Pesticides: EPA's Efforts to Collect and Take Action on Exposure Incident
Data (Letter Report, 07/12/95, GAO/RCED-95-163).

Pursuant to a congressional request, GAO reviewed the Environmental
Protection Agency's (EPA) monitoring of human exposures to pesticides,
focusing on whether EPA: (1) collects data on exposure arising from the
use on nonagricultural pesticides; (2) takes action in response to
potential health risks from such exposure; and (3) receives sufficient
information to assess whether unacceptable risks are occurring.

GAO found that: (1) EPA has collected pesticide exposure data from
pesticide registrants and public and private entities since the 1970s
and, in 1992, it implemented a computerized system to organize and track
such data; (2) EPA has not assigned full-time staff to data collection
and processing, therefore, the system has a data entry backlog, which
limits its effectiveness; (3) EPA acted in 19 instances between 1989 and
1994 to protect the public from pesticide risks; (4) EPA often cannot
assess whether a pesticide poses an unacceptable health risk, since
incident reports frequently lack key data, may not be representative, or
are not submitted; (5) an EPA work group is developing a long-term plan
to collect and manage exposure data, but it has yet to develop a plan
for putting the most cost-effective improvements into effect; (6) to
improve the number and quality of exposure reports, EPA has proposed a
rule that requires pesticide registrants to submit more detailed data on
exposure incidents and clarifies the registrants' responsibilities; (7)
EPA is determining the feasibility of having registrants who submit
large numbers of reports to submit them electronically; and (8) the
current exposure monitoring system includes data on both agricultural
and nonagricultural pesticides, since EPA collects and processes the
same information for those chemicals.

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

 REPORTNUM:  RCED-95-163
     TITLE:  Pesticides: EPA's Efforts to Collect and Take Action on 
             Exposure Incident Data
      DATE:  07/12/95
   SUBJECT:  Pesticides
             Environmental monitoring
             Reporting requirements
             Health hazards
             Computerized information systems
             Data collection operations
             Information processing operations
             Federal/state relations
             Data integrity
IDENTIFIER:  EPA Incident Data System
             EPA Pesticide Incident Monitoring System
             California
             Washington
             New York
             Oregon
             Texas
             NIOSH Sentinel Event Notification System for Occupational 
             Risks Program
             
**************************************************************************
* This file contains an ASCII representation of the text of a GAO        *
* report.  Delineations within the text indicating chapter titles,       *
* headings, and bullets are preserved.  Major divisions and subdivisions *
* of the text, such as Chapters, Sections, and Appendixes, are           *
* identified by double and single lines.  The numbers on the right end   *
* of these lines indicate the position of each of the subsections in the *
* document outline.  These numbers do NOT correspond with the page       *
* numbers of the printed product.                                        *
*                                                                        *
* 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.                                          *
*                                                                        *
* A printed copy of this report may be obtained from the GAO Document    *
* Distribution Facility by calling (202) 512-6000, by faxing your        *
* request to (301) 258-4066, or by writing to P.O. Box 6015,             *
* Gaithersburg, MD 20884-6015. We are unable to accept electronic orders *
* for printed documents at this time.                                    *
**************************************************************************


Cover
================================================================ COVER


Report to Senator Harry M.  Reid,
U.S.  Senate

July 1995

PESTICIDES - EPA'S EFFORTS TO
COLLECT AND TAKE ACTION ON
EXPOSURE INCIDENT DATA

GAO/RCED-95-163

EPA's Data on Exposure to Pesticides


Abbreviations
=============================================================== ABBREV

  EPA - Environmental Protection Agency
  FIFRA - Federal Insecticide, Fungicide, and Rodenticide Act

Letter
=============================================================== LETTER


B-261118

July 12, 1995

The Honorable Harry M.  Reid
United States Senate

Dear Senator Reid: 

Pesticides--herbicides, insecticides, fungicides, and rodenticides,
among others--are designed to kill and control living organisms such
as unwanted species of plants, insects, and animals.  Nonagricultural
pesticides--those not intended for use in producing or preserving
foods or crops--are used in places where people live, work, and play,
such as homes, buildings, gardens, lawns, parks, and golf courses. 
Because pesticides are designed to destroy or control living
organisms, exposure to them can be hazardous. 

