Evaluation of sampling methods for the detection of pathogenic bacteria on pre-harvest leafy greens

A Validated Preharvest Sampling Simulation Shows that Sampling Plans with a Larger Number of Randomly Located Samples Perform Better than Typical Sampling Plans in Detecting Representative Point-Source and Widespread Hazards in Leafy Green Fields

Commercial leafy greens customers often require a negative preharvest pathogen test, typically by compositing 60 produce sample grabs of 150 to 375 g total mass from lots of various acreages. This study developed a preharvest sampling Monte Carlo simulation, validated it against literature and experimental trials, and used it to suggest improvements to sampling plans. The simulation was validated by outputting six simulated ranges of positive samples that contained the experimental number of positive samples (range, 2 to 139 positives) recovered from six field trials with point source, systematic, and sporadic contamination. We then evaluated the relative performance between simple random, stratified random, or systematic sampling in a 1-acre field to detect point sources of contamination present at 0.3% to 1.7% prevalence. Randomized sampling was optimal because of lower variability in probability of acceptance. Optimized sampling was applied to detect an industry-relevant point source [3 log(CFU/g) over 0.3% of the field] and widespread contamination [-1 to -4 log(CFU/g) over the whole field] by taking 60 to 1,200 sample grabs of 3 g. More samples increased the power of detecting point source contamination, as the median probability of acceptance decreased from 85% with 60 samples to 5% with 1,200 samples. Sampling plans with larger total composite sample mass increased power to detect low-level, widespread contamination, as the median probability of acceptance with -3 log(CFU/g) contamination decreased from 85% with a 150-g total mass to 30% with a 1,200-g total mass. Therefore, preharvest sampling power increases by taking more, smaller samples with randomization, up to the constraints of total grabs and mass feasible or required for a food safety objective. IMPORTANCE This study addresses a need for improved preharvest sampling plans for pathogen detection in leafy green fields by developing and validating a preharvest sampling simulation model, avoiding the expensive task of physical sampling in many fields. Validated preharvest sampling simulations were used to develop guidance for preharvest sampling protocols. Sampling simulations predicted that sampling plans with randomization are less variable in their power to detect low-prevalence point source contamination in a 1-acre field. Collecting larger total sample masses improved the power of sampling plans in detecting widespread contamination in 1-acre fields. Hence, the power of typical sampling plans that collect 150 to 375 g per composite sample can be improved by taking more, randomized smaller samples for larger total sample mass. The improved sampling plans are subject to feasibility constraints or to meet a particular food safety objective.

Evaluation of the Impact of Skewness, Clustering, and Probe Sampling Plan on Aflatoxin Detection in Corn

Probe sampling plans for aflatoxin in corn attempt to reliably estimate concentrations in bulk corn given complications like skewed contamination distribution and hotspots. To evaluate and improve sampling plans, three sampling strategies (simple random sampling, stratified random sampling, systematic sampling with U.S. GIPSA sampling schemes), three numbers of probes (5, 10, 100, the last a proxy for autosampling), four clustering levels (1, 10, 100, 1,000 kernels/cluster source), and six aflatoxin concentrations (5, 10, 20, 40, 80, 100 ppb) were assessed by Monte-Carlo simulation. Aflatoxin distribution was approximated by PERT and Gamma distributions of experimental aflatoxin data for uncontaminated and naturally contaminated single kernels. The model was validated against published data repeatedly sampling 18 grain lots contaminated with 5.8-680 ppb aflatoxin. All empirical acceptance probabilities fell within the range of simulated acceptance probabilities. Sensitivity analysis with partial rank correlation coefficients found acceptance probability more sensitive to aflatoxin concentration (-0.87) and clustering level (0.28) than number of probes (-0.09) and sampling strategy (0.04). Comparison of operating characteristic curves indicate all sampling strategies have similar average performance at the 20 ppb threshold (0.8-3.5% absolute marginal change), but systematic sampling has larger variability at clustering levels above 100. Taking extra probes improves detection (1.8% increase in absolute marginal change) when aflatoxin is spatially clustered at 1,000 kernels/cluster, but not when contaminated grains are homogenously distributed. Therefore, taking many small samples, for example, autosampling, may increase sampling plan reliability. The simulation is provided as an R Shiny web app for stakeholder use evaluating grain sampling plans.

In Silico Models for Design and Optimization of Science-Based Listeria Environmental Monitoring Programs in Fresh-Cut Produce Facilities

Food facilities need time- and cost-saving methods during the development and optimization of environmental monitoring for pathogens and their surrogates. Rapid virtual experimentation through in silico modeling can alleviate the need for extensive real-world, trial-and-error style program design. Two agent-based models of fresh-cut produce facilities were developed as a way to simulate the dynamics of Listeria in the built environment by modeling the different surfaces of equipment and employees in a facility as agents. Five sampling schemes at three time points were evaluated in silico on their ability to locate the presence of Listeria contamination in a facility with sample sites for each scheme (i.e., scenario, as modeled using scenario analysis) based on the following: the facilities' current environmental monitoring program (scenario 1), Food and Drug Administration recommendations (scenario 2), random selection (scenario 3), sites exclusively from zone 3 (i.e., sites in the production room but not directly adjacent to food contact surfaces) (scenario 4), or model prediction of elevated risk of contamination (scenario 5). Variation was observed between the scenarios on how well the Listeria prevalence of the virtually collected samples reflected the true prevalence of contaminated agents in the modeled operation. The zone 3 only (scenario 4) and model-based (scenario 5) sampling scenarios consistently overestimated true prevalence across time, suggesting that those scenarios could provide a more sensitive approach for determining if Listeria is present in the operation. The random sampling scenario (scenario 3) may be more useful for operations looking for a scheme that is most likely to reflect the true prevalence. Overall, the developed models allow for rapid virtual experimentation and evaluation of sampling schemes specific to unique fresh-cut produce facilities. IMPORTANCE Programs such as environmental monitoring are used to determine the state of a given food facility with regard to the presence of environmental pathogens, such as Listeria monocytogenes, that could potentially cross-contaminate food product. However, the design of environmental monitoring programs is complex, and there are infinite ways to conduct the sampling that is required for these programs. Experimentally evaluating sampling schemes in a food facility is time-consuming, costly, and nearly impossible. Therefore, the food industry needs science-based tools to aid in developing and refining sampling plans that reduce the risk of harboring contamination. Two agent-based models of two fresh-cut produce facilities reported here demonstrate a novel way to evaluate how different sampling schemes can be rapidly evaluated across multiple time points as a way to understand how sampling can be optimized in an effort to locate the presence of Listeria in a food facility.