Statistical Techniques to Analyze Pesticide Data Program Food

Jun 14, 2018 - Statistical Techniques to Analyze Pesticide Data Program Food. Residue Observations. Arpad Z. Szarka,* Carol G. Hayworth, Tharacad S...
0 downloads 0 Views 816KB Size
Subscriber access provided by - Access paid by the | UCSB Libraries

Food Safety and Toxicology

Statistical Techniques to Analyze Pesticide Data Program Food Residue Observations Arpad Z. Szarka, Carol G. Hayworth, Tharacad S. Ramanarayanan, and Robert S. I. Joseph J. Agric. Food Chem., Just Accepted Manuscript • DOI: 10.1021/acs.jafc.8b00863 • Publication Date (Web): 14 Jun 2018 Downloaded from http://pubs.acs.org on June 19, 2018

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

Page 1 of 31

1

Journal of Agricultural and Food Chemistry

Statistical Techniques to Analyze Pesticide Data Program Food Residue Observations

2

3

Arpad Z. Szarka*, Carol G. Hayworth, Tharacad S. Ramanarayanan, and Robert S. I. Joseph

4

Operator and Consumer Safety, Syngenta Crop Protection, LLC, Greensboro, NC, 27419.

5

(*) Corresponding author: A. Z. Szarka, e-mail: [email protected]

6 7

ABSTRACT

8

The U.S. EPA conducts dietary risk assessments to ensure that levels of pesticides on food are

9

safe in the U.S. food supply. Often these assessments utilize conservative residue estimates,

10

Maximum Residue Levels (MRLs), and a high-end estimate derived from the registrant-

11

generated field trial data sets. A more realistic estimate of consumers’ pesticide exposure from

12

food may be obtained by utilizing residues from food monitoring programs such as the Pesticide

13

Data Program (PDP) of the US Department of Agriculture. A substantial portion of food residue

14

concentrations in PDP monitoring programs is below the limits of detection (left-censored) which

15

makes the comparison of regulatory field trial and PDP residue levels difficult. In this paper we

16

present a novel adaption of established statistical techniques, Kaplan-Meier Estimator (K-M),

17

Robust Regression on Ordered Statistic (ROS) and Maximum Likelihood Estimator (MLE), to

18

quantify the pesticide residue concentrations in the presence of heavily censored data sets.

19

The examined statistical approaches include the most commonly used parametric and non-

20

parametric methods for handling left-censored data that have been used in the field of medical

21

and environmental sciences. This work presents a case study in which thiamethoxam bell

22

pepper residue data generated from the registrant field trials were compared with PDP

23

monitoring residue values.

24

compared with commonly used simple substitution methods for determination of summary

The results from the statistical techniques were evaluated and

ACS Paragon Plus Environment

Journal of Agricultural and Food Chemistry

25

statistics. It was found that the maximum likelihood estimator (MLE) is the most appropriate

26

statistical method to analyze this residue data set. Using the MLE technique, the data analyses

27

showed that the median and mean PDP bell pepper residue levels are approximately 19 times

28

and 7 times lower, respectively than the corresponding statistic of field trial residues.

29 pesticide food monitoring; censored data; Kaplan-Meier survival analysis,

30

KEYWORDS:

31

maximum likelihood estimator, regression on ordered statistic, dietary risk assessment

32

ACS Paragon Plus Environment

Page 2 of 31

Page 3 of 31

33

Journal of Agricultural and Food Chemistry

1. INTRODUCTION

34

The U.S. EPA conducts dietary risk assessments to ensure that the levels of pesticides

35

in food are safe in the U.S. food supply. Dietary risk assessment outcomes depend upon the

36

toxicity of pesticide, food intake, and the magnitude of pesticide residues in food, including the

37

raw agricultural commodities (RACs). The magnitude of residues in RACs is an estimate of the

38

pesticide concentration present in consumers’ food intake.

39

estimation of residue concentrations in RACs will afford a more accurate and reliable

40

assessment of dietary risk. In recognizing the importance of quantifying potential pesticide

41

concentrations in food, current regulations require registrants to generate pesticide field trial

42

residue data during the development of a pesticide in order to register the product with the EPA.

43

These data are also used to establish maximum legal limits i.e. Maximum Residue Levels

44

(MRLs) or tolerances for crops. The generated field trial data become part of the body of

45

knowledge associated with the use of the product. In addition to evaluation of dietary risk, a

46

wider variety of scientific tests are required by law to ensure the safe use of the pesticide,

47

including efficacy, product chemistry, potential human health, environmental effects, and the

48

impact on non-target organisms.

Consequently, improving the

49

To conserve resources, the U.S. EPA has implemented a tiered approach in dietary risk

50

assessment. 1 The lowest tier assessments utilize conservative residue estimates, MRLs, and a

51

high-end estimate derived from the registrant generated field trial data sets. At the next tier, the

52

MRLs are replaced by the average of the field residue values for chronic assessments and the

53

entire distribution of residues for the acute assessments. The field trials are conducted under

54

worst-case scenarios i.e. at the maximum application rates, maximum number of applications

55

and minimum pre-harvest intervals (PHIs), specified in the product label.

56

residues generated under these conditions represent conservative, worst-case estimates. A

57

more realistic estimate of consumers’ pesticide exposure from food may be obtained utilizing

ACS Paragon Plus Environment

Therefore, food

Journal of Agricultural and Food Chemistry

58

residues from food monitoring programs. Monitoring programs sample, test, and report on

59

pesticide residues in agricultural commodities. In the USA, monitoring data are available from

60

the USDA’s Pesticide Data Program (PDP) and registrant-supported market basket studies.

61

PDP is a national program that collects residue data on selected agricultural commodities in the

62

food supply based on a rigorous statistical design to ensure that sample collections provide

63

reliable estimates of pesticide residues in the U.S. food supply. The PDP data can provide

64

more realistic estimates of residues in RACs as opposed to the worst case field trial data. The

65

PDP data are reflective of the actual pesticide use and usage scenarios, including the percent of

66

crop treated (%CT). PDP samples are taken close to the point of consumption and the data

67

take into account the reduction of residues that may occur between pesticide treatment and

68

consumption; therefore, it provides a more realistic measure of the pesticide residues. While

69

food monitoring data have been frequently used as refinement in residue estimates, a

70

quantitative assessment of residue reduction has not been available for the PDP data.

71

In general, the PDP data are reflective of multiple residue reduction factors.

72

Consequently, the observed pesticide residues are significantly lower compared to registrant

73

generated field trial residues. In addition, the PDP data frequently include values reported as

74

“less than detection limit”. These values are known to be less than some value (e.g. 15%) of residue data set contains

97

observations