In response to your concerns about the potential health risks to
consumers of exposure to pesticides through accidents, misuse, or
lack of awareness of their hazards, we agreed to determine whether
the Environmental Protection Agency (EPA) (1) collects information on
exposure to or incidents arising from the use of nonagricultural
pesticides in order to evaluate the extent of risk to human health,
(2) takes action when the information it receives on such incidents
indicates potential health risks, and (3) receives sufficient
information to assess whether unacceptable risks are occurring. 
While we focused our review on exposure to nonagricultural
pesticides, our discussion of EPA's monitoring activities also
encompasses agricultural pesticides because the agency's current
system for monitoring incidents of exposure includes both
agricultural and nonagricultural pesticides. 


   RESULTS IN BRIEF
------------------------------------------------------------ Letter :1

Since the 1970s, EPA has collected data on incidents of exposure to
pesticides.  Beginning in 1992, the agency has used a computerized
management information system--the Incident Data System--to organize
and track reports on such incidents. 

Using these data, EPA has taken a number of measures to protect
public health.  For example, after analyzing data from emergency
rooms, hospitals, and poison control centers, the agency determined
that most uses of arsenical (arsenic-based) ant baits could no longer
be used in homes because of the potential high risk to children. 

Although EPA has been able to take some actions using the data
collected on incidents, the reports it receives frequently contain
insufficient information for the agency to determine whether the
pesticide involved presents an unacceptable risk to human health. 
Key data such as whether the pesticide product was misused, what
symptoms the victim exhibited, and how the exposure occurred are
often missing from the information EPA receives.  As a result, there
is often no clear evidence of cause and effect, and EPA has little or
no basis to assess risk and determine whether new or revised
regulatory measures are needed. 

EPA has recognized the limitations of its data on incidents of
exposure to pesticides.  In September 1994, it established a work
group to develop a long-term plan for collecting and managing these
data.  Although the work group has already identified several
specific actions that could improve the agency's ability to collect
data on such incidents, it is still gathering data and has not yet
developed a long-term plan with milestones for putting the most
cost-effective improvements into effect.  EPA has also proposed a
rule that would require pesticide registrants to submit more detailed
data, when available, on incidents of exposure to the agency.  EPA
believes this rule will improve the quality of reports on incidents,
as well as increase their number, since it will clarify the
responsibilities of those who register the pesticides with
EPA--primarily the producers. 


   BACKGROUND
------------------------------------------------------------ Letter :2

Nonagricultural pesticides encompass a wide range of
products--including home and garden insecticides and fungicides,
sterilants, insect repellents, and household cleaning agents--and the
potential for exposure is significant.  The effects of exposure on
humans depend on the characteristics of the pesticide, dosage,
duration of the exposure (usually through inhalation, skin contact,
or ingestion), and physiological reaction of the person affected. 
Some people suffer no effects; others experience symptoms ranging
from relatively mild headaches, skin rashes, eye irritation, and
general flu-like symptoms to more serious chemical burns, paralysis,
and even death.  Chronic and delayed-onset illnesses such as cancer
may only appear years after repeated exposure to small doses of a
pesticide. 

Under the Federal Insecticide, Fungicide, and Rodenticide Act
(FIFRA), EPA is responsible for ensuring that pesticides, when
properly used, do not have any unreasonable adverse effects on the
environment (any unreasonable risk to man or the environment, taking
into account the economic, social, and environmental costs and
benefits of the use of any pesticide).\1 The act authorizes EPA to
register pesticide products, specify the terms and conditions of
their use before they are marketed, and remove unreasonably hazardous
pesticides from the marketplace.  Thus, registrations are basically
licenses for specified uses of pesticide products.  The act also
requires that EPA reassess and reregister thousands of older
pesticide products on the basis of current scientific standards. 

The process requires the pesticides' registrants to complete studies
of various health and environmental effects, which are then reviewed
by EPA to determine whether the products can be reregistered and thus
remain on the market.  Section 6(a)(2) of FIFRA also requires that
registrants of pesticides report to EPA any additional factual
information that they may obtain about unreasonable adverse effects
that their registered pesticides have on the environment.  According
to EPA, the additional information on adverse effects that the
registrants must report includes toxicology studies, human
epidemiological and exposure studies, and efficacy studies, as well
as incidents of pesticide exposure. 

In addition, the act requires that EPA monitor, among other things,
the extent to which humans, animals, and the environment are
incidentally exposed to pesticides, trends over time, and the sources
of contamination.  According to EPA, the data on incidents of
pesticide exposure often augment the extensive studies performed by
registrants as part of reregistration.  This review focused on the
data on incidents of exposure reported to EPA. 

