Food Safety
FDA Needs to Reassess Its Approach to Reducing an Illness Caused by Eating Raw Oysters
Gao ID: GAO-11-607 September 8, 2011
Vibrio vulnificus (V. vulnificus) is a bacterium that occurs naturally in the Gulf of Mexico. On average, since 2000, about 32 individuals a year in the United States have become ill from eating raw or undercooked oysters containing V. vulnificus, and about half have died. The Food and Drug Administration (FDA) is responsible for ensuring oyster safety and works with the Interstate Shellfish Sanitation Conference (ISSC), which includes representatives from FDA, states, and the shellfish industry to establish guidelines for sanitary control of the shellfish industry. GAO was asked to determine the extent to which FDA and the ISSC agree on the V. vulnificus illness reduction goal, use a credible approach to measure progress toward the illness rate reduction goal, have evaluated the effectiveness of their actions in reducing V. vulnificus illnesses, and whether the Gulf Coast oyster industry has adequate capacity to postharvest process oysters harvested April through October. GAO reviewed data and documents and interviewed officials in FDA, the ISSC, Florida, Louisiana, and Texas.
FDA and the ISSC do not agree on a common V. vulnificus illness reduction goal. In October 2009, FDA announced its intention to change its approach to V. vulnificus illnesses from reducing them to largely eliminating them. To do so, FDA would require states to use postharvest processing methods, which include a mild heat treatment known as low temperature pasteurization. FDA's announced approach was a change from the 60 percent illness rate reduction goal established by the ISSC in 2001, with FDA concurrence. In a November 2009 letter to FDA, the ISSC expressed disappointment that FDA had not followed a 1984 memorandum of understanding that calls for FDA and the ISSC to consult on such matters. If FDA and the ISSC are not in agreement on the illness reduction goal and strategies to achieve it, it will be difficult for the Gulf Coast states to move forward to significantly reduce the number of consumption-related V. vulnificus illnesses. The approach FDA and the ISSC have been using to measure progress toward the previously agreed upon V. vulnificus illness rate reduction goal established in 2001 has limitations that undermine its credibility. For example, the ISSC continues to include California's results in its illness rate reduction calculation along with Florida, Louisiana, and Texas. Doing so overstates the effectiveness of consumer education and time and temperature controls--FDA's and the ISSC's primary strategies for reducing V. vulnificus illnesses--because California, unlike these other states, requires that all raw Gulf Coast oysters harvested during the summer and sold in the state be processed to reduce V. vulnificus to nondetectable levels, which has reduced V. vulnificus illnesses to nearly zero. FDA and the ISSC have taken few steps to evaluate the effectiveness of their consumer education efforts since 2004. Likewise, they have not directly evaluated the effectiveness of the time and temperature controls implemented in 2010, which call for harvesters to ensure that oysters are cooled to specific temperatures within certain times to reduce V. vulnificus growth. Although data are not available, our discussions with state and oyster industry officials suggest 100 percent compliance with the controls is highly unlikely. Moreover, our analysis shows--even assuming 80 percent compliance in the summer months-- it is unlikely that these controls will lead to the level of illness reduction estimated by a model developed by FDA. The Gulf Coast oyster industry does not have sufficient capacity to process all of its oysters intended for raw consumption that are harvested from April through October to reduce V. vulnificus to nondetectable levels, according to an FDA-commissioned report. The report concluded that it will take a minimum of 2 to 3 years to develop the infrastructure needed to process these oysters. However, the report has some limitations that call into question the completeness of its cost and timeline estimates. For example, the report's cost estimates did not include some construction costs and costs associated with purchasing land needed to expand existing processing facilities or build new ones. Without this information, the full cost of developing sufficient processing capacity will not be known. GAO recommends that FDA work with the ISSC to agree on an illness reduction goal, improve its approach to measuring progress in reducing V. vulnificus illnesses, regularly evaluate its illness reduction strategies, and address the limitations in the FDA-commissioned report. FDA and the ISSC generally agreed with our recommendations.
Recommendations
Our recommendations from this work are listed below with a Contact for more information. Status will change from "In process" to "Open," "Closed - implemented," or "Closed - not implemented" based on our follow up work.
Director:
Lisa R. Shames
Team:
Government Accountability Office: Natural Resources and Environment
Phone:
(202) 512-2649
GAO-11-607, Food Safety: FDA Needs to Reassess Its Approach to Reducing an Illness Caused by Eating Raw Oysters
This is the accessible text file for GAO report number GAO-11-607
entitled 'Food Safety: FDA Needs to Reassess Its Approach to Reducing
an Illness Caused by Eating Raw Oysters' which was released on October
11, 2011.
This text file was formatted by the U.S. Government Accountability
Office (GAO) to be accessible to users with visual impairments, as
part of a longer term project to improve GAO products' accessibility.
Every attempt has been made to maintain the structural and data
integrity of the original printed product. Accessibility features,
such as text descriptions of tables, consecutively numbered footnotes
placed at the end of the file, and the text of agency comment letters,
are provided but may not exactly duplicate the presentation or format
of the printed version. The portable document format (PDF) file is an
exact electronic replica of the printed version. We welcome your
feedback. Please E-mail your comments regarding the contents or
accessibility features of this document to Webmaster@gao.gov.
