Children and Pesticides: New Approach to Considering Risk Is Partly in
Place (Letter Report, 09/11/2000, GAO/HEHS-00-175).

Pursuant to a congressional request, GAO provided information on the
Environmental Protection Agency's (EPA) efforts to reduce children's
exposure to pesticides by implementing the requirements of the Food
Quality Protection Act (FQPA), focusing on the: (1) approach EPA has
developed for making decisions about applying the new safety factor; (2)
progress that has been made in considering aggregate exposure and
cumulative effects; and (3) progress that has been made in reassessing
tolerances for pesticide residues.

GAO noted that: (1) when FQPA became law in 1996, EPA immediately began
efforts to consider the additional safety factors for children, using
available methods and data in an interim approach that has evolved over
time; (2) an internal committee now recommends whether to apply the
additional safety factor in pesticide reviews, based on data
completeness and evidence of increased susceptibility in children; (3)
using this approach, EPA has made decisions about applying the
additional safety factor for 105 of the more than 450 pesticides to be
reassessed; (4) it determined that an additional safety factor was
necessary in 49 cases and not necessary in the remaining 56 cases; (5)
EPA also had interim procedures in place for considering aggregate
exposure, which incorporate available data on exposures from food,
drinking water, and residential uses; (6) data on nonfood exposures have
been lacking for most pesticides, however, and methods for estimating
and combining such exposures are still being developed; (7) EPA has not
yet begun to consider cumulative effects in the regulatory process; (8)
it has determined that one group of pesticides that is considered to be
high-risk, called the organophosphates, will need to be assessed for
cumulative effect, but methods for doing so are still under development;
(9) potential effects of considering aggregate exposure and cumulative
effects are beginning to emerge; (10) on June 8, 2000, after applying
the additional safety factor and conducting an aggregate exposure
assessment for chlorpyrifos, EPA announced a need to substantially
reduce children's exposure to this pesticide by reducing its use on
foods frequently eaten by children and by eliminating nearly all
household uses; (11) EPA has reported progress in reassessing existing
tolerances for pesticide residues on foods, but relatively few of these
allowable limits have changed as a result of considering FQPA's new
requirements; (12) FQPA called for reassessing one-third of all existing
tolerances by August 1999--a goal EPA met; (13) GAO analyzed a larger
group, those counted as reassessed through April 2000; (14) for about 47
percent of these tolerances, the manufacturer agreed with EPA to
eliminate the tolerances and withdraw the pesticides from those uses,
before the additional safety factor or aggregate exposures were
considered; (15) in most of these cases the pesticide was no longer
being used on a particular food crop or the manufacturer decided not to
maintain the ability to use it on a particular food crop; and (16) in
reassessing tolerances, EPA has given priority to high-risk chemicals.

--------------------------- Indexing Terms -----------------------------

 REPORTNUM:  HEHS-00-175
     TITLE:  Children and Pesticides: New Approach to Considering Risk
	     Is Partly in Place
      DATE:  09/11/2000
   SUBJECT:  Pesticide regulation
	     Agricultural chemicals
	     Product safety
	     Children
	     Food and drug law
	     Safety regulation
	     Safety standards
IDENTIFIER:  EPA Office of Pesticide Programs

******************************************************************
** This file contains an ASCII representation of the text of a  **
** GAO Testimony.                                               **
**                                                              **
** No attempt has been made to display graphic images, although **
** figure captions are reproduced.  Tables are included, but    **
** may not resemble those in the printed version.               **
**                                                              **
** Please see the PDF (Portable Document Format) file, when     **
** available, for a complete electronic file of the printed     **
** document's contents.                                         **
**                                                              **
******************************************************************
GAO/HEHS-00-175

Appendix I: Methodology

24

Appendix II: Three Pesticides Evaluated Under FQPA's New Requirements:
Dicofol, Methomyl, and Phosmet

26

Appendix III: GAO Contact and Staff Acknowledgments

34

Table 1: Key Pesticide Regulatory Requirements Added by FQPA 7

Table 2: Examples of Questions Considered in Safety Factor Reviews 11

Table 3: Safety Factor Decisions 13

Table 4: Results of Tolerance Reassessments Involving FQPA Considerations 20

Figure 1: Overview of Risk Assessment and Mitigation for Pesticide Exposure
6

EPA Environmental Protection Agency

FQPA Food Quality Protection Act of 1996

ILSI International Life Sciences Institute

OPP Office of Pesticide Programs

Health, Education, and
Human Services Division

B-284334

September 11, 2000

The Honorable Edward M. Kennedy
The Honorable Barbara A. Mikulski
The Honorable Frank R. Lautenberg
The Honorable Jack Reed
United States Senate

Managing the risk of exposure to pesticides is important for all Americans,
but especially for children, whose developing systems can be more
susceptible to harm. The Food Quality Protection Act of 1996 (FQPA) requires
the Environmental Protection Agency (EPA), which regulates the use of
pesticides at the federal level, to reevaluate the amounts of pesticide
residues allowed on or in food--known as tolerances. EPA must ensure that
the tolerances are safe, that is, that there is a reasonable certainty that
no harm will result from exposure from all food and nonfood sources. In
doing so, unless another safety factor is determined to be appropriate, EPA
is required to apply an additional 10-fold safety factor in setting
tolerances to ensure the safety of foods for children. EPA is also required
to ensure that there is reasonable certainty that no harm will result to
children specifically from aggregate exposure to a pesticide from all
sources (such as lawn treatments, household uses, and drinking water, as
well as food). In addition, EPA must consider available information
concerning the combined or cumulative effects on children from groups of
pesticides that may act on the body in similar harmful ways. The law
requires EPA to consider all these factors in reassessing tolerances for
pesticide residues in foods, and in doing so, EPA must give priority to
pesticides that appear to pose the greatest risk to public health.

You asked us to examine how EPA is applying these requirements of the FQPA.
We focused our efforts on three questions:

ï¿½ What approach has EPA developed for making decisions about applying the
new safety factor?

ï¿½ What progress has been made in considering aggregate exposure and
cumulative effects?

ï¿½ What progress has been made in reassessing tolerances for pesticide
residues?

This report is based in part on a review of documents related to safety
factor determinations and pesticide risk assessments, as well as a review of
EPA's database for tracking tolerance reassessments. We supplemented this
analysis with interviews at EPA, as well as with federal health agencies,
chemical industry and environmental groups, and outside experts. We did not
evaluate EPA's regulatory decisions or the quality of the data behind them.
Appendix I provides further details on our methodology. We performed our
work from October 1999 through July 2000 in accordance with generally
accepted government auditing standards.

When FQPA became law in 1996, EPA immediately began efforts to consider the
additional safety factor for children, using available methods and data in
an interim approach that has evolved over time. An internal committee now
recommends whether to apply the additional safety factor in pesticide
reviews, based on data completeness and evidence of increased susceptibility
in children. Using this approach, EPA has made decisions about applying the
additional safety factor for 105 of the more than 450 pesticides to be
reassessed. It determined that an additional safety factor was necessary in
49 cases and not necessary in the remaining 56 cases.