When EPA identifies risks during its review of data on incidents, the
agency may initiate one or more actions.  These actions include
restricting pesticide uses by placing specific instructions for use
on the product's label (for example, requiring protective equipment),
canceling specific uses of the pesticide, and/or canceling the
pesticide's registration, thus removing the pesticide from the
marketplace. 


--------------------
\1 7 U.S.C.  136(bb). 


   EPA COLLECTS DATA ON INCIDENTS
   FROM REGISTRANTS AND OTHER
   SOURCES
------------------------------------------------------------ Letter :3

From 1978 through 1981, EPA coordinated and collected information on
incidents of pesticide exposure through its Pesticide Incident
Monitoring System.  The system's reports originated from registrants
and from sources such as state and local agencies, poison control
centers, health clinics, and hospitals that provide this information
voluntarily.  After this system was eliminated because of funding
cuts, EPA continued to receive reports of incidents involving
pesticides from registrants and from the voluntary sources.  However,
the agency did not have an automated system for monitoring data on
such incidents until 1992, when it developed the Incident Data System
to organize and track data originating from both pesticide
registrants and the voluntary sources.  This system stores data on
incidents involving humans, domestic animals, wildlife (fish, birds,
and mammals), and groundwater and surface water. 

Although most--about 87 percent, according to an Office of Pesticide
Programs official--of the reports on incidents in EPA's system come
from registrants, EPA also receives supplementary data from voluntary
sources.  FIFRA does not require states or sources other than
registrants to collect or submit data on exposures.  However, some
states have established mandatory reporting regulations specifically
for pesticide-related illnesses.  EPA currently receives data on
incidents routinely from five of these states--either directly or
indirectly.  California and Washington voluntarily send annual
summary reports to EPA directly, while the agency receives quarterly
reports on incidents in New York, Oregon, and Texas from the National
Institute of Occupational Safety and Health,\2 which collects data
from these states.\3 According to an EPA health statistician, other
states may report some data on incidents to EPA, but not routinely. 

Written reports on incidents are forwarded to a single location at
EPA headquarters, where they are cataloged and screened to determine
whether they warrant detailed attention and/or consideration in
registration or reregistration reviews.  Aggregate reports are
periodically generated from the data entered into the computerized
system to determine if patterns are emerging that could cause
concern. 

However, EPA has a backlog of data to be entered into the Incident
Data System, thus limiting the effective use of the data it receives. 
Although the agency currently has a number of people involved in
collecting and analyzing data on pesticide incidents, only a portion
of each individual's work time is spent dealing with incidents, and
no one has been assigned full-time to data collection efforts such as
entering data into the system.  Since the system became operational
in June 1992, EPA has received about 12,575 reports.  While about
8,125 of the reports had been entered into the system as of April
1995, information on about 3,250 incidents had not yet been entered
because of limited staff resources.  Another estimated 1,200 reports,
which the registrants say contain confidential information, will not
be entered into the system until the agency determines the validity
of these claims. 


--------------------
\2 The National Institute for Occupational Safety and Health, part of
the Centers for Disease Control and Prevention, is the federal
institute responsible for conducting research and making
recommendations for the prevention of work-related illnesses and
injuries. 

\3 These data are collected under the Sentinel Event Notification
System for Occupational Risks program, which is designed to encourage
the development, implementation, and evaluation of state-level
surveillance systems and to implement measures to prevent certain
work-related illnesses. 


   EPA HAS TAKEN MEASURES TO
   PROTECT PUBLIC HEALTH USING
   DATA ON INCIDENTS
------------------------------------------------------------ Letter :4

According to EPA staff, data on incidents of exposure played a
significant part in 19 instances in which the agency took measures to
protect the public health between 1989 and 1994.  For example, after
analyzing data from emergency rooms, hospitals, and poison control
centers, the agency determined that most uses of arsenical
(arsenic-based) ant baits could no longer be used in homes because of
the potential high risk to children.  In another instance, EPA, after
reviewing cases involving the deaths of two individuals who died when
they entered structures treated with methyl bromide, required that
the product's label be revised to extend the period before people are
allowed to reenter a treated area. 

