Characterizing the underlying distributions in produce collected from the United States side of the United States - Mexico Border Public

Abstract

Despite the many health benefits of eating fresh fruits and vegetables, there is a risk of foodborne illness. Fresh produce, since it is consumed raw, never receives a kill step to rid of harmful pathogens. Being able to predict the risk of illness associated with fresh produce is important to prevention. However, before any inferences can be made, the underlying statistical distribution of pathogens, and their associated indicators, needs to be understood in order to make accurate risk predictions. This study assessed the fit of 5 commonly used distributions (normal, lognormal, Poisson, gamma, negative binomial) among 4 indicators (aerobic plate count, coliforms, Escherichia coli, Enterococci spp.), sampled from cabbage (n= 109), cantaloupe (n= 42) and cilantro (n= 141), which were collected on the U.S. side of the United States - Mexico border. Distributions were assessed by comparing the Pearson's chi-square values, along with the Akaike's information criterion, to determine which distributions fit each of the indicators. If more than one distribution fit an indicator, the best fitting distribution was determined. Of the 12 different sets of indicator-produce combinations, 10 were found to fit at least one of the assessed distributions. The lognormal fit all 10 of these indicator-produce combinations, while the gamma and negative binomial also fit 6 of the 10 indicator-produce combinations. The normal and Poisson did not fit any of the indicator-produce combinations. For the indicators in which more than one distribution fit, the lognormal was consistently found to have the best fit. This study emphasizes the value in assessing different distributions before making any risk predictions.

Table of Contents

LITERATURE REVIEW

Introduction. 1

Sources of Contamination. 1

Sampling Methods. 4

Testing procedures, post sampling: 5

Indicator Organisms. 7

Contamination Distributions. 9

Needs. 13

Goals and Aims. 14

Significance. 14

INTRODUCTION.. 17

METHODS. 20

Sample Collection. 20

Sample Processing. 20

Statistical Analysis. 21

RESULTS. 24

Goal 24

Histograms. 24

Distribution Fitting. 26

DISCUSSION.. 29

The lognormal distribution. 29

The gamma and negative binomial distributions. 30

The lognormal distribution was consistently the best of the 5 distributions. 31

Strengths and Limitations. 32

Implications. 32

Conclusions. 33

References. 34

Tables. 37

Figures. 41

APPENDIX A: IRB CLEARANCE 44

About this Master's Thesis

Rights statement

• Permission granted by the author to include this thesis or dissertation in this repository. All rights reserved by the author. Please contact the author for information regarding the reproduction and use of this thesis or dissertation.

School	Rollins School of Public Health
Department	Epidemiology
Degree	MSPH
Submission	Master's Thesis
Language	English
Research Field	Health Sciences, Public Health
Mot-clé	fresh produce
Committee Chair / Thesis Advisor	Leon, Juan, Emory University
Partnering Agencies	Does not apply (no collaborating organization)

Dernière modification

Primary PDF

Thumbnail	Title	Date Uploaded	Actions
	Characterizing the underlying distributions in produce collected from the United States side of the United States - Mexico Border ()	2018-08-28 13:45:00 -0400	Press to Select an action Download

Supplemental Files

Thumbnail	Title	Date Uploaded	Actions