EPA also has interim procedures in place for considering aggregate exposure,
which incorporate available data on exposures from food, drinking water, and
residential uses. Data on nonfood exposures have been lacking for most
pesticides, however, and methods for estimating and combining such exposures
are still being developed. EPA has not yet begun to consider cumulative
effects in the regulatory process. It has determined that one group of
pesticides that is considered to be high-risk, called the organophosphates,
will need to be assessed for cumulative effects, but methods for doing so
are still under development. The potential effects of considering aggregate
exposure and cumulative effects are beginning to emerge. On June 8, 2000,
after applying the additional safety factor and conducting an aggregate
exposure assessment for chlorpyrifos (an organophosphate sold as Dursban,
and the most widely used household pesticide in the United States), EPA
announced a need to substantially reduce children's exposure to this
pesticide by reducing its use on foods frequently eaten by children and by
eliminating nearly all household uses. EPA has not completed aggregate
exposure reviews for all 39 organophosphates individually, but when it does,
a cumulative assessment will be required for the group, which may identify
the need for additional changes.

EPA has reported progress in reassessing existing tolerances for pesticide
residues on foods, but relatively few of these allowable limits have changed
as a result of considering FQPA's new requirements. The act called for
reassessing one-third of all existing tolerances by August 1999--a goal EPA
met. We analyzed a somewhat larger group, those counted as reassessed
through April 2000. For about 47 percent of these tolerances, the
manufacturer agreed with EPA to eliminate the tolerances and withdraw the
pesticides from those uses (a pesticide has a separate tolerance for each
food crop it is used on), before the additional safety factor or aggregate
exposures were considered. In most of these cases the pesticide was no
longer being used on a particular food crop or the manufacturer decided not
to maintain the ability to use it on a particular food crop. Most of the
remaining reassessments in the group we analyzed resulted in no change. In
reassessing tolerances, EPA has given priority to high-risk chemicals. Some
high-risk pesticides have been reassessed. However, the only tolerances
counted as reassessed for the high-risk organophosphate pesticides were ones
that were canceled voluntarily, because the organophosphates will require a
cumulative assessment before existing tolerances can be formally reassessed.

In essence, evaluating and managing the risk from exposure to pesticides
involve determining the maximum safe level of exposure to a pesticide and
assessing whether expected actual exposure is below this maximum level.1
Figure 1 shows how these two steps relate to each other. As long as the
expected exposure remains lower than the maximum safe exposure, the risk
created by use of the particular pesticide is within acceptable limits and
usually no action is required. However, once expected actual exposure levels
exceed the maximum safe amount, EPA must determine the best ways to reduce
exposure below the safe level to mitigate the risk. Many possible risk
mitigation actions may be applied, ranging from prohibiting an agricultural
or residential use of the pesticide to changing directions for its use (such
as spraying less often). These mitigation steps are intended to reduce
overall exposure from all sources, including exposure through pesticide
residues on foods.

aIf this pesticide is part of a group that acts on the body in similar
harmful ways, EPA must assess risk based on cumulative effects, that is, the
combined aggregate exposure for each pesticide in the group.

FQPA made several fundamental changes in how EPA assesses and manages
pesticide exposure risks to humans.2 Under FQPA, EPA must reevaluate
existing tolerances for pesticide residues in foods within 10 years. In
doing so, EPA is required to (1) apply an additional 10-fold safety factor
in setting tolerances to ensure the safety of foods for children, unless
reliable data support a different factor; (2) ensure that there is
reasonable certainty that no harm will result to children from aggregate
exposure to a pesticide from food, drinking water, and residential sources;
and (3) consider available information concerning the cumulative effects on
children of pesticides that act in a similar harmful way (known as a common
mechanism of toxicity). These FQPA requirements also apply in setting
tolerances for new pesticides that are being registered and for new uses of
existing pesticides. In reassessing existing tolerances, EPA must give
priority to pesticides that appear to pose the greatest risk to public
health. Table 1 provides a brief overview of these requirements.

 Requirement                     Explanation
                                 To account for differences in sensitivity
                                 between humans and test animals, as well
                                 as differences in sensitivity among
                                 different people, EPA routinely applies a
                                 safety factor for humans that sets the
 Applying an additional safety   maximum acceptable or safe level of
 factor for children             exposure to a pesticide at 1/100 of the
                                 maximum amount observed as safe in
                                 animals. The additional safety factor for
                                 children generally reduces the safe level
                                 by another factor of 10, thereby lowering
                                 it to 1/1,000 of the amount for animals.
                                 Aggregate exposure is the exposure to a
                                 single pesticide that a person would be
                                 likely to face from all sources, including
                                 food, drinking water, and home and garden
 Considering aggregate exposure  use of the pesticide. Before FQPA, EPA was
                                 required only to assess exposure from
                                 food, although in practice the agency
                                 sometimes considered other sources as
                                 well.
                                 The cumulative effects of exposure are
                                 those that a person would be likely to
                                 face from the combined aggregate exposures
                                 to several pesticides that act on the body
                                 in similar harmful ways. Groups or classes
 Considering cumulative effects  of pesticides that act in a similar way
                                 are said to have a common mode of action
                                 or a common mechanism of toxicity. When
                                 EPA determines that a group of pesticides
                                 has a common mechanism of toxicity, it
                                 must consider the aggregate exposure from
                                 every pesticide in the group.
                                 A tolerance is the maximum legal amount of
                                 a pesticide residue that is allowed to
                                 remain on a food commodity that has been
                                 treated with the pesticide. It is usually
                                 expressed in parts per million or parts
                                 per billion. The allowed residue from the
 Reassessing tolerances          use of one pesticide on one specific crop,
                                 such as apples or asparagus, represents
                                 one tolerance. Thus a single pesticide may
                                 have many tolerances. FQPA requires EPA to
                                 reassess all tolerances in effect prior to
                                 passage of the law in August 1996 (9,721
                                 tolerances for more than 450 pesticides)
                                 to ensure that they are safe for children.

Among the difficulties EPA has faced in implementing FQPA requirements is
the fact that the scientific knowledge necessary to accomplish some of these
new mandates did not exist in 1996. EPA's pesticide regulatory process has
traditionally focused on exposures from food and considered each pesticide
separately. Under FQPA the agency has been required to develop the methods
and data to perform the new aggregate exposure and cumulative effects
assessments, which incorporate exposures from nonfood sources and the
combined effects of entire classes of pesticides that were previously
considered individually.

The first class of pesticides to be affected by cumulative assessment under
FQPA will likely be the organophosphates. EPA has identified the
organophosphates as a class of pesticides requiring cumulative assessment
because they can impair nervous system function by inhibiting the enzyme
cholinesterase. The organophosphates are older pesticides, and EPA considers
some of them to be more hazardous (although not all older pesticides are
necessarily more hazardous). They are of special concern because of their
toxicity and widespread use both in agriculture and in homes and gardens,
according to the National Research Council's 1993 report, Pesticides in the
Diets of Infants and Children.3 One of them is the single most widely used
household pesticide in the United States--chlorpyrifos, which is sold under
such names as Dursban and Lorsban.

Implementation

EPA's Office of Pesticide Programs (OPP) has the lead responsibility for
implementing the new FQPA requirements within its existing system of
pesticide regulation. This system includes registering or licensing new
pesticide products for use in the United States and reevaluating older
pesticides to ensure that they meet current health standards and that their
risks are adequately mitigated (a process required to reregister the
pesticide for continued use).4 Nearly 900 people organized into nine OPP
divisions carry out these activities. Two divisions have the main
responsibility for managing pesticide risk assessments: the Registration
Division (for assessing new chemicals and new uses of existing chemicals)
and the Special Review and Reregistration Division (for assessing most

conventional chemical pesticides for reregistration and for reassessing
tolerances as required by FQPA).5

To help conduct these risk assessments, the two divisions use analyses
provided primarily by another division, the Health Effects Division.
Scientists in the division examine the substantial body of studies and data
reports that under regulation are required to be submitted by the
pesticide's registrant (that is, the applicant for registration, usually the
manufacturer), along with other available data, to ensure the reliability of
the studies, assess the toxicity of the pesticide under review, and estimate
the risks of exposure. Sources of possible exposure that are considered
include water and residential contamination, in addition to the traditional
focus on food exposures. The risk assessments are subject to internal peer
review by the Health Effects Division staff.