In a third case, EPA determined that many reports of adverse
reactions to pet care products likely resulted from misuse of the
product or accidental exposure.  Specifically, it appeared that some
animals and humans had reacted adversely as a result of overdoses or
repeated applications at too frequent intervals, or simultaneous
applications of multiple pesticide products to pets and their
environment.  In several incidents, cats were injured by pet care
products intended for dogs only.  In this case, the aggregate number
of incidents and other data in the Incident Data System on all pet
care products led EPA to draft a proposed Pesticide Regulation
Notice.  EPA intends for the proposed notice to provide registrants
of pesticide products with instructions on how a product's label
should be changed to reflect the proper intervals for repeated use of
the product and to restrict the use of the product to animals for
which it was specifically intended.  At the time of our review, the
proposed notice had not been finalized.  (App.  I lists other
examples of actions that EPA has taken using data on incidents
involving nonagricultural pesticides.)


   INFORMATION THAT EPA COLLECTS
   ON EXPOSURE INCIDENTS MAY NOT
   ALWAYS BE SUFFICIENT
------------------------------------------------------------ Letter :5

Although EPA has been able to take some actions using data on
incidents of exposure, the data the agency receives may not always be
sufficient and its ability to assess risk and take action based on
such data may be limited.  The reports on incidents that EPA receives
from registrants, as well as some of the voluntary reports such as
those received from states, often vary in detail and lack key
information needed to assess risk.  For example, the reports
frequently lack information on what pesticide caused the incident,
how the exposure occurred, and what symptoms the victim suffered. 
EPA believes this type of information is essential in assessing risks
and thus determining whether the label on a product should be changed
or its use restricted or cancelled. 

Also, EPA cannot be sure that the reports it receives from
registrants and voluntary sources are representative of incidents of
exposure occurring nationwide.  In addition, according to experts
involved in these issues, underreporting of such incidents is
widespread because, among other things, health care professionals may
not always be adequately trained to recognize pesticide poisoning. 


      EPA'S DATA OFTEN LACK
      INFORMATION NECESSARY TO
      ASSESS RISK
---------------------------------------------------------- Letter :5.1

Although pesticide registrants are required to report to EPA any
additional factual information on the unreasonable adverse effects of
their registered pesticides, their incident reports vary in detail. 
Section 6(a)(2) of FIFRA, which requires the registrants to report to
EPA, does not require specific information, and EPA does not require
standardized formats.  An official in EPA's Office of Pesticide
Programs said that registrants interpret FIFRA's reporting
requirements in a variety of ways.  Also, some registrants report
frequently, while others do not. 

In reviewing recent reports received by EPA, we found that some
registrants do not always include important information such as
whether the product was misused or how frequently the victim was
exposed to a pesticide.  For example, one registrant submitted
several reports that identified the pesticide involved and described
the symptoms suffered but did not mention whether the product was
used according to the label's instructions or whether the victim was
exposed to the pesticide once or repeatedly.  EPA believes some
reports may lack important data simply because the data was
unavailable to the registrants, while other reports may exclude data
due to registrant interpretation of reporting requirements. 

The data that the states provide to EPA voluntarily also frequently
lack important information, such as whether the product was misused,
whether the victim was repeatedly exposed to the pesticide, what
symptoms the victim suffered, how the exposure occurred, and--in some
cases--what pesticide caused the incident.  Information on laboratory
tests, which would help confirm the exposure and health effects, is
seldom present.  In reviewing some of the data received by EPA, we
found that although two states, in their 1994 quarterly reports,
summarized the number of pesticide-related incidents, they did not
provide detailed information about the exposures.  One state reported
11 occupational (work-related) pesticide poisonings for the quarter,
of which 3 were confirmed (that is, cause and effect had been
determined), but did not disclose the names of the pesticides
involved or other details of the exposures.  Another state's
quarterly report summarized several incidents of occupational
pesticide poisonings in that state but revealed the name of only one
pesticide.  The report indicated that state agencies were further
investigating some incidents to determine what action should be
taken. 

Although EPA believes that any information about pesticide exposures
can be useful, without some of the significant details about an
incident of exposure EPA is unable to identify trends or patterns
among pesticides that cause problems, assess their potential risks,
or take corrective action.  When the information EPA receives from
the registrants, as well as voluntary sources such as states, does
not have much of the data needed for assessing risk, it is of limited
use. 

In this connection, officials in the Office of Pesticide Programs
emphasized that FIFRA does not mandate that the states have
mechanisms for collecting data on incidents and does not require
states to report incidents to EPA.  The officials also said that
although EPA receives some data from states, the agency does not
depend on the states for reports of incidents. 