This is a work of the U.S. government and is not subject to copyright
protection in the United States. It may be reproduced and distributed
in its entirety without further permission from GAO. Because this work
may contain copyrighted images or other material, permission from the
copyright holder may be necessary if you wish to reproduce this
material separately.
United States Government Accountability Office:
GAO:
Report to the Honorable Rosa L. DeLauro, House of Representatives:
September 2011:
Food Safety:
FDA Needs to Reassess Its Approach to Reducing an Illness Caused by
Eating Raw Oysters:
GAO-11-607:
GAO Highlights:
Highlights of GAO-11-607, a report to the Honorable Rosa L. DeLauro,
House of Representatives.
Why GAO Did This Study:
Vibrio vulnificus (V. vulnificus) is a bacterium that occurs naturally
in the Gulf of Mexico. On average, since 2000, about 32 individuals a
year in the United States have become ill from eating raw or
undercooked oysters containing V. vulnificus, and about half have
died. The Food and Drug Administration (FDA) is responsible for
ensuring oyster safety and works with the Interstate Shellfish
Sanitation Conference (ISSC), which includes representatives from FDA,
states, and the shellfish industry to establish guidelines for
sanitary control of the shellfish industry. GAO was asked to determine
the extent to which FDA and the ISSC agree on the V. vulnificus
illness reduction goal, use a credible approach to measure progress
toward the illness rate reduction goal, have evaluated the
effectiveness of their actions in reducing V. vulnificus illnesses,
and whether the Gulf Coast oyster industry has adequate capacity to
postharvest process oysters harvested April through October. GAO
reviewed data and documents and interviewed officials in FDA, the
ISSC, Florida, Louisiana, and Texas.
What GAO Found:
FDA and the ISSC do not agree on a common V. vulnificus illness
reduction goal. In October 2009, FDA announced its intention to change
its approach to V. vulnificus illnesses from reducing them to largely
eliminating them. To do so, FDA would require states to use
postharvest processing methods, which include a mild heat treatment
known as low temperature pasteurization. FDA‘s announced approach was
a change from the 60 percent illness rate reduction goal established
by the ISSC in 2001, with FDA concurrence. In a November 2009 letter
to FDA, the ISSC expressed disappointment that FDA had not followed a
1984 memorandum of understanding that calls for FDA and the ISSC to
consult on such matters. If FDA and the ISSC are not in agreement on
the illness reduction goal and strategies to achieve it, it will be
difficult for the Gulf Coast states to move forward to significantly
reduce the number of consumption-related V. vulnificus illnesses.
The approach FDA and the ISSC have been using to measure progress
toward the previously agreed upon V. vulnificus illness rate reduction
goal established in 2001 has limitations that undermine its
credibility. For example, the ISSC continues to include California‘s
results in its illness rate reduction calculation along with Florida,
Louisiana, and Texas. Doing so overstates the effectiveness of
consumer education and time and temperature controls-”FDA‘s and the
ISSC‘s primary strategies for reducing V. vulnificus illnesses-”
because California, unlike these other states, requires that all raw
Gulf Coast oysters harvested during the summer and sold in the state
be processed to reduce V. vulnificus to nondetectable levels, which
has reduced V. vulnificus illnesses to nearly zero.
FDA and the ISSC have taken few steps to evaluate the effectiveness of
their consumer education efforts since 2004. Likewise, they have not
directly evaluated the effectiveness of the time and temperature
controls implemented in 2010, which call for harvesters to ensure that
oysters are cooled to specific temperatures within certain times to
reduce V. vulnificus growth. Although data are not available, our
discussions with state and oyster industry officials suggest 100
percent compliance with the controls is highly unlikely. Moreover, our
analysis shows-”even assuming 80 percent compliance in the summer
months”-it is unlikely that these controls will lead to the level of
illness reduction estimated by a model developed by FDA.
The Gulf Coast oyster industry does not have sufficient capacity to
process all of its oysters intended for raw consumption that are
harvested from April through October to reduce V. vulnificus to
nondetectable levels, according to an FDA-commissioned report. The
report concluded that it will take a minimum of 2 to 3 years to
develop the infrastructure needed to process these oysters. However,
the report has some limitations that call into question the
completeness of its cost and timeline estimates. For example, the
report‘s cost estimates did not include some construction costs and
costs associated with purchasing land needed to expand existing
processing facilities or build new ones. Without this information, the
full cost of developing sufficient processing capacity will not be
known.
What GAO Recommends:
GAO recommends that FDA work with the ISSC to agree on an illness
reduction goal, improve its approach to measuring progress in reducing
V. vulnificus illnesses, regularly evaluate its illness reduction
strategies, and address the limitations in the FDA-commissioned
report. FDA and the ISSC generally agreed with our recommendations.
View [hyperlink, http://www.gao.gov/products/GAO-11-607] or key
components. For more information, contact Lisa Shames at (202) 512-
3841 or shamesl@gao.gov.