Soon after FQPA became law in 1996, EPA began to include consideration of
the additional safety factor for children in its pesticide risk assessments,
as required. EPA developed interim guidelines for determining whether this
additional safety factor should be applied, and these procedures have
evolved over time.6 Under this approach, an internal review committee of
scientists, managers, and other experts within OPP--the FQPA Safety Factor
Committee--takes lead responsibility for recommending whether the additional
safety factor should be applied, with OPP management making the final
decisions. By March 2000, this committee had reviewed and prepared safety
factor recommendations for 150 pesticides, and OPP management had made final
safety factor decisions for 105 of them.

1996

EPA began to consider the additional safety factor in its pesticide reviews
and tolerance reassessments soon after FQPA was passed. By October 1996, EPA
had drafted an initial version of its approach to applying the additional
safety factor, and by January 1997 it had issued a notice providing detailed
guidelines for manufacturers on how pesticide reviews for registration and
reregistration would proceed, taking the new FQPA requirements into
account.7 In March 1997, EPA published an implementation plan for FQPA that
addressed how it would consider the additional safety factor for children.
The implementation plan called for applying the following approach:

ï¿½ EPA would require the additional 10-fold safety factor for children if the
agency lacked complete and reliable data to assess pre- or postnatal
toxicity relating to infants and children or if the data indicated pre- or
postnatal effects of concern.

ï¿½ If data were incomplete, an additional safety factor between 3 and 10
would be applied, with the size of the factor depending on how much
information was incomplete and the seriousness of any concerns about
effects.

ï¿½ If data were sufficient to demonstrate no potential pre- or postnatal
effects of concern, no additional safety factor would be applied.

To make recommendations to OPP management about applying the safety factor
in individual pesticide risk assessments, OPP established a Safety Factor
Committee in February 1998.8 This internal peer review group is composed of
risk assessors (including toxicologists and exposure experts) from OPP
science divisions and risk managers (staff responsible for risk mitigation
activities) from the divisions that regulate most of the chemicals. The
committee's procedures call for systematically reviewing both toxicology and
exposure data for each chemical, focusing on two overriding concerns: (1)
uncertainties in the data used for the toxicology and exposure assessments
(data gaps) and (2) evidence of increased susceptibilities in infants and
children (the potential for pre- and postnatal toxicity).9 Examples of the
types of questions considered in these subject areas are presented in table
2.

 Subject area           Questions
                        Do we have adequate hazard (toxicity) studies to
                        evaluate risk to infants and children?

 Toxicology             Do these studies show enhanced susceptibility in
                        infants and children? That is, do the effects in
                        the young occur at doses that do not cause effects
                        in adults?
                        What kinds of residue databases are available for
                        each crop (for example, field study data or
                        monitoring data) and what are their sources?

 Food exposure          Is information available on percentage of crop
                        treated?

                        According to food consumption data, which crops
                        contribute significantly to the diet for adults?
                        For infants and children?
                        Are models or monitoring data used to estimate
                        drinking water exposure?

                        If models are used, what scenarios are used in the
 Drinking water         model, and what are the resulting estimated
 exposure               environmental concentrations?
                        If monitoring data are used, what kinds of data
                        were collected and under what conditions (for
                        example, from vulnerable areas at maximum label
                        rates)?
                        Is the compound used around the home in such a way
                        that children and infants may be exposed? What are
                        the frequency and rate of application?

 Residential exposure   Are reliable biologically based exposure data or
                        epidemiology data available to support the results
                        of the assessment (for example, incident reports or
                        Centers for Disease Control and Prevention
                        biomonitoring data)?

Committee members are encouraged to apply scientific judgment as well as
qualitative and quantitative data in reaching consensus on whether to apply,
reduce, or remove the safety factor. The committee considers written reports
and oral presentations and seeks to reach a consensus in each case on the
FQPA safety factor it will recommend to OPP managers.10

As of March 2000 the committee had reviewed 150 pesticides and submitted
safety factor recommendations to OPP managers. In reviewing the committee's
justifications for its recommendations, we found that when the committee
identified both toxicology data gaps and evidence of increased
susceptibility in children, the pesticides were most likely to receive a
recommendation for a 10-fold safety factor. When there was no evidence of
increased susceptibility, but incomplete data, a safety factor was also
recommended, but it was less than 10 when the data suggested that a lower
safety factor was sufficient. Pesticides with neither increased
susceptibility nor data gaps usually received a recommendation for no
additional safety factor.

While EPA has incorporated consideration of the new safety factor in its
pesticide reviews, it has continued efforts to refine its policies on
applying the safety factor. A formal policy document on the safety factor
has been developed (it is not a regulation), which discusses in detail the
legal framework, overall approach, and related toxicology and exposure
issues. It is much more extensive than the guidelines under which the Safety
Factor Committee has been operating, but is consistent with them. An EPA
official explained that the two documents serve somewhat different purposes,
with the policy document providing comprehensive discussion of the issues
and the operating procedures translating those policies into practical
guidelines. The safety factor policy document was released for public
comment in 1999. As of July 2000, an agency official told us that EPA was
still assessing the comments it had received, and the document had not yet
been issued in final form.

OPP senior managers make the final decisions about whether to apply the
additional safety factor for children, based on the Safety Factor
Committee's recommendations and other considerations. As of the end of March
2000, OPP had made safety factor decisions for 105 of the 150 pesticides the
Safety Factor Committee had reviewed. OPP determined that a safety factor to
protect children, in addition to the routinely applied 100-fold safety
factor, was necessary in 49 cases and that available evidence was sufficient
to show that an additional safety factor was not required in 56 cases (see
table 3). For the organophosphate pesticides, OPP decided to apply the
additional safety factor in 24 cases and not to apply it in 15.

                               OPP decision
                                                                Factor
 Type of pesticide  Number     No additional 3-fold   10-fold   greater
                    reviewed   factor        factor   factor    than
                                                                10-fold
 Organophosphates   39         15            12       10        2
 All others         66         41            11       11        3
 Total              105        56            23       21        5
 Percentage of
 total                         53%           22%      20%       5%

In most cases, OPP managers adopted the committee's recommendations for the
level of safety factor to protect children, but in some cases they increased
those levels. In 19 of the 105 decisions, the factor was increased (made
more protective) to account for other types of uncertainties. OPP officials
said these uncertainties most often related to serious data gaps or special
concerns about the severity of a pesticide's health effects. In 5 cases, OPP
increased the combined safety and uncertainty factors to greater than
10-fold.11

To provide an indication of whether EPA was following its procedures, we
selected three high-risk pesticides of different types (including one
organophosphate) that had gone through the safety factor review process and
asked a consultant with expertise in environmental toxicology to review the
process and results. These three examples are described in appendix II. The
consultant concluded that in all three cases EPA's actions were thorough and
its conclusions reasonable.