      REPORTS TO EPA MAY NOT BE
      REPRESENTATIVE OF INCIDENTS
      OCCURRING NATIONWIDE
---------------------------------------------------------- Letter :5.2

Reports on incidents of exposure that EPA receives from registrants
and from voluntary sources may not be representative of incidents
occurring nationwide.  For example, the nation's poison control
centers typically receive far more reports of exposure than EPA does. 
These centers recorded over 150,000 incidents of humans being exposed
to pesticides in 1992-93.  In contrast, about 12,575\4 incidents of
humans and animals being exposed to pesticides have been reported to
EPA since 1992. 

EPA has sometimes used data from a data base maintained by the
American Association of Poison Control Centers,\5 but the agency has
generally not had funds to routinely pay the fees for such data.  The
association's data base contains considerable amounts of data on
individual exposures, including the type of substance or product,\6

age of the patient, means of exposure, symptoms, and type of
treatment--if any--and the medical outcome.  While the association
publishes summary data annually in the September issue of the
American Journal of Emergency Medicine, it charges a fee for detailed
data.  For example, exposure data on a single poison for 1990-93
would cost $4,400.  Abstracts of individual case records, when
available, are priced at $150. 

As an alternative to purchasing these data directly, however, EPA can
require registrants to purchase the data when the agency determines
that a pesticide poses a high risk to public health.  In 1993, for
example, EPA's Acute Worker Risk Strategy Work Group identified 28
chemicals as acutely toxic to agricultural workers--based on data
from California, data on toxicity, and data on usage.  In this case,
EPA issued a data call-in notice\7 requiring the pesticides'
registrants to submit data from the American Association of Poison
Control Centers.  Using data from California and from the poison
control centers, EPA's worker risk group has proposed measures to
reduce risk for aldicarb, azinphos-methyl, carbofuran, methamidophos,
and methomyl pesticides.\8


--------------------
\4 This figure includes the backlog of about 4,450 incidents of
exposure not yet entered into EPA's Incident Data System.  Also, the
total number of incidents on which EPA receives data may actually be
higher than 12,575 because some summary data involving multiple
similar incidents may be entered into the system and counted as one
incident. 

\5 The American Association of Poison Control Centers, a nationwide
organization of poison centers, maintains a national poisoning
surveillance data base containing information on the volume and
characteristics of incidents of poisoning voluntarily reported to
regional and local centers.  According to an association official,
these centers are usually funded by states and/or community or
university hospitals.  While the centers contribute to an extensive
data base on incidents of exposure, they operate primarily to provide
information on poisons, consultation, and outreach to health
professionals and the general public.  In 1993, 64 centers
contributed to the association's data base, representing portions of
43 states and the District of Columbia and covering about 70 percent
of the incidents of human poisoning reported to poison centers in the
United States. 

\6 Although the association collects data on individual brands, the
association's policy generally prohibits the release of data that
identify brands. 

\7 A formal request for data. 

\8 Based on EPA's review of data on incidents from the poison control
centers and California for 28 pesticides, aldicarb, azinphos-methyl,
methamidophos, mevinphos, and methomyl were determined to be among
pesticides that are most acutely toxic to farm workers. 
Subsequently, in 1994, uses of mevinphos were voluntarily cancelled
because of the high risk of serious poisoning.  Additionally, data
from the poison control centers showed that carbofuran was one of the
agricultural pesticides provoking the highest number of incidents. 


      DATA ARE UNDERREPORTED
---------------------------------------------------------- Letter :5.3

Apart from pesticide registrants, FIFRA does not give EPA authority
to require individuals, states, or organizations to report exposure
to or incidents involving pesticides to EPA.  The voluntary nature of
the data collection system is a major contributor to underreporting
of incidents.  However, underreporting also results from a lack of
training within the medical community in recognizing pesticide
poisonings and lack of familiarity with state reporting requirements. 

In our 1993 report on agricultural pesticides,\9 we reported that
state officials cited underreporting as a serious problem because,
among other reasons, health care professionals lacked adequate
training in recognizing and diagnosing pesticide-related illnesses
and were unfamiliar with state reporting requirements and/or
unwilling to report cases to state officials.  State and federal
officials indicated that even when reports were made, it was
frequently difficult to verify incidents and determine their cause
because of delays in reporting and a lack of information about the
circumstances of these illnesses. 

While these reasons were cited for agricultural pesticides and farm
workers, the same appears to be true for nonagricultural pesticides
and consumers.  For example, an EPA Health Statistician told us that
he believed the medical community's incomplete understanding or
recognition of pesticide poisonings was one reason why the data that
EPA collected on incidents were not sufficient in helping the agency
take the necessary action. 