[End of section]
Contents:
Letter:
Background:
FDA and the ISSC Do Not Currently Agree on a V. Vulnificus Illness
Reduction Goal:
The Approach FDA and the ISSC Use for Measuring Progress toward Their
Illness Rate Reduction Goal Has Limitations That Undermine Its
Credibility:
FDA and the ISSC Have Taken Few Steps to Evaluate the Effectiveness of
Consumer Education and Time and Temperature Controls:
Extent of Industry Compliance with Time and Temperature Controls Is
Unknown:
Extent of Illness Reduction from Time and Temperature Controls Depends
on Compliance Levels:
Adequate Capacity Does Not Exist to Use Postharvest Processing on Gulf
Coast Oysters, and Questions Exist about the Feasibility of Developing
Such Capacity:
Conclusions:
Recommendations for Executive Action:
Agency Comments and Our Evaluation:
Appendix I: Estimating the Impact of Time and Temperature Controls on
the Number of Illnesses from V. Vulnificus:
Appendix II: Comments from the Department of Health and Human Services:
Appendix III: Comments from the Interstate Shellfish Sanitation
Conference:
Appendix IV: GAO Contact and Staff Acknowledgments:
Tables:
Table 1: Time and Temperature Controls Implemented by Florida,
Louisiana, and Texas Beginning May 1, 2010:
Table 2: Comparison of Estimates between WHO/FAO Risk Simulation Model
and GAO Replication of the Model for July, August, and September:
Table 3: Time and Temperature Controls for 2010, by State and Month:
Table 4: Comparison of FDA's Risk Calculator and GAO's Modification of
WHO/FAO Risk Simulation Model Concerning Estimated Number of V.
Vulnificus Illnesses per 100,000 Servings of Oysters at Baseline:
Table 5: Estimated Percentage Reduction in the Number of V. Vulnificus
Illnesses as a Result of Time and Temperature Controls, by State and
Month:
Table 6: Probability That the Time and Temperature Controls Will
Reduce V. Vulnificus Illnesses to the Number Estimated by FDA's Risk
Calculator or Lower, by Estimated Compliance Rate:
Figures:
Figure 1: Gulf Coast States and Oyster Harvest Areas:
Figure 2: Amount and Value of Oysters Harvested by Gulf Coast States
in 2009:
Figure 3: Total Number of V. Vulnificus Oyster Consumption-Related
Illnesses Reported Nationally, by month, from 2000-2010:
Figure 4: Number of V. Vulnificus Oyster Consumption-Related Illnesses
Reported Nationally, by Year, from 2000-2010:
Figure 5: Sample Log Sheet:
Figure 6: Sample Oyster Labels:
Figure 7: Estimated Reduction in the Number of V. Vulnificus Illnesses
as a Result of Time and Temperature Controls for Louisiana in August:
Figure 8: Key Components of WHO/FAO Risk Simulation Model and GAO
Modifications:
Abbreviations:
CDC: Centers for Disease Control and Prevention:
FDA Food and Drug Administration:
ISSC: Interstate Shellfish Sanitation Conference:
NOAA: National Oceanic and Atmospheric Administration:
RTI: Research Triangle Institute International:
V. vulnificus: Vibrio vulnificus:
WHO/FAO: World Health Organization and the Food and Agriculture
Organization:
[End of section]
United States Government Accountability Office:
Washington, DC 20548:
September 8, 2011:
The Honorable Rosa L. DeLauro:
House of Representatives:
Dear Ms. DeLauro:
Illnesses due to Vibrio vulnificus (V. vulnificus), a naturally
occurring bacterium that is commonly found in oysters harvested from
the Gulf of Mexico and that grows quickly--especially during warmer
months (April through November)--are the most common cause of death
from seafood consumption in the United States. Individuals with
compromised immune systems can develop a severe and potentially fatal
infection from eating raw or undercooked oysters contaminated with V.
vulnificus. According to data provided by the Interstate Shellfish
Sanitation Conference (ISSC)--a voluntary organization formed in 1982
by state officials from 22 states to promote uniform national
shellfish safety policies--the number of V. vulnificus illnesses
associated with raw oyster consumption has averaged about 32 a year
nationwide since 2000. V. vulnificus illnesses are fatal about 50
percent of the time, according to the Centers for Disease Control and
Prevention (CDC).
Under ISSC procedures, FDA, the federal agency responsible for
ensuring the safety of oysters and other shellfish, must concur with
the ISSC's proposed shellfish safety policies before they can be
incorporated into the National Shellfish Sanitation Program's
shellfish safety guidelines.[Footnote 1] In 2001, the ISSC approved,
with FDA concurrence, a change in the shellfish safety guidelines
providing that individual states are to develop V. vulnificus risk
management plans if two or more confirmed V. vulnificus illnesses
since 1995 could be traced to the consumption of commercially
harvested raw or undercooked oysters that originated from the state's
waters. Initially, four states--Alabama, Florida, Louisiana, and
Texas--exceeded that threshold and therefore, in accordance with the
shellfish safety guidelines, were expected to develop such plans.
[Footnote 2] The states' risk management plans relied primarily on
consumer education aimed at high-risk (i.e., immune-compromised)
individuals to reduce the rate of V. vulnificus illness and time and
temperature controls to cool oysters to a specific temperature within
a certain period after harvesting to reduce the growth of V.
vulnificus.