Cumulative Effects

EPA has interim procedures in place for considering aggregate exposure in
its pesticide reviews and tolerance reassessments. These procedures
incorporate available data on exposures from drinking water and residential
uses, along with food exposures. Efforts are being made to improve available
data on nonfood exposures and the methods for estimating combined exposures
to individual pesticides from all sources. Efforts to consider the
cumulative effects of exposure to groups of similar pesticides have not
progressed as far as those for aggregate exposure. EPA has adopted policies
for identifying classes of pesticides that have a common mechanism of
toxicity, but methods for conducting cumulative assessments for these
classes of pesticides are still under development. As a result, EPA has not
yet considered cumulative effects in its pesticide risk assessments. In the
case of the organophosphates--and chlorpyrifos in particular--the potential
effects of aggregate exposure and cumulative assessments, in terms of needed
mitigation steps, are beginning to emerge.

Although formal policy guidance for performing aggregate exposure
assessments has not yet been issued in final form, EPA has interim
procedures in place for considering aggregate exposure using available data
and methods.12 Traditionally, EPA has assessed the risk of food use
pesticides on the basis of estimated exposure from all foods containing
residues of the pesticide. Under FQPA, EPA must also take into account the
amount of exposure to each pesticide that is likely to occur from drinking
water and from uses in and around the home. Common residential uses include
lawn and gardening uses, pet applications, and roach and termite treatments.
Not all pesticides have residential uses, but for those that do, adding
those types of exposures to food and water exposures might push the total
beyond the maximum safe level of exposure, leading to a need for mitigation
steps and possible changes in tolerances.13

Because of its traditional focus on pesticide exposures from foods, EPA's
data and methods for estimating food exposures are relatively highly
developed, but for most pesticides the agency has lacked the data and
methods to estimate nonfood exposures from drinking water and residential
uses. Moreover, EPA has lacked a method for combining exposures from these
sources to estimate aggregate exposure. While such data and methods are
being developed, the agency is using an interim approach that relies on
available data and conservative scientific judgments to protect health; that
is, in most cases the high-end estimates for drinking water and residential
exposures are added to the estimated food exposure.

Estimating exposures from pesticides used in and around the home has been a
particular challenge for EPA. In a working paper on assessing these
residential exposures, EPA stated that it relies primarily on the scientific
literature and industry sources because it lacks data for most pesticides to
characterize exposures from nonfood sources. Many types of needed data still
are not available. For example, an official told us that results from an
effort initiated in 1995 to collect data on outdoor residential exposures
(mainly from lawn chemicals) are only now coming in, and other efforts to
collect data on indoor residential exposures and commercial pesticide
applications are still under way. Methods for using such data to estimate
residential exposures are being developed, which, when approved by EPA's
Scientific Advisory Panel, will apply to both aggregate exposure and
cumulative assessments. EPA intends that ongoing development and refinement
will follow, as the agency gains experience with the methods.

Developing ways to assess both aggregate exposure and cumulative effects has
been more difficult and time-consuming than EPA anticipated, but developing
approaches to cumulative effects assessment has proved particularly
difficult. Experts in toxicology, exposure assessment, and risk assessment
methodologies have indicated that the science necessary to successfully
factor in these types of exposures, especially for cumulative effects, is a
work in progress. Beginning in 1997, EPA contracted with the International
Life Sciences Institute (ILSI)14 to convene workshops that EPA hoped would
bring together representatives of industry and academia and other interested
parties to participate in developing the new exposure assessment policies
required under FQPA. An ILSI group worked on aggregate exposure assessment
methods in 1997 and 1998, and another workshop reported on a framework for
cumulative risk assessment in 1999.

Because of the complexity of the scientific issues involved, EPA has
included considerable review by experts both inside EPA (including staff and
advisory committees) and in the academic and research communities in its
development of ways to measure aggregate exposure and cumulative effects.
According to one EPA official, while this peer review likely provided
benefits, it also slowed the process. The review has come from such groups
as EPA's Scientific Advisory Panel and the Tolerance Reassessment Advisory
Committee, which also have provided review of policies related to other
aspects of implementing FQPA.15 The Scientific Advisory Panel, which has
been the main source of ongoing peer review, now meets about every 2 months
for 4 to 5 days, the official told us. All panel meetings are open to the
public, industry, and environmental groups. Obtaining review from the
Tolerance Reassessment Advisory Committee required substantial time and
resources, according to the official, because many background and policy
documents had to be prepared.

To put cumulative assessment in place, EPA first needed methods to identify
groups of pesticides that act on the body in similar ways to cause adverse
health effects. A January 1999 document laid out the principles EPA applies
to determine if a group of pesticides acts through a common mechanism of
toxicity. Using these principles, EPA has identified the organophosphates as
one such group of pesticides because they impairnervous system function by
inhibiting the enzyme cholinesterase.16 The next step, developing the
methods for actually conducting cumulative assessments, has been more
difficult and time-consuming, and the first

draft of the methodology was not released for public comment until June
2000.17

EPA has not yet conducted a cumulative assessment. In addition to the need
for an acceptable methodology, a cumulative assessment requires aggregate
exposure assessments for each of the individual pesticides in the cumulative
effects group. While aggregate exposure assessments are in process for the
39 organophosphates, they are not all complete. Nonetheless, EPA agency
staff expect to present a pilot test of the proposed cumulative assessment
methodology to the Scientific Advisory Panel in September 2000, using a case
study of 25 organophosphate pesticides (including 7 chemicals with
residential uses and 2 with water residues). The pesticides will not be
named, to encourage focus on the assessment process, but the data will be
real and fairly complete.

Beginning to Emerge

EPA currently lacks the methods to consider cumulative effects associated
with classes of pesticides that have a common mechanism of toxicity, such as
the organophosphates, but the potential effects of aggregate exposure
assessment can be seen in the example of chlorpyrifos (sold under such names
as Dursban and Lorsban), a major organophosphate pesticide that has many
food and residential uses. Chlorpyrifos is found in many insect sprays and
is the single most widely used household pesticide in the United States. It
is also used by many commercial growers. In this instance, EPA has applied
an additional 10-fold safety factor to protect children and has assembled
considerable data about aggregate risk from the many sources of possible
exposure to chlorpyrifos. At a technical briefing on June 8, 2000, EPA
announced agreement with the pesticide's manufacturer to eliminate all home,
lawn, and garden uses of the pesticide, to eliminate the majority of termite
control uses, and to significantly lower allowable pesticide residues on
several foods regularly eaten by children, such as apples, grapes, and
tomatoes. These mitigation steps are intended to reduce expected aggregate
exposure below the maximum safe level.

Whether actions similar to those for chlorpyrifos will result from
considering aggregate exposures to other less widely used organophosphate
pesticides is unknown. However, when EPA conducts a cumulative assessment,
combining aggregate exposures for all the pesticides in this group,
additional mitigation steps may be necessary to protect children's health.

EPA has made some progress in reassessing existing tolerances, as required
by FQPA, but relatively few of these allowable limits for pesticide residues
have changed as a result of considering the law's new requirements. As of
April 2000, EPA reported that it had reassessed nearly 3,500 tolerances for
about 300 pesticides.18 However, nearly half of these tolerance
reassessments did not require consideration of the additional safety factor
for children or aggregate exposure, because the manufacturer agreed with EPA
to voluntarily eliminate the tolerances and withdraw the pesticides from
those uses. Most of the other reassessed tolerances were unchanged. Although
EPA has given priority to reassessing tolerances for high-risk pesticides,
reassessments for the high-risk organophosphate pesticides cannot be
completed until a cumulative assessment has been done for the group.