With respect to health care professionals' familiarity with state
reporting requirements, a toxicologist at the University of
California at Berkeley reported that physicians in California--the
state with the most comprehensive registry of pesticide-related
illnesses in the nation--are often not aware that such illnesses must
be reported to the appropriate local health officers.  According to
the report he coauthored, Preventing Pesticide-related Illness in
California Agriculture,\10

one-quarter of physicians surveyed in rural California did not know
that suspected and confirmed pesticide-related illnesses must be
reported to county health officers. 


--------------------
\9 Pesticides on Farms:  Limited Capability Exists to Monitor
Occupational Illnesses and Injuries (GAO/PEMD-94-6, Dec.  15, 1993). 

\10 William S.  Pease, Rachel A.  Morello-Frosch, David S.  Albright,
Amy D.  Kyle, and James C.  Robinson (Berkeley, California, The
Regents of the University of California:  1993). 


   EPA HAS TAKEN INITIATIVES TO
   IMPROVE COLLECTION OF DATA ON
   INCIDENTS
------------------------------------------------------------ Letter :6

EPA has recognized that its approach to data collection needs
improvement, and in September 1994, its Office of Pesticide Programs
established a work group to focus on potential improvements.  This
work group was established to develop a long-term plan for
collecting, storing, manipulating, and using data on incidents. 

EPA recently completed the first phase of this effort, in which the
work group identified the (1) critical and desirable data elements,
(2) use and potential use of the data collected, (3) current and
potential sources of data, and (4) gaps between the data EPA needs
and the data it already has. 

A second phase--to identify potential improvements in data collection
and analysis--will include identifying (1) how much different system
configurations would cost, (2) who should have access to these
systems, (3) whether one or more data collection systems are needed,
(4) how the agency should be structured internally for the data
collection system, and (5) who should operate the system.  Further
efforts by the work group will include exploring the potential for
more routinely requiring registrants to purchase data from the poison
control centers as part of specific projects. 

A December 1994 report by the work group indicated that additional
phases may also be undertaken.  Although the work group coordinator
said the group plans to establish deadlines for the second phase, as
of May 1995 EPA did not have a formal plan with milestones for
completing any of the phases for this group's work or for
implementing any improvements the work group identified. 

EPA has also proposed a new rule, which it calls the 6(a)(2) rule,
aimed at improving the quality of the data on incidents the agency
receives from pesticide registrants and making the processing of this
information easier for the registrants and the agency.\11 Although
registrants are required under FIFRA to submit any factual data on
adverse effects they may have, EPA is concerned that incidents may be
underreported by the industry as a whole.  The currently available
guidance on reporting on incidents, developed in the 1970s, is not
very detailed.  On the basis of the proposed rule, registrants will
be given specific regulatory requirements on what data they must
report to EPA on incidents of exposure, when such data are available. 
For example, the specific information being requested in the proposed
rule includes the name of the company submitting the information to
EPA, the EPA registration (or identification) number of the pesticide
involved, and a detailed summary including specific information about
the incident being reported.  EPA believes its new rule will clarify
the registrants' responsibilities and should result in significantly
greater numbers of reports on incidents.  EPA expects the new rule to
be finalized in 1995. 

In addition, officials from the Office of Pesticide Programs said
that the office is considering a major reorganization as part of an
effort to streamline operations and that options for managing
information on incidents will be considered as part of this effort. 

Furthermore, EPA staff have been working with four companies that
submit large numbers of reports on incidents of exposure to determine
the feasibility of electronic submission of reports.  Officials in
the Office of Pesticide Programs believe that if the registrants put
the data in a format compatible with the data in the agency's
Incident Data System, staff will be able to enter these data directly
into the system.  The officials also said that they plan to ask these
companies to consider electronically resubmitting reports they had
previously submitted on paper.  Eliminating the need to manually key
these data into the system could help reduce most of the backlog. 
EPA believes this effort is a cost-effective method of improving its
handling of incidents of exposure. 


--------------------
\11 The proposed rule was published in the Federal Register for
comment on September 24, 1992 (57 F.R.  44290).  EPA calls it the
6(a)(2) rule because it would codify EPA's interpretation of section
6(a)(2) of FIFRA. 