The shellfish safety guidelines also included goals for reducing the
rate of illness for four reporting states[Footnote 3]--California,
Florida, Louisiana, and Texas; for those states, the guidelines
specified illness rate reduction goals of 40 percent by the end of
2006 and 60 percent by the end of 2008.[Footnote 4] If the 60 percent
goal was not achieved, the Gulf Coast states (Alabama, Florida,
Louisiana, Mississippi, and Texas) were to implement one or more
illness reduction strategies identified in the guidelines--postharvest
processing methods (such as a mild heat treatment known as low
temperature pasteurization) to reduce V. vulnificus to nondetectable
levels, closing oyster harvest areas, or removing oysters from their
shells (i.e., shucking) for cooking prior to consumption--or
"equivalent" strategies. According to FDA and the ISSC, however, by
the end of 2008, the four reporting states achieved about a 35 percent
V. vulnificus illness rate reduction and therefore missed the 60
percent goal by about 25 percent. Instead of imposing the illness
reduction strategies specified in the guidelines if the illness rate
reduction goals were not met, in May 2009, the ISSC approved, with FDA
concurrence, new more stringent time and temperature controls. These
controls were intended to help achieve the approximately 25 percent
illness rate reduction needed to meet the 60 percent goal by the end
of 2010.
By October 2009, FDA announced its intent to change its approach to V.
vulnificus illnesses from reducing them to largely eliminating them.
Specifically, FDA said that, in spite of the ISSC's efforts, the
number of V. vulnificus illnesses had not significantly declined and
strategies that fall well short of eliminating V. vulnificus were no
longer sufficient. FDA has also raised concerns with the current
approach for measuring V. vulnificus illness reductions. Furthermore,
in an October 2009 letter to the ISSC, FDA went on to say that
academia, the oyster industry, and government, with the support of the
ISSC, had developed postharvest processing technologies that could
largely eliminate V. vulnificus illnesses and that the Gulf Coast
oyster industry has the capacity to use postharvest processing on 100
percent of Gulf Coast oysters intended for raw consumption.
In your role during the 111th Congress as the Chairwoman of the House
Appropriations Subcommittee on Agriculture, Rural Development, Food
and Drug Administration, and Related Agencies, you asked us to review
efforts of FDA and the ISSC to reduce illnesses caused by oysters
contaminated with V. vulnificus bacteria. Accordingly, this report
examines the extent to which (1) FDA and the ISSC are currently in
agreement on a V. vulnificus illness reduction goal, (2) FDA and the
ISSC use a credible approach to measure progress toward their V.
vulnificus illness rate reduction goal, (3) FDA and the ISSC have
evaluated the effectiveness of consumer education programs and time
and temperature controls in reducing V. vulnificus consumption-related
illnesses, and (4) the Gulf Coast oyster industry has adequate
capacity to use postharvest processing on oysters harvested during
warmer months and intended for raw consumption.
To determine the extent to which FDA and the ISSC are in agreement on
a V. vulnificus illness reduction goal, we reviewed the ISSC's meeting
minutes and FDA's responses to ISSC proposals regarding illness
reduction goals. To determine the extent to which FDA and the ISSC
have used a credible approach to measure progress toward their V.
vulnificus illness rate reduction goals, we analyzed the number of
states used in determining V. vulnificus illness rate reduction, the
effectiveness of primary V. vulnificus illness rate reduction
strategies, and the effect of such factors as natural and manmade
disasters on V. vulnificus illness rate reduction. To determine the
extent to which FDA and the ISSC have evaluated the effectiveness of
consumer education programs in reducing V. vulnificus consumption-
related illnesses, we reviewed relevant ISSC-commissioned surveys and
studies and the V. vulnificus consumer education activities of
Florida, Louisiana, and Texas. We also interviewed officials from
these states' health agencies and members of the ISSC's Vibrio
education committee. To determine the extent to which FDA and the ISSC
have evaluated the effectiveness of time and temperature controls in
reducing V. vulnificus consumption-related illnesses, we reviewed
scientific literature and reports. We also analyzed the data in the
model that was the basis for a tool that FDA developed for Florida,
Louisiana, and Texas to use in determining their time and temperature
controls. We also replicated the model to determine its validity and
then modified it to simulate the impact of the time and temperature
controls implemented by Florida, Louisiana, and Texas in May 2010 (see
appendix I). In addition, we interviewed a nonprobability sample of 11
leading Vibro researchers about the scientific underpinnings of time
and temperature controls[Footnote 5] and law enforcement officials
from agencies in Florida, Louisiana, and Texas regarding their time
and temperature enforcement activities. To determine the extent to
which there is adequate capacity to use postharvest processing on Gulf
Coast oysters harvested during warmer months and intended for raw
consumption, we analyzed the 2011 FDA-commissioned Research Triangle
Institute International (RTI) report that addressed the feasibility
and economic impacts of requiring postharvest processing.[Footnote 6]
We also interviewed the lead author of the RTI report. To address all
four objectives, we interviewed knowledgeable officials from FDA, the
ISSC, and state regulatory agencies in Florida, Louisiana, and Texas,
as well as representatives from the Gulf Coast oyster industry. We
also conducted site visits to Florida, Louisiana, and Texas and
attended key ISSC meetings in Alabama and Florida.
We conducted this performance audit from May 2010 to September 2011 in
accordance with generally accepted government auditing standards.
Those standards require that we plan and perform the audit to obtain
sufficient, appropriate evidence to provide a reasonable basis for our
findings and conclusions based on our audit objectives. We believe
that the evidence obtained provides a reasonable basis for our
findings and conclusions based on our audit objectives.