FQPA requires EPA to reassess all food use tolerances that were in effect
prior to passage of the law in August 1996--9,721 tolerances--to ensure that
the maximum residue levels they allow reflect any changes that might result
from the act's new protections. EPA must complete these tolerance
reassessments within 10 years of FQPA's enactment on a specific schedule:
one-third by August 1999, two-thirds by August 2002, and the rest by August
2006.

According to EPA, 3,290 tolerances (34 percent of the total) had been
reassessed by August 1999, the date of its first deadline. EPA also
announced that 2,178 (66 percent) of those reassessed tolerances were for
pesticides in the highest risk group. By April 2000, when we conducted our
analysis, the number of tolerance reassessments stood at 3,471, or 36
percent of the total.

Our analysis of the 3,471 tolerances that EPA counted as reassessed in April
2000 showed that nearly half of them--1,638 tolerances, or 47 percent--did
not involve consideration of the new FQPA requirements. Most of these
(1,257) were eliminated or canceled by EPA, with the manufacturer's
agreement, before risk assessments for the associated pesticides were
completed.19 Tolerance reassessments considered to be voluntary removals or
cancellations generally fell into two categories:

ï¿½ Tolerance no longer needed. When a particular use (for example, on apples)
has been removed from the list of registered uses for a pesticide, a
tolerance is no longer needed for that use. There were cases in which
tolerances for previously removed uses had not yet been canceled, and if
they were not needed for imported foods, EPA completed the cancellation
process.20

ï¿½ Manufacturer withdraws support. Manufacturers may withdraw support for
certain tolerances for a variety of reasons. For example, they may determine
that the costs of continued registration of the pesticide for that
use--including the costs of additional testing and registration fees--are
not justified by market conditions. An EPA official told us that in a number
of these cases, risk concerns that the agency expressed about the associated
pesticide contributed to the manufacturer's decision to drop the tolerance.

Unchanged

Fifty-three percent of the tolerance reassessments (1,833) were based on
pesticide risk assessments that considered aggregate exposure and the
additional safety factor for children. Most of these tolerance
reassessments--1,421 tolerances, or 77.5 percent--resulted in no change (see
table 4). The remainder of the tolerances were revoked (eliminating the
use), lowered (allowing less residue), or raised (allowing more residue).

                   Tolerance decision
 Total reassessed  Kept              Lowered (allows     Raised (allows
                   same     Revoked  less residue)       more residue)
 Number     1,833  1,421    98       139                 175
 Percentage        77.5%    5%       8%                  9.5%

EPA officials indicated that only a small percentage of tolerances were
lowered, even after the additional safety factor and aggregate exposures
were considered, because historically the agency has set tolerances
conservatively. As a result, they said, many tolerances were already at
levels that would pass FQPA's more stringent requirements. Likewise, EPA
officials told us that their decisions to raise 175 tolerances (that is, to
allow an increased concentration of pesticide residue to remain on the food)
do not represent an unacceptable risk to children or the general population.
Instead, the raised tolerances reflect new data from additional studies or
field trials that allowed EPA to perform more refined analyses of pesticide
exposure and risk.

Completed

FQPA required EPA to give priority to reassessing tolerances for high-risk
chemicals, and in August 1997 the agency published a Federal Register
notice21 that divided the pesticides with tolerances requiring reassessment
into three priority groups by level of risk. The highest priority group,
Group 1, which EPA considers to be the highest risk, included the
organophosphates, probable cancer-causing chemicals, and other pesticides of
particular concern. This group accounts for the largest proportion, about 57
percent, of all tolerances that need to be reassessed.22

Of the 3,471 tolerances that EPA has counted as reassessed through April
2000, two-thirds (2,286, or 66 percent) were for pesticides in Group 1. This
represents reassessment of 41 percent of all tolerances for the high-risk
pesticides. Less than 30 percent (483 of 1,691) of tolerances for the
high-risk organophosphate pesticides were counted as reassessed, and most of
these were canceled voluntarily. Even though safety factor decisions have
been made for 39 organophosphates and risk assessments including aggregate
exposure are in process, EPA has been unable to finalize the pesticide risk
assessments and their associated tolerance reassessments, because the
individual reviews must be combined in a cumulative assessment for all of
the organophosphates.

FQPA brought substantial changes to EPA's pesticide regulatory process, and
these changes are still works in progress. Some of the tools needed by EPA
to implement FQPA were not available when the law was enacted. EPA set about
developing the necessary procedures, methodologies, and data almost
immediately. The agency has adopted a series of interim approaches while
specifying, with the assistance of peer reviewers, more refined permanent
methods, which are now nearing completion. EPA has made progress in
reviewing pesticides and reassessing tolerances since 1996, but so far
relatively few tolerances have changed as a result of considering the new
FQPA requirements. While it is too early to tell what the future effects of
FQPA may be, the next few years could bring substantial changes, as the
organophosphates and other groups of high-risk pesticides are reconsidered
in the light of their aggregate exposures and cumulative effects. It appears
that EPA's recent decision on chlorpyrifos, for example, will result in
major changes in the uses of that pesticide that are intended to protect
people, and children in particular, from potentially adverse health effects.

We provided a draft of this report for comment to EPA, which supplied
technical comments that we incorporated where appropriate.

We will send copies of this report to the Honorable Carol Browner, EPA
Administrator, appropriate congressional committees, and other interested
parties. We will also make copies available to others on request.

If you or your staffs have any questions about this report, please call me
at (202) 512-7119. Major contributors to this report are listed in appendix
III.

Janet Heinrich
Associate Director, Health Financing
and Public Health Issues

Methodology

To examine how EPA is making decisions about applying the new safety factor
for children, we obtained documentation for each FQPA safety factor
determination. This consisted of three documents: (1) a summary log of
recommendations made by the Safety Factor Committee as of March 21, 2000,
with justifications for these recommendations, (2) a list of final decisions
made by OPP managers for regulatory purposes, and (3) a "Safety Factor
Report" dated March 22, 2000, which explains the differences between the
Safety Factor Committee recommendations and the final safety factor
decisions. We synthesized information from these lists with other
documentary evidence obtained from EPA, as well as information from EPA's
Tolerance Reassessment Tracking System (discussed below), to summarize the
results of FQPA safety factor decisions made to date. In addition, we
obtained detailed documentation and support material for three pesticides
that were reviewed by the Safety Factor Committee and asked a consultant
with expertise in environmental toxicology, H.B. Matthews, Ph.D., Society of
Toxicology Congressional Fellow, to review those cases in depth. We did this
to provide specific examples of EPA processes and their results and to
determine whether EPA's processes for assigning FQPA safety factors seemed
reasonable.

To determine what progress EPA has made in considering aggregate exposure
and cumulative effects, we obtained documents showing the development of
policies and procedures, including numerous interim drafts, and interviewed
EPA officials regarding the history behind the development of these
policies. We also held numerous discussions with EPA officials to determine
the extent to which EPA has assessed aggregate exposure and cumulative
effects in the pesticides reviewed since passage of FQPA, and the schedule
for the completion and implementation of the policies. Information on EPA's
plans, schedule, and progress in reassessing the group of organophosphate
pesticides is available on the OPP Web site at
www.epa.gov/pesticides/op/status.htm.

