMR across hospitals. ^{Reference Maina, Tosas-Auguet and McKnight49}

Farming

There is ample evidence of AMR associated with livestock. The Wakiso District in Uganda has the highest livestock production in the county, and when environmental, human, and livestock samples were screened for extended-spectrum-beta-lactamase-producing E.coli, 80% of households had at least 1 positive sample. Seventy percent had animals carrying the drug-resistant bacteria, 62% had humans carrying the drug-resistant bacteria, and 10% had environmental samples carrying the drug-resistant bacteria. ^{Reference Muleme, Musoke and Balugaba50} The use of a protected water source for drinking, using containers with lids, and having a clean household were negatively associated with the presence of drug-resistant bacteria. ^{Reference Muleme, Musoke and Balugaba50} A survey in Tanzania revealed that poultry farmers consistently use oxytetracycline via drinking water for disease prevention, treatment, or growth promotion, and a majority of cabbage farmers use manure from livestock treated with antibiotics to enrich their soil. ^{Reference Mongi, Meshi and Ntwenya40} Similarly, a study conducted in a Zambian poultry farm revealed that 10% of the samples contained cefotaxime-resistant E. coli and that all of these samples were also resistant to at least 2 other antimicrobial classes. ^{Reference Shawa, Furuta and Paudel51} Further analysis showed that 20% of the poultry isolates were closely related to 25% of human inpatient isolates, sharing 2 plasmids that contained 14 distinct AMR genes, pointing to transmission between poultry and humans. ^{Reference Shawa, Furuta and Paudel51} Another study conducted on lettuce farms in Ghana found E. coli in all lettuce samples, regardless of the water source used for irrigation, with most (82%) of the samples resistant to more than 1 antimicrobial. ^{Reference Quarcoo, Boamah Adomako and Abrahamyan52} About one-fifth of raw meat samples (beef, sheep, goat, and chicken) from slaughterhouses in Ethiopia contained E. coli that were resistant to commonly used antibiotics, with nearly one-half showing multiple drug resistance. ^{Reference Messele, Abdi, Yalew, Tegegne, Emeru and Werid53}

Food safety

A survey of ready-to-eat foods in Accra, Ghana, showed that E. cloacae was present in 17% of tested food samples and showed resistance to commonly used antibiotics in the region. ^{Reference Dela, Egyir and Behene54} Similarly, a study of ready-to-eat shrimp in Nigeria showed that 93% of ready-to-eat shrimp contained Vibrio species, with the most common being Vibrio parahaemolyticus (38%), Vibrio vulnificus (12%), and Vibrio fluvialis (10%). ^{Reference Beshiru, Okareh, Okoh and Igbinosa55} All isolates were 100% sensitive to colistin and gentamycin but broadly resistant to other antibiotics commonly used in the region. ^{Reference Beshiru, Okareh, Okoh and Igbinosa55} Likewise, a study from a different part of Nigeria showed that 13% of barbecued beef samples from local markets contained E. coli. ^{Reference Datok, Ishaleku, Tsaku, Agya and Adoga56} All the E. coli strains were resistant to ampicillin, and 75% of the cultures that tested positive for E. coli showed resistance to multiple different antimicrobials. ^{Reference Datok, Ishaleku, Tsaku, Agya and Adoga56}