Background:
Gulf Coast oysters are commercially harvested from the waters of the
Gulf of Mexico adjacent to Alabama, Florida, Louisiana, Mississippi,
and Texas and shipped throughout the United States. Figure 1 shows the
Gulf Coast states and the location of the primary oyster harvest areas
in the Gulf of Mexico.
Figure 1: Gulf Coast States and Oyster Harvest Areas:
[Refer to PDF for image: illustrated U.S. map]
The map highlights the locations of oyster harvest areas in the states
of Alabama, Florida, Louisiana; Mississippi, and Texas.
Source: RTI.
[End of figure]
According to statistics from the Department of Commerce's National
Oceanic and Atmospheric Administration (NOAA), in 2009, the Gulf Coast
region produced about 23 million pounds of oysters, approximately 63
percent of the nation's total domestic production, valued at about $72
million. Figure 2 shows the amount and value of oysters harvested by
Gulf Coast states in 2009, the most recent year for which these data
are available.
Figure 2: Amount and Value of Oysters Harvested by Gulf Coast States
in 2009:
[Refer to PDF for image: 2 vertical bar graphs]
Amount harvested:
State: Louisiana;
Amount harvested: 14.7 million pounds.
State: Texas;
Amount harvested: 2.7 million pounds.
State: Mississippi;
Amount harvested: 2.2 million pounds.
State: Florida (west coast);
Amount harvested: 2.9 million pounds.
State: Alabama;
Amount harvested: 0.023 million pounds.
Total Amount harvested: 23 million pounds.
Value:
State: Louisiana;
Value: $50 million.
State: Texas;
Value: $9.4 million.
State: Mississippi;
Value: $6.1 million.
State: Florida (west coast);
Value: $6.9 million.
State: Alabama;
Value: $0.077 million.
Total Value: $72 million.
Source: NOAA.
Note: The amounts and value of oysters harvested are rounded.
[End of figure]
Because V. vulnificus is more abundant in oysters harvested during the
warmer-weather months (April through November), consumers who eat raw
oysters harvested during this period are likely to be exposed to
greater amounts of V. vulnificus. Although most healthy people do not
become ill from V. vulnificus, people with certain medical conditions--
such as chronic liver disease, hemochromatosis,[Footnote 7] cancer,
kidney disease, diabetes, and human immunodeficiency virus/acquired
immune deficiency syndrome--are at risk of developing a potentially
fatal bloodstream infection known as septicemia, which is
characterized by fever and chills, life-threatening low blood
pressure, and blistering skin lesions. Figure 3 shows that V.
vulnificus consumption-related illnesses peak during April through
November and remain quite low from December through March.
Figure 3: Total Number of V. Vulnificus Oyster Consumption-Related
Illnesses Reported Nationally, by month, from 2000-2010:
[Refer to PDF for image: line graph]
Illnesses peak during warmer months, May through October.
Month: January;
Number of reported illnesses: 0.
Month: February;
Number of reported illnesses: 3.
Month: March;
Number of reported illnesses: 4.
Month: April;
Number of reported illnesses: 38.
Month: May;
Number of reported illnesses: 33.
Month: June;
Number of reported illnesses: 42.
Month: July;
Number of reported illnesses: 58.
Month: August;
Number of reported illnesses: 68.
Month: September;
Number of reported illnesses: 41.
Month: October;
Number of reported illnesses: 38.
Month: November;
Number of reported illnesses: 40.
Month: December;
Number of reported illnesses: 2.
Source: GAO analysis, including preliminary 2010 ISSC data.
[End of figure]
According to the ISSC's data, since 2000, 348 V. vulnificus
consumption-related illnesses have been reported nationally. As figure
4 shows, the number of V. vulnificus consumption-related illnesses
reported nationally from 2000 to 2010 have been relatively consistent
annually--with the exception of 2005 and 2010, when Hurricane Katrina
and the Deepwater Horizon oil spill, respectively, severely reduced
the oyster harvest.
Figure 4: Number of V. Vulnificus Oyster Consumption-Related Illnesses
Reported Nationally, by Year, from 2000-2010:
[Refer to PDF for image: line graph]
Calendar year: 2000;
Number of yearly illnesses: 30.
Calendar year: 2001;
Number of yearly illnesses: 40.
Calendar year: 2002;
Number of yearly illnesses: 35.
Calendar year: 2003;
Number of yearly illnesses: 32.
Calendar year: 2004;
Number of yearly illnesses: 32.
Calendar year: 2005[A];
Number of yearly illnesses: 13.
Calendar year: 2006;
Number of yearly illnesses: 39.
Calendar year: 2007;
Number of yearly illnesses: 34.
Calendar year: 2008;
Number of yearly illnesses: 30.
Calendar year: 2009;
Number of yearly illnesses: 40.
Calendar year: 2010[B];
Number of yearly illnesses: 23.
Source: GAO analysis, including preliminary 2010 ISSC data.
[A] Hurricane Katrina struck the Gulf Coast region in August 2005,
causing significant damage to the oyster industry. Oyster beds and
vessels along the Gulf Coast were extensively damaged, if not
destroyed, by siltation and contamination related to Hurricane Katrina.
[B] The Deepwater Horizon oil spill in the Gulf of Mexico occurred in
April 2010, causing significant damage to the oyster industry. The oil
spill resulted in numerous harvest area closures and significant death
of oysters from freshwater diversions that were used to prevent oil
from reaching shorelines.