Finally, to identify what progress has been made in reassessing tolerances,
we obtained EPA's Tolerance Reassessment Tracking System database current to
April 11, 2000. EPA created this database to track the agency's progress in
meeting the deadlines associated with FQPA's requirement to reassess all
tolerances. The tracking system contains extensive information on all
permanent pesticide tolerances registered as of August 1996, as well as data
on each pesticide associated with the tolerances. In order to examine the
agency's use of the FQPA safety factor in assessing risk for each pesticide,
we added to the database a field containing the specific safety factor
decisions for each chemical. Because most pesticides in the tracking system
have more than one tolerance, we associated each pesticide's safety factor
with all tolerances for that chemical.

We used the tracking system to identify groups of tolerances and pesticides
with specific attributes by creating a series of database filters. For
example, by selecting for tolerances EPA counted as reassessed, which had an
FQPA safety factor decision, but were not reassessed administratively
through notices in the Federal Register,23 we identified those tolerances
that were reassessed as the result of a complete pesticide risk assessment,
including an FQPA safety factor decision and consideration of aggregate
exposures. Similarly, we used various criteria to determine other attributes
of the tolerances EPA has counted as reassessed, such as pesticide type
(organophosphate, carbamate, organochlorine, and so on), risk priority group
(Group 1, 2, or 3), and the resulting tolerance reassessment actions (raise,
lower, same, or revoke).

Three Pesticides Evaluated Under FQPA's New Requirements: Dicofol, Methomyl,
and Phosmet

We asked a consultant with expertise in environmental toxicology, H.B.
Matthews, Ph.D., Society of Toxicology Congressional Fellow, to help us
review three pesticide cases to provide detailed examples of how OPP
considers the new FQPA requirements in its pesticide risk assessments. In
these examples, we focused primarily on the work of the FQPA Safety Factor
Committee, but we considered other aspects of OPP's review and
decision-making process as well.

We selected three pesticides for review, using the following criteria: the
pesticides selected must have received a final safety factor decision; must
have tolerances to be reassessed under FQPA; and must be high-risk Group 1
pesticides24 of different types, including an organophosphate; preferably
should have multiple uses (tolerances), including nonfood residential uses
requiring aggregate exposure assessment; and must be likely to affect
children through food and other exposures. The pesticides we selected were
dicofol, an organochlorine; methomyl, a carbamate; and phosmet, an
organophosphate.

We obtained documents to describe the OPP review process for these three
pesticides from the OPP officials who manage the FQPA Safety Factor
Committee. We provided these documents to our consultant for his review,
which focused on such basic questions as the following:

ï¿½ As an overall conclusion, based on the input reports and committee
deliberations, does the Safety Factor Committee's recommended safety factor
appear to be reasonable and reasonably well justified?

ï¿½ Did the committee follow its own review criteria in a systematic way?

ï¿½ Did the committee adequately justify its decisions on (1) data
completeness and reliability (data gaps) and (2) evidence of increased
susceptibility in children?

ï¿½ How did the committee consider aggregate exposures?

ï¿½ Was cumulative assessment addressed in any way?

Our expert reviewer prepared comments addressing these questions for each of
the pesticide examples, and in some cases provided additional information
and opinion based on his own knowledge and experience. Those comments have
been combined with a description of the pesticide and a summary of its
review history in the sections below.

Dicofol is an organochlorine pesticide in EPA's high-risk Group 1. It is
used primarily on cotton, apples, and citrus crops and has nonresidential
uses on lawns and ornamental shrubs (for example, it may be used by
professional applicators on golf courses and landscaping, but it may not be
used by homeowners). Dicofol was first registered as a pesticide in the
United States in 1957. There are 50 current food use tolerances registered
for dicofol, all of which have been reassessed under FQPA requirements.

Dicofol was reviewed at one of the first meetings of the FQPA Safety Factor
Committee, on March 30, 1998. Information was excerpted from the pesticide's
reregistration document, which was nearly "Toxicological Considerations for
FQPA Safety Factor Selection."

The reregistration document's "FQPA considerations" section presented the
following conclusions: (1) the data provided no indication of increased
sensitivity in young animals, but (2) a developmental neurotoxicity study
was required, but not available (a data gap), because dicofol is an
organochlorine, is structurally related to DDT (which is neurotoxic), and is
considered an endocrine disruptor. A safety factor of 3 was recommended. The
main concern in terms of exposure was for occupational users. However, at
that time there were two homeowner uses, and no data were available to
assess residential exposure. Because of this lack of data, the
reregistration document recommended that residential use of dicofol be
discontinued. The document stated that EPA did not have the methods or the
data to consider potential cumulative effects from dicofol and other members
of the organochlorine class of pesticides.

The Safety Factor Committee reviewed the information from the reregistration
document for dicofol in March 1998 but was unable to reach a consensus
decision because it was concerned that its recommendation could set a
precedent for other endocrine disruptors. After seeking guidance from OPP
division directors, the committee recommended a 3-fold safety factor.

OPP managers accepted this recommendation, and the revised (July 2, 1998)
and final (November 1998) dicofol reregistration documents issued by EPA
included the 3-fold FQPA safety factor. Reasons given were as follows: (1)
aggregate exposure concerns were reduced because the two residential uses
were dropped, (2) cumulative effects from dicofol and other organochlorines
could not be considered, (3) strong concerns regarding occupational exposure
remained, and (4) significant risk mitigation actions were required,
including voluntary cancellation of some uses by the registrant. Eligibility
for reregistration was contingent on the results of a dermal toxicity study
and a dislodgeable foliar residue study (to be submitted).

Our reviewer noted that dicofol is an old pesticide, which is only
moderately toxic and not extremely persistent in the environment, with no
remaining residential uses. It raises concerns because of its structural
relation to DDT. Having considered the documents described above, he
reported that EPA's review of dicofol was thorough, although this was one of
the first pesticides the Safety Factor Committee reviewed and its procedures
were not as explicit as they later became. In the reviewer's judgment, data
for dicofol appeared to be complete, with the only identified data gap being
the need for a developmental neurotoxicity study; and the decision that
there was no evidence of increased susceptibility in children was well
justified. He concluded that EPA and the Safety Factor Committee responded
appropriately to FQPA requirements to consider aggregate exposures for
dicofol. EPA was not prepared to consider cumulative effects at that time.
The reviewer noted that because pesticides related to dicofol have been
removed from the market, cumulative effects from dicofol and other
organochlorines would be effectively limited.

Methomyl is a carbamate insecticide, also in EPA's high-risk Group 1. It has
a wide variety of registered uses on field, vegetable, and orchard crops,
turf farms, livestock quarters, and commercial premises and refuse
containers. It was first registered in the United States in 1968. There are
80 current food use tolerances for methomyl but no homeowner residential
uses. Ornamental and greenhouse uses were canceled voluntarily during the
course of the pesticide's review in 1998. Tolerance reassessment under FQPA
has been completed.

Methomyl was reviewed by the Safety Factor Committee on April 6, 1998. The
committee received a report from the internal toxicology review committee
(known as the Hazard Identification Assessment Review Committee), along with
a document known as the "FQPA Responses," which was prepared to address the
committee's specific questions in its standard operating procedures.

The toxicology review committee's report reviewed the toxicology database,
including a new study submitted by the registrant (21-day dermal toxicity in
rabbits). Data gaps were identified: acute and subchronic neurotoxicity
studies were required but not available. Consequently, data on
cholinesterase inhibition, behavioral effects, and nervous system effects by
the pesticide were missing. The neurotoxicity studies were required because
methomyl is a carbamate (a class of pesticides with known neurotoxic
effects) and neurotoxic effects from methomyl were seen in dogs and rabbits.
The requirement for a developmental neurotoxicity study was noted as
"reserved," pending the results of the acute and subchronic studies.

