Lessons learned about study design, sampling and testing techniques to describe fecal contamination routes on produce at the farm level

Although farms and packing sheds have been implicated in several outbreaks caused by contaminated produce, contamination routes are not well understood. To address this, we completed two epidemiological field studies in farms and packing sheds on the US-Mexico border. In May, we pilot tested a novel study design in Mexico, to build on our past work. In our past studies, we collected 923 samples of produce and enumerated fecal indicators, including Escherichia coli (E. coli), by culture based methods, with limits of detection of 10 CFU/g. E. coli were found in 6 of 14 produce types, including cantaloupe. Produce type, packing step, country of origin, and season were significantly associated with E. coli contamination. Samples from the harvesting bin were nearly seven times more likely to be contaminated than those sampled directly from the field (OR = 6.77, 95% CI: 2.87 – 15.96), suggesting contamination occurred during harvest. Compared to the bin, wash-tank samples were a third as likely to be contaminated (OR = 0.32, 95% CI: 0.12 – 0.81), suggesting that produce washing decreased fecal contamination. A key limitation of our past work is that fecal contamination is often too low or focal to be observed at the production level. We will describe how we addressed these challenges in our current study by focusing on high-risk crops (jalapeño peppers, cantaloupes, tomatoes), designing novel field sampling methods, and by using molecular epidemiology techniques like Bacteriodales 16S rDNA microbial source tracking and E. coli strain-typing to identify the source of contamination.