[End of figure]
Although the number of V. vulnificus consumption-related illnesses is
small, the costs of the disease are high because of the high mortality
rate--about 50 percent, according to CDC--costing the nation about
$124 million annually, according to FDA. However, a senior ISSC
official said FDA's estimate overstates the annual costs related to V.
vulnificus consumption-related illnesses because it does not factor in
the age and pre-existing health condition of the victims.
As the federal agency responsible for ensuring the safety of
shellfish, including oysters, in March 1984, FDA entered into a
memorandum of understanding with the ISSC recognizing it as the
primary voluntary national organization of state shellfish regulatory
officials that provides guidance and counsel on matters related to the
sanitary control of shellfish. The ISSC provides a formal structure
for state regulatory authorities to establish guidelines, and
procedures for applying those guidelines, for the sanitary control of
the oyster industry. These guidelines must be reviewed by FDA for
consistency with existing laws, regulations, and policies before they
can be adopted. In addition to FDA and state regulatory officials, the
ISSC also includes members from the shellfish industry and other
federal agencies.
Postharvest processing, closing oyster harvest areas, and shucking can
all be expected to either substantially reduce or essentially
eliminate exposure to V. vulnificus bacteria by consumers of raw
oysters.[Footnote 8] However, when the 60 percent illness rate
reduction goal was not met by the end of 2008, instead of implementing
these strategies, FDA and the ISSC relied on estimates generated by
FDA's V. vulnificus risk calculator in adopting time and temperature
controls that they considered to be an equivalent strategy.[Footnote
9] Senior FDA and ISSC officials told us that although time and
temperature controls are not equivalent to the other strategies in the
guidelines in terms of the total amount of illness reduction each can
achieve, they considered the new time and temperature controls to be
equivalent in that, according to the risk calculator's estimations,
they would equivalently help the states to achieve the approximately
25 percent illness rate reduction needed to meet the 60 percent goal
by the end of 2010. Table 1 shows the time and temperature controls
implemented in Florida, Louisiana, and Texas on May 1, 2010.
Table 1: Time and Temperature Controls Implemented by Florida,
Louisiana, and Texas Beginning May 1, 2010:
State: Florida;
Period controls were in effect: May through October;
Maximum time from harvest to refrigeration (hours): 6[A];
Maximum time from refrigeration to internal oyster meat temperature of
55°F (hours): 2[A].
State: Louisiana;
Period controls were in effect: May through October;
Maximum time from harvest to refrigeration (hours): 1;
Maximum time from refrigeration to internal oyster meat temperature of
55°F (hours): 6.
State: Texas;
Period controls were in effect: May and October;
Maximum time from harvest to refrigeration (hours): 4;
Maximum time from refrigeration to internal oyster meat temperature of
55°F (hours): 6.
State: Texas;
Period controls were in effect: June and September;
Maximum time from harvest to refrigeration (hours): 3;
Maximum time from refrigeration to internal oyster meat temperature of
55°F (hours): 6.
State: Texas;
Period controls were in effect: July and August;
Maximum time from harvest to refrigeration (hours): 1;
Maximum time from refrigeration to internal oyster meat temperature of
55°F (hours): 6.
Source: GAO analysis of Florida, Louisiana, and Texas 2010 V.
vulnificus risk management plans.
[A] This is the rapid cooling control option, one of four time and
temperature control options authorized in Florida's V. vulnificus risk
management plan and the only option used during the summer of 2010 by
the Florida oyster industry, according to a senior Florida regulatory
official.
[End of table]
Although FDA had concurred with the use of new time and temperature
controls earlier in 2009, in October of that year, a senior FDA
official stated that the agency would require postharvest processing
to reduce V. vulnificus to nondetectable levels. There are currently
four methods for processing oysters after they have been harvested to
reduce V. vulnificus to nondetectable levels: (1) high-pressure
processing, (2) a mild heat treatment known as cool pasteurization,
(3) cryogenic quick freezing, and (4) irradiation. Each of these
processes--except irradiation--is currently in limited, voluntary
commercial use in the Gulf Coast region. The senior FDA official
indicated that the postharvest processing requirement would apply to
all Gulf Coast oysters harvested during the warmer months of the year
beginning with the 2011 harvest season. However, in response to
concerns expressed by some members of Congress and the ISSC, among
others, FDA suspended its plan to require postharvest processing until
a study was done to determine how postharvest processing can be
implemented in the fastest, safest, and most economical way. In 2010,
FDA contracted with RTI to study the feasibility and economic impacts
of requiring postharvest processing of Gulf state (Alabama, Florida
Louisiana, Mississippi, and Texas) oysters harvested from April
through October and intended for raw consumption.[Footnote 10]
FDA and the ISSC Do Not Currently Agree on a V. Vulnificus Illness
Reduction Goal:
In October 2009, a senior FDA official announced in a speech before
the ISSC that, under FDA's Hazard Analysis and Critical Control Point
rules, beginning in May 2011, FDA intended to require postharvest
processing of all Gulf Coast oysters harvested during warmer months,
when higher levels of V. vulnificus are more likely to be present, to
reduce V. vulnificus to nondetectable levels.[Footnote 11]
According to FDA officials, the agency took this action for two
primary reasons. First, consumer education activities and time and
temperature controls, which had been in use by Louisiana, Florida, and
Texas since 2001, had not achieved the 60 percent goal by the 2008
deadline.[Footnote 12] Second, validated methods of postharvest
processing technology had become available. FDA noted that in
California, since the state began requiring postharvest processing of
Gulf Coast oysters in 2003, there had been zero consumption-related V.
vulnificus illnesses. A senior FDA official said FDA now believes that
postharvest processing of oysters is the control measure that best
meets the intent of its Hazard Analysis and Critical Control Point
seafood safety requirement to prevent, eliminate, or reduce to an
acceptable level the occurrence of pathogens such as V. vulnificus.
