Antimicrobial-resistant Salmonella and Escherichia coli in fresh vegetables from Ouagadougou, Burkina Faso: Prevalence, β-lactamase genes, and food safety implications

Fresh vegetables consumed raw are a recognized but underappreciated vehicle for foodborne pathogens and antimicrobial-resistant (AMR) bacteria, yet data on AMR determinants in fresh produce remain scarce for West Africa. This cross-sectional study characterized the prevalence of Escherichia coli and Salmonella spp . in 300 fresh vegetable samples from farms, markets, restaurants, and supermarkets in Ouagadougou, Burkina Faso, along with their antimicrobial resistance profiles, selected β-lactamase genes, and associated food-handling risk factors. E. coli was detected in 63.0% of samples (95% CI: 57.4–68.3%) and Salmonella spp. in 10.0% (95% CI: 6.9–13.9%). Prevalence did not differ significantly by season for either species (E. coli: 61.3% dry season vs 64.7% rainy season, χ²(1) = 0.36, p = 0.55; Salmonella: 8.0% vs 12.0%, χ²(1) = 1.33, p = 0.25). Per-agent resistance rates were likewise broadly similar between seasons for both species (e.g., Salmonella tetracycline resistance 23.3% dry vs 26.7% rainy). Salmonella isolates showed higher resistance for several individual agents than E. coli isolates, most notably tetracycline (25.0% vs 15.3%) and trimethoprim-sulfamethoxazole (21.7% vs 12.8%; see Table 1). Multidrug resistance (non-susceptibility to ≥ 3 antimicrobial classes) was observed in 60% of resistant Salmonella isolates. Among eight resistant isolates screened by multiplex polymerase chain reaction (four E. coli and four Salmonella, selected to represent the range of resistance phenotypes observed), blaTEM was detected in 100% (8/8), blaOXA in 50.0% (4/8), and blaCTX-M in 50.0% (4/8), with co-occurrence of all three genes in 37.5% (3/8) of isolates. The detection of the extended-spectrum β-lactamase-associated blaCTX-M gene provides new surveillance data on vegetable-associated bacteria in Burkina Faso, an environmental compartment not previously included in national AMR monitoring. Risky food-handling practices were widespread: 98% of vendors stored vegetables at ambient temperature, 85% applied no disinfection, and 80% of settings presented cross-contamination risk with raw meat. A screening-level model yielded an indicative 2.0% exposure-risk estimate for antibiotic-resistant Salmonella per raw salad serving; this is a screening-level indicator rather than a confirmed infection probability (see Methods). A HACCP-based control framework with five critical control points is proposed as a preliminary, field-validation-pending tool. These findings support the integration of fresh produce into One Health antimicrobial resistance surveillance strategies in West Africa.