   CONCLUSIONS
------------------------------------------------------------ Letter :7

While EPA has a system for collecting, reviewing, and acting on
incidents of exposure to pesticides and has taken action on some data
on incidents, the system does not currently ensure that EPA always
has sufficient information to determine whether action to protect
public health is necessary.  Although EPA has been able to take some
actions using its data on incidents, the agency may not be
appropriately responding to all cases of adverse health effects
caused by pesticide use.  Better, more complete data on incidents
involving pesticides would help EPA determine whether additional
actions are necessary to protect public health. 

EPA has already begun to take some steps to improve its collection
and analysis of data, and its work group is continuing to identify
additional areas for improvements.  We support the agency's efforts
because they should lead to better management of data on incidents. 
Similarly, EPA's proposed 6(a)(2) rule should lead to an improvement
in the quality of data submitted by registrants. 


   AGENCY COMMENTS
------------------------------------------------------------ Letter :8

We requested comments on a draft of this report from EPA.  On June
12, 1995, we met with a section head, Policy and Special Projects
Staff, Office of Pesticide Programs, to obtain the agency's comments
on the draft report.  During this meeting, we were provided with
comments from the Director, Office of Pesticide Programs.  EPA
believes our report accurately explains that EPA regards data on
incidents of exposure as an important supplement to laboratory
studies, and is seeking ways to improve the quality and quantity of
the data submitted to the agency, as well as for improved ways of
managing and using the data in making regulatory decisions. 

EPA believes the draft report did not clearly state the importance of
its proposed 6(a)(2) rule, which is to accomplish two significant
objectives.  First, the rule will explain to registrants exactly what
facts EPA wants them to report.  Secondly, the rule is intended to
solve the perceived problem of underreporting by registrants due to
lack of clear guidance in the form of an enforceable regulation.  The
agency pointed out that the proposed rule does not place new or
additional requirements on registrants, but only clarifies what is
already required under FIFRA.  We agree that the rule is important
for improving the quality of data on incidents. 

EPA was also concerned that in a period of serious resource
constraints, it will be very difficult to make all the improvements
to its collection of data on incidents that would be desirable.  As
noted in our report, acquiring adequately detailed information from
nonregistrant sources can cost substantial amounts of money.  EPA
believes that managing increased numbers of reports will require the
investment of scarce funds and personnel in data management systems. 
In its comments, EPA said that although electronic data submission
and other reporting innovations may help to achieve economies, some
improvements may not be possible at all if resources are cut
significantly in the future. 

EPA also provided some technical comments, and we have made changes
in appropriate sections of our report to accommodate these comments. 


   OBJECTIVES, SCOPE, AND
   METHODOLOGY
------------------------------------------------------------ Letter :9

Our objectives were to determine whether EPA collects data on
incidents of exposure to pesticides and takes action based on these
data, and whether such data are sufficient to allow the agency to
determine if unacceptable risks to public health are occurring.  To
accomplish these objectives, we interviewed officials from EPA's
Office of Pesticide Programs, including the Chief, Special Projects
and Coordination; Incident Data Officer for Humans and Domestic
Animals; Coordinator, Ecological Incident Monitoring; Chief,
Certification and Training Branch; and Section Head of Special Review
and Groundwater.  We also reviewed documents and records from EPA's
Incident Data System. 

To obtain views on incidents of pesticide exposure from others
outside of EPA, we discussed the adverse health effects of
nonagricultural pesticides with representatives of industry and of
environmental and other nonprofit organizations.  In addition, we
visited California, Florida, and Oregon, and collected and reviewed
these states' data on incidents of exposure.  We selected these
states because they collect data on such incidents and because two of
these states--California and Florida--have climates in which a
greater use of nonagricultural pesticides is likely to be required. 

We conducted our review between March 1994 and May 1995 in accordance
with generally accepted government auditing standards. 


---------------------------------------------------------- Letter :9.1

As arranged with your office, we plan no further distribution of this
report until 10 days after the date of this letter unless you
publicly announce its contents earlier.  We will then send copies to
the Administrator of EPA.  We will also make copies available to
others on request. 

Please call me at (202) 512-4907 if you or your staff have any
questions.  Major contributors to this report are listed in appendix
II. 

Peter F.  Guerrero
Director, Environmental
 Protection Issues


ACTIONS EPA HAS TAKEN ON
NONAGRICULTURAL PESTICIDES
=========================================================== Appendix I

While EPA does not routinely receive complete data on incidents
involving nonagricultural pesticides, it sometimes receives
information on specific cases that is detailed enough to assist it in
taking actions to protect public health.  Table I.1 lists examples of
EPA's use of such data to take actions between 1989 and 1994. 