In a November 2009 letter to FDA, the ISSC expressed disappointment
that FDA had unilaterally decided to announce its intent to change its
policy and had not followed the 1984 memorandum of understanding that
calls for FDA and the ISSC to exchange information concerning the
shellfish safety program and resolve problems of interpretation and
policy. According to the letter, the ISSC was concerned that FDA was
now proposing to abandon the V. vulnificus risk management plans
adopted in 2001 by the ISSC with FDA concurrence. Furthermore, the
ISSC, with FDA's concurrence, had already agreed to implement new time
and temperature controls to address V. vulnificus beginning in May
2010. The ISSC letter also stated that if FDA continued its effort
without ISSC support, it was likely that many Gulf Coast states would
choose not to exercise their enforcement responsibilities under the
shellfish safety program with regard to postharvest processing, and
instead might implement intrastate programs that could allow
consumption of raw oysters produced within their state without the
controls necessary to substantially reduce V. vulnificus illnesses. In
its April 2010 response to the ISSC, FDA acknowledged the ISSC's
concerns and agreed to work collaboratively with it to identify the
steps needed before implementing a postharvest processing requirement
for Gulf Coast oysters harvested during the warmer months.
Specifically, FDA agreed to fund an independent study, which RTI later
conducted, to assess how postharvest processing and equivalent
controls could be implemented in the fastest, safest, and most
economical way.
Nevertheless, FDA and the ISSC have not yet agreed on a new illness
reduction goal and the strategies for achieving that goal. As we noted
in our October 2005 report on practices that can help agencies enhance
and sustain collaboration, agencies need to define and articulate the
common outcome they are seeking to achieve that is consistent with
their respective agency goals and missions.[Footnote 13] Also, to
achieve the common outcome, collaborating agencies need to establish
strategies that work in concert with those of their partners or are
joint in nature. Furthermore, trust is a necessary element for a
collaborative relationship and it is critical to involve all key
stakeholders in decision-making. In summary, our October 2005 report
indicates that absent effective collaboration, it is unlikely that
agencies can develop and implement joint agreements. If FDA and the
ISSC cannot agree on the V. vulnificus illness reduction goal and
strategies to achieve the goal, it is unlikely that the states'
efforts to significantly reduce the number of consumption-related V.
vulnificus illnesses will be effective.
The Approach FDA and the ISSC Use for Measuring Progress toward Their
Illness Rate Reduction Goal Has Limitations That Undermine Its
Credibility:
The approach FDA and the ISSC have been using to measure progress
toward their 60 percent illness rate reduction goal established in
2001 has three main limitations that undermine its credibility: the
limited number of states used in determining V. vulnificus illness
reduction, overstatement of the effectiveness of the primary V.
vulnificus illness reduction strategies, and not controlling for the
effect of such factors as natural and man-made disasters.
Limited number of states used in determining V. vulnificus illness
rate reduction. First, the approach FDA and the ISSC use for measuring
progress toward their illness rate reduction goal is based on the
inclusion of V. vulnificus illness data from four states: California,
Florida, Texas, and Louisiana. V. vulnificus illnesses related to raw
oyster consumption occur in other states, but the FDA and ISSC
measurement approach does not capture either the scope of such
illnesses from oysters harvested from the entire Gulf Coast region or
the national scope of V. vulnificus illnesses. According to a senior
ISSC official, the ISSC selected the four states because of the
quality of their illness reporting systems since each had been
consistently reporting V. vulnificus illnesses for the longest time
period and because most other states were not reporting V. vulnificus
illnesses. Since 2007, annually, about 20 states have reported V.
vulnificus illnesses to CDC. Senior FDA officials told us they advised
the ISSC to begin including more states in the V. vulnificus illness
calculation to better reflect the occurrence of V. vulnificus
illnesses nationally. According to FDA officials, the ISSC has not
responded to their recommendation. A senior ISSC official acknowledged
to us that analyzing national data would provide a more representative
measure of progress toward the illness rate reduction goal than the
current approach. The official told us that the ISSC is meeting in
October 2011 to discuss, among other things, developing an alternative
approach to measuring progress toward the illness rate reduction goal.
Overstatement of the effectiveness of primary V. vulnificus illness
rate reduction strategies. In addition to not reflecting the national
scope of V. vulnificus illnesses, the FDA and ISSC approach overstates
the effectiveness of their primary V. vulnificus illness rate
reduction strategies--consumer education and time and temperature
controls--by including V. vulnificus illness data from California.