The Safety Factor Committee decided to recommend a 3-fold safety factor for
methomyl, based on (1) no indication of increased susceptibility in children
and (2) data gaps--specifically, the lack of acute and subchronic
neurotoxicity studies. The committee reviewed exposure data for food and
drinking water, but not for residential exposures because there were no
residential uses. Data quality was considered generally high, and realistic
assumptions (conservative models) were used.

The final reregistration document for methomyl, dated December 1998,
reflected the 3-fold FQPA safety factor decision. Because there were no
homeowner uses, the aggregate exposure assessment did not consider
residential exposures. However, because methomyl is produced when another
pesticide, thiodicarb, degrades, the aggregate risk assessment considered
methomyl residues from applications of both methomyl and thiodicarb. The
document also states: "The Agency does not have, at this time, available
data to determine whether methomyl has a common mechanism of toxicity with
other substances or how to include this pesticide in a cumulative risk
assessment. For the purposes of this assessment, therefore, the Agency has
not assumed that methomyl has a common mechanism of toxicity with other
substances."

On the basis of his review of the EPA documents, our reviewer felt that a
3-fold safety factor was appropriately conservative for methomyl. Input data
were provided to address the Safety Factor Committee's review questions, and
the committee's report indicated a thorough and careful review. That report
provided justification for reducing the 10-fold safety factor to 3-fold by
referring to sections of the source reports. Regarding the data gap for
acute and subchronic neurotoxicity studies in rats that was the reason for
the 3-fold safety factor, the reviewer noted that the gap did not seem to be
a pressing need, because similar data were available for two other species,
dogs and rabbits. His opinion was that if EPA had been willing to accept the
data for dogs and rabbits as an alternative to the rat data, the safety
factor might have been removed. The lack of a cumulative risk assessment in
this case was not likely a problem, the reviewer said, because of the short
half-life of the chemical, meaning that it would dissipate rapidly.

Phosmet is a member of the largest class of insecticides, the
organophosphates. It is a broad-spectrum insecticide that causes systemic
toxicity by inhibiting cholinesterase, and there also have been concerns
about carcinogenicity. Phosmet is marketed for agricultural uses,
nonagricultural occupational uses, and homeowner uses to control pests,
including moths, beetles, weevils, lice, flies, fleas, and ticks. It is used
on a variety of fruit and vegetable crops (especially apples and peaches),
tree crops, nut trees, cotton, and ornamentals and in forestry. In addition,
phosmet is used for direct animal treatments on cattle, swine, and dogs
(flea and tick treatments). There are 43 current food use tolerances for
phosmet, of which 1 has been revoked voluntarily during the course of the
review. The remainder have yet to be reassessed, pending the required
cumulative assessment for the organophosphates.

The Safety Factor Committee has reviewed phosmet twice since the law passed
in 1996. The first review was through what OPP calls its "OP
[Organophosphate] Marathon Meetings," in which all of the organophosphate
pesticides, including phosmet, were reviewed together. The Safety Factor
Committee held its marathon meeting on June 15 and 16, 1998. In the case of
phosmet, it concurred with the findings of the May 1998 toxicology review
committee marathon meeting and recommended a 3-fold FQPA safety factor,
based on data gaps noted by the toxicology group.25 Specifically, the
marathon meetings found that there was no evidence of increased
susceptibility to phosmet in young animals, but there were data gaps for two
types of neurotoxicity studies, and the requirement for a developmental
neurotoxicity study would depend on the results of those other studies.
Because no data were available to assess neurotoxicity, cholinesterase
inhibition, behavioral effects, or neuropathology for phosmet, the 3-fold
safety factor was recommended.

The second review of phosmet took place in summer 1999 after the registrant
submitted new data, including the acute and subchronic neurotoxicity
studies. On the basis of the results of these studies and additional
information from the registrant, the toxicology review committee determined
that the developmental neurotoxicity study was not required. The Safety
Factor Committee then concluded that there were no remaining data gaps and
no evidence of increased susceptibility. The committee reported that
adequate actual data, surrogate data, and/or modeling outputs were available
to satisfactorily assess dietary food and residential exposures26 and to
provide a screening level of drinking water exposure assessment.
Consequently, the committee recommended that the FQPA safety factor be
removed.27

OPP's regulatory decision agreed with the Safety Factor Committee that no
additional FQPA safety factor was needed for phosmet, and the reregistration
document was revised to reflect the new data and decisions (see the version
of October 7, 1999, and the February 9, 2000, public release version). The
risk assessment for phosmet concluded that (1) dietary food and water risks
are not a concern, even when combined, from either acute or chronic
exposure, (2) risks from residential exposure to treated dogs and garden
uses are a concern, especially for toddlers exposed to treated dogs, and (3)
there also are concerns for workers handling the pesticide and regarding
some ecological hazards (to birds, water, and honey bees). This latest
revised risk assessment describes the risks associated with use of phosmet
alone, and it may be revised again when EPA has conducted the necessary
cumulative assessment for the organophosphates.

Our reviewer concluded that through repeated reviews and reports, EPA's
consideration of the data on phosmet has been very thorough. The Safety
Factor Committee reviews followed the standard operating procedures
systematically, and decisions on the completeness and reliability of the
data and the lack of data gaps were adequately justified. The committee's
conclusion that there was no apparent developmental or reproductive toxicity
also was adequately justified and supported removal of the 10-fold safety
factor. Our reviewer stated that phosmet is rapidly metabolized and degrades
quickly under most environmental conditions; therefore it is seldom detected
in food or water, and exposures usually are very low. This helps explain why
aggregate exposure to phosmet is low. In our reviewer's opinion, calculated
exposure of children resulting from contact with treated dogs appeared to be
conservative. Regarding the postponed cumulative risk assessment, he noted:
"Having worked on the problem of cumulative risk assessment, I realize that
issues relating to cumulative risks present a very complex and controversial
problem--a problem that is not going to be easily solved. And the solution,
when it comes, is not likely to be to anybody's complete satisfaction."

Incidents of human poisoning from phosmet apparently are relatively common,
because our reviewer said that phosmet accounts for the largest number of
residential exposures to pesticides that result in treatment in a health
care facility. But he observed that these incidents are not an indication of
unusual toxicity as much as they are a result of incorrect use. Almost all
of the poisoning incidents resulted from failure to properly dilute a
concentrated formulation of phosmet for use as flea dip treatment for dogs.
Regarding the question of whether phosmet is likely to cause cancer, our
reviewer noted that the evidence is limited to reports of increased
incidences of tumors in mouse livers (which were marginally statistically
significant) and that there are differences of opinion as to the relevance
of increased mouse liver tumors to human health risks.

GAO Contact and Staff Acknowledgments

Janet Heinrich, (202) 512-7119

The following staff made key contributions to this work: Ellen M. Smith,
Matthew W. Byer, Stanley G. Stenersen, and Katherine M. Iritani.

(101887)

Table 1: Key Pesticide Regulatory Requirements Added by FQPA 7

Table 2: Examples of Questions Considered in Safety Factor Reviews 11

Table 3: Safety Factor Decisions 13

Table 4: Results of Tolerance Reassessments Involving FQPA Considerations 20

Figure 1: Overview of Risk Assessment and Mitigation for Pesticide Exposure
6
  

1. The word "pesticide" is used broadly here to include insecticides,
herbicides, fungicides, rodenticides, and the like, which are designed to
prevent, destroy, repel, or reduce pests.