                                    Table I.1
                     
                      EPA's Actions to Protect Public Health

                           Data collected, used,
                           and/or
Pesticide                  analyzed by EPA            Action taken
-------------------------  -------------------------  --------------------------
Chlorine                   EPA reviewed data from     EPA restricted the use of
                           hospitals' emergency       chlorine in swimming
                           rooms, newspaper           pools.
                           clippings generated by
                           manufacturers, and field
                           information from state
                           agencies to identify the
                           types and severity of
                           poisonings that could
                           result from the use of
                           chlorine in swimming
                           pools.

Lawn care products         Through an increase in     EPA developed guidance for
                           the number of incidents    the states on how to
                           reported by the National   establish posting and
                           Pesticide                  notification programs for
                           Telecommunications         lawn care products.
                           Network,\a EPA identified
                           a public perception of
                           risk from lawn care
                           pesticides.

Pet care products with     Through its Incident Data  EPA has completed a
adverse effects on human   System, EPA identified a   Pesticide Registration
and animal health          large number of pets       Notice instructing
                           being adversely affected   registrants to clarify
                           by consumers' misuse of    warnings and instructions
                           these products. The data   on the products' labels to
                           also revealed that human   prevent misuse by
                           health was being           consumers.
                           adversely affected.

Personal-use insect        Using information          EPA distributed a
repellent                  collected from EPA's       physician's advisory
                           regional offices and from  through the Centers for
                           state agencies, EPA found  Disease Control and poison
                           cases in which certain     centers as well as a
                           insect repellents were     consumer brochure on
                           causing adverse            proper use.
                           reactions.

Mercury\b                  On the basis of (1)        EPA canceled all uses of
                           reports on a child with    mercury in household
                           acrodynia,\c (2) over 40   paints.
                           publications on the
                           relationship between that
                           disease and mercury, and
                           (3) levels of mercury
                           that the Centers for
                           Disease Control found in
                           household air and
                           occupants' urine in
                           Detroit homes, EPA
                           assessed the risk of
                           acrodynia resulting from
                           the use of mercury in
                           household paint.

Arsenical ant bait         EPA used data from         EPA canceled most uses of
                           hospitals' emergency       sodium arsenate in
                           rooms, hospitals, a        household ant bait.
                           poison control center,
                           and the state of Texas to
                           determine that this
                           pesticide product had a
                           small margin of safety
                           for young children.

Disulfoton                 A parent informed EPA of   EPA required the
                           an incident involving a    manufacturer to retest the
                           child who overcame a       product's child-resistant
                           child-resistant package    packaging for efficacy.
                           containing 2 percent
                           disulfoton powder (a
                           pesticide used on
                           ornamental plants and
                           house plants).

Methyl bromide             EPA learned of an          EPA required revisions to
                           investigation of two       the pesticide's label
                           cases (one in California   requiring longer
                           and one in Iowa) in which  ventilation periods before
                           two people died after      people reentered treated
                           reentering structures      structures.
                           treated with methyl
                           bromide.

Boric acid                 Data reviewed by a poison  EPA required revisions to
                           control center permitted   the product's label to
                           EPA to determine how much  restrict the number of
                           boric acid powder or how   tablets used in one
                           many tablets resulted in   application of the
                           poisonings of children.    product.
--------------------------------------------------------------------------------
\a The National Pesticide Telecommunications Network is a toll-free
telephone hotline funded by EPA to provide information on pesticide
use, toxicology, health effects, and safety to the general public and
professional communities (including physicians, veterinarians, poison
control centers, attorneys, and commercial pesticide applicators). 
Subcontractors linked by telephone to the network also provide
medical emergency services for humans and domestic animals. 

\b Mercury was added to paints to preserve the paint in the can by
controlling the growth of microbes, principally bacteria, and to
preserve the paint from mildew attack after it was applied to an
exterior surface. 

\c Acrodynia is a rare form of childhood mercury poisoning. 


MAJOR CONTRIBUTORS TO THIS REPORT
========================================================== Appendix II

RESOURCES, COMMUNITY, AND ECONOMIC
DEVELOPMENT DIVISION WASHINGTON,
D.C. 

Lawrence J.  Dyckman, Associate Director
J.  Kevin Donohue, Assistant Director
Raymond M.  Ridgeway, Evaluator-In-Charge
Jennifer W.  Clayborne, Evaluator
Phyllis Turner, Communications Analyst

