Text

theses

The causes of most fresh produce outbreaks in U.S. are unknown, but cross contamination during washing or improper storage temperatures during retail storage, distribution or home storage may play a role. The first objective of our research was to integrate and compare published data, published models and data from the ComBase database relevant to Salmonella growth in fresh and fresh-cut produce. The second objective of our research was to develop a cross contamination model that predicts the concentration of contaminated produce and the concentration of non- contaminated produce after washing using literature data. A literature research was conducted to find relevant data on the growth of Salmonella on fresh cut produce. Data for Salmonella growth in a variety of fruit and vegetable products was also extracted from ComBase. Calculated growth rates were converted to square-root growth rates for comparative purposes and analyzed. Four published Salmonella growth models (Koseki and Isobe on iceberg lettuce; Pan and Schaffner on cut tomatoes; Li et al on cut melons; and Sant'Ana et al on lettuce) were compared to the extracted data. The most conservative model (Koseki and Isobe, 2005) was fail-safe for all but 5.5% (6/109) of the extracted data, predicting faster growth that that actually observed. A literature research was conducted to find relevant published data on the cross contamination rates between contaminated produce after wash, wash water and non- contaminated produce after wash. Data were converted to the same units, log transformed, used to create histograms and figures using Microsoft Excel. GInaFit and BestFit software were used to select suitable distributions. The software program @RISK was used to build a risk model. The simulation model predicted that when tomatoes were contaminated at 4 log CFU/tomato, after washing at 100 ppm chlorine, those same tomatoes contained ~1.0 log CFU/tomato, while contaminated cantaloupes contain ~2.8 log CFU/cantaloupe after washing at 0 ppm chlorine. The simulation model also predicted that uncontaminated tomatoes after washing at 0 ppm chlorine with contaminated tomatoes will contain ~ -0.59 log CFU/tomato (or 1 in 4 tomatoes containing ≥ 1 CFU), while uncontaminated cantaloupes after washing at 100 ppm chlorine with contaminated cantaloupes will contain ~ -2.83 log CFU/cantaloupe (or 1 in 676 tomatoes containing ≥ 1 CFU).

ETD_5260

M.S.

Includes bibliographical references

by Fang-Yu Lin

doi:10.7282/T35X2712

ETD graduate

The author owns the copyright to this work.