2. FQPA amended existing laws that are intended to protect the public from
harmful exposures to pesticides. The amended laws were primarily the Federal
Insecticide, Fungicide, and Rodenticide Act (7 U.S.C. 136 et seq.) and the
Federal Food, Drug, and Cosmetic Act (21 U.S.C. 301 et seq.).

3. National Research Council, Pesticides in the Diets of Infants and
Children (Washington, D.C.: National Academy Press, 1993), pp. 17, 245-46.

4. As directed by the 1988 amendments to the Federal Insecticide, Fungicide,
and Rodenticide Act, EPA has been conducting a comprehensive review of
pesticides initially registered before November 1, 1984, to determine their
eligibility for reregistration.

5. While OPP regulates all types of pesticides, we focused our review on the
largest group, the conventional pesticides, leaving aside antimicrobials and
biopesticides. These other types of pesticides are the responsibility of
OPP's Antimicrobial Division and Biopesticides and Pollution Prevention
Division, respectively. Those two divisions both assess and manage the risks
associated with chemicals under their purview. FQPA standards apply, and
reassessed tolerances for antimicrobials and biopesticides are counted in
the EPA Tolerance Reassessment Tracking System.

6. EPA's interim guidelines are operating policies and procedures that the
agency follows to make the decisions required by FQPA, such as whether to
apply the additional safety factor for children, while the process of
developing more complete, formal policies and procedures continues.
Decisions made under interim guidelines may be revisited, if appropriate, as
methods and data are improved and policies refined.

7. Pesticide Registration Notice 97-1 stated that data requirements for
pesticide registrations and reregistrations would not be revised immediately
to reflect the new FQPA requirements. The notice included, however, a
detailed outline of the additional data and studies that EPA was requesting
from registrants on a voluntary basis to support timely pesticide reviews
under FQPA.

8. EPA officials told us that, up to that time, two predecessor groups had
conducted the reviews. The first was the Reference Dose Committee in the
Health Effects Division, which was responsible for reviewing pesticide
toxicity data at the time FQPA was passed. It continued its reviews,
including consideration of the new safety factor, until summer 1997, when a
new group was established for that purpose. This second group was the Hazard
Identification Assessment Review Committee, a committee that still reviews
the toxicology database for each pesticide to determine what adverse health
effects the pesticide might cause. This committee performed a limited number
of reviews, based on toxicity data only.

9. According to an EPA official, the data studies that are most useful in
assessing pre- and postnatal toxicity are the prenatal developmental
toxicity studies in rats and rabbits, the two-generation reproductive
toxicity study in rats, and, when available, the developmental neurotoxicity
study in rats. All of these studies are done using pregnant and young
animals. The first two studies are always required for assessing food use
pesticides, and the third study (the developmental neurotoxicity study) is
conditionally required.

10. In the few cases in which the committee could not reach a consensus
recommendation, a memorandum was prepared to division directors for their
decision. An example of this situation is dicofol, a pesticide case
described in app. II.

11. OPP most often applied a higher safety factor when a risk assessment
used a "lowest-observed-adverse-effect level" in animal studies as a
threshold no-harm level, instead of the preferred
"no-observed-adverse-effect level," and thus was less certain and
protective. Safety factors for one pesticide, for example, were increased by
the FQPA 10-fold factor for children and multiplied by an additional 10-fold
factor for uncertainties.

12. The latest draft of the policy guidance for performing aggregate
exposure assessments, dated October 1999, was being revised to respond to
public comments for release in summer 2000.

13. In app. II, phosmet provides an example of a pesticide with residential
uses of concern, in this case as an insecticide for home and garden use and
a flea dip treatment for dogs.

14. ILSI is a nonprofit worldwide foundation established in 1978 to advance
the understanding of scientific issues relating to nutrition, food safety,
toxicology, risk assessment, and the environment. Its Risk Science
Institute, established in 1985, seeks to advance and improve the scientific
basis of risk assessment.

15. The Scientific Advisory Panel is a scientific peer review group that
advises EPA on major issues. It has reviewed and commented on science policy
issues and their related guidance documents, including the policies and
procedures for applying the safety factor and methods for aggregate exposure
and cumulative assessments. Panel members are mainly from academia or other
government agencies. The Tolerance Reassessment Advisory Committee met from
May 1998 through October 1999. It included representatives of the pesticide
industry, agricultural interests, and environmental groups who were convened
by EPA to identify key science policy issues affecting risk assessment. It
was replaced by a new group, the Committee to Advise on Reassessment and
Transition, in June 2000.

16. According to an EPA official, OPP has identified two other groups of
pesticides that have a common mechanism of toxicity and are candidates for
cumulative risk assessment: the cholinesterase-inhibiting carbamates and
some of the chloroacetanilides.

17. The draft was released for public comment on June 29, 2000. Following
the comment period, EPA plans to revise the draft and publish it in final
form, but no date has been set.

18. EPA sets a tolerance for each use of a pesticide on a single food
product; thus pesticides applied to many types of food will have multiple
tolerances. Some pesticides have more than 100 tolerances.

19. The remainder of the tolerances that were canceled in this way involved
biological pesticides regulated by another EPA division, inert ingredients
and some other pesticides exempted from FPQA's requirements (for example,
for not applying to food, such as use on tobacco), and other special
circumstances.

20. An import tolerance is maintained for imported foods for which there is
no U.S. registration for the pesticide in question. FQPA applies the same
standards to imports as it does to other tolerances.

21. Federal Register, vol. 62, no. 149, pp. 42020-30, Aug. 4, 1997.

22. Group 1 (228 pesticides/5,546 tolerances), the highest-risk chemicals,
includes the organophosphates, carbamates, organochlorines, and probable
carcinogens. Group 2 (93 pesticides/1,928 tolerances) includes the
lower-risk possible carcinogens and all remaining reregistration chemicals
(those that were first registered before 1984). Group 3 (148
pesticides/2,247 tolerances) includes the remaining pre-FQPA pesticides with
reregistration eligibility decisions, the remaining post-1984 pesticides,
biological pesticides, and the remaining inerts.

23. EPA officials told us that the tolerances counted as reassessed
administratively through notices in the Federal Register were not the result
of a full pesticide risk assessment and hence were not affected by an FQPA
safety factor decision.

24. Group 1 pesticides are those identified by EPA as the highest risk and
whose tolerances require priority reassessment under FQPA.

25. There was also a separate toxicology review committee meeting on phosmet
on September 4, 1997, as part of phosmet's reregistration process. The
discussion covered FQPA requirements, such as data availability and data
gaps--specifically, the need for a developmental neurotoxicity study;
possible endocrine disruption; and possible human carcinogenic effects. The
toxicology review committee recommended a 3-fold safety factor at that time.

26. There were no chemical-specific data for phosmet to assess potential
exposures to children following outdoor residential applications. Therefore
the draft Standard Operating Procedures for Residential Exposure Assessments
was used with surrogate data. This type of analysis is intended to represent
worst-case or screening-level assessments.

27. Phosmet's potential for causing cancer in humans was also reassessed in
1999 through a separate internal committee process. As a result, phosmet was
classified in the category "suggestive evidence of carcinogenicity, but not
sufficient to assess human carcinogenic potential." This was consistent with
the previous classification, and therefore no change was needed in the risk
assessment, which already took this level of cancer risk into account.
*** End of document. ***