Since 2003, California has required postharvest processing of all raw
Gulf Coast oysters harvested from April through October and sold in
the state and has reported two consumption-related V. vulnificus
illnesses since the requirement took effect. A senior ISSC official
acknowledged that California's postharvest processing requirement has
reduced the number of V. vulnificus illnesses in that state. This
official also acknowledged that including California's results
contributed significantly to achieving the interim 2006 40 percent
illness rate reduction goal. For this reason, both California and FDA
officials have requested that the ISSC no longer include California
data in its illness rate reduction calculation. According to a senior
ISSC official, however, California data should be included because
reporting states should not be excluded based on the states' chosen V.
vulnificus illness rate reduction strategies.
Lack of control for the effect of such factors as natural and man-made
disasters. The FDA and ISSC measurement approach does not control for
the effect of such factors as natural and manmade disasters. FDA,
ISSC, and state officials we spoke with agree that the level of V.
vulnificus illnesses is associated with the level of oyster production
and consumption. When oyster production decreases as a result of
factors such as natural or man-made disasters like Hurricane Katrina
in 2005 and the Deepwater Horizon oil spill in 2010, the level of
oyster consumption also decreases and, with it, the rate of V.
vulnificus illnesses. Not controlling for the effect of factors
external to the V. vulnificus illness rate reduction strategies chosen
by FDA and the ISSC gives a misleading indication of the success of
those strategies. In 2000, the ISSC considered a proposal to calculate
illness rate as the number of illnesses divided by oyster production.
According to a senior ISSC official, the ISSC did not approve the
proposal because oyster production data were not readily available,
which is no longer the case. After rejecting the 2000 proposal to
account for production, in 2001 the ISSC adopted a proposal to
calculate the illness rate as the number of illnesses per unit of
population. A senior FDA official told us that FDA initially agreed
with this proposal because illnesses per unit of population is a
standard measure used by CDC for tracking the prevalence of many
illnesses. In retrospect, however, FDA and ISSC officials told us that
population should not be part of the calculation. A senior FDA
official explained that tracking illnesses per unit of population is
meaningful for certain types of illnesses but is not meaningful for
others. For example, he told us that tracking illnesses per unit of
population makes sense for illnesses that are passed from person-to-
person or for food-borne illnesses associated with foods that are
widely consumed but that it does not make sense for illnesses
associated with foods like oysters, which are a specialty food and not
widely consumed throughout the population.
In 2009, the ISSC adopted a proposal to change its measure of
effectiveness from illness rate reduction to risk reduction, which
would be based on the risk per serving of raw or undercooked oysters.
Under the proposal, the revised goal would be to reduce the risk per
serving to a level equivalent to the current 60 percent illness rate
reduction goal. FDA initially opposed the proposal but later
concurred, stating that the change would eliminate the problems
associated with the current approach for measuring V. vulnificus
illness rate reduction. In March 2010, the ISSC appointed a workgroup
to explore implementation of the proposal. According to a senior ISSC
official, as of March 2011, the work group had held one conference
call but had not yet determined how the concept of risk per serving
would be applied and measured in the V. vulnificus illness context.
According to the ISSC official, the proposal is scheduled to be
implemented in January 2012.
FDA and the ISSC Have Taken Few Steps to Evaluate the Effectiveness of
Consumer Education and Time and Temperature Controls:
FDA and the ISSC have performed either very limited or no evaluations
of the effectiveness of their key V. vulnificus illness reduction
strategies. Specifically, the ISSC has not evaluated the effectiveness
of consumer education efforts in reducing V. vulnificus illnesses
since 2004, and FDA has not conducted any evaluations of its own. In
addition, although the V. vulnificus risk calculator developed by FDA
estimates that time and temperature controls can reduce V. vulnificus
illnesses, FDA and the ISSC have not directly evaluated the
effectiveness of the May 2010 time and temperature controls that the
ISSC approved, with FDA concurrence, for the states to use in reducing
consumption-related V. vulnificus illnesses.
The ISSC Has Not Evaluated the Effectiveness of Consumer Education
Efforts Since 2004, and FDA Has Not Conducted Its Own Evaluations:
The ISSC conducted consumer surveys in 2002 and 2004 that were
intended to measure the extent to which (1) V. vulnificus education
programs increased consumer awareness of the risks of eating raw
oysters and (2) high-risk consumers refrained from eating raw oysters
for health reasons. The 2002 survey of raw oyster consumers
established baseline information on consumers' beliefs about raw
oysters, consumption patterns, and knowledge of risks associated with
eating Gulf Coast raw oysters. The 2004 follow-up survey measured
whether raw oyster consumers changed their raw oyster consumption
patterns during the previous 2 years as a result of the ISSC's and
states' (Florida, Louisiana, and Texas) V. vulnificus consumer
education efforts. The 2004 survey found no significant increase in
overall consumer knowledge about the risk of eating raw oysters or the
proportion of high-risk consumers who stopped eating them.
A senior FDA official said that the agency has not conducted its own
evaluation of the effectiveness of V. vulnificus consumer education
efforts; instead it relied on the ISSC's surveys to determine the
impacts of consumer education efforts. FDA officials told us that
their review of consumer education efforts is limited to checking the
V. vulnificus risk management plans implemented by Florida, Louisiana,
and Texas to ensure the plans include a consumer education component.
According to FDA and state officials, the states' V. vulnificus
education efforts have included a variety of activities such as online
V. vulnificus education courses for physicians, nurses, and
dieticians; public service announcements for broadcast on television
and radio; advisories included with the drug prescriptions of high-
risk consumers; and brochures targeting high-risk consumers that
contained information about the risk of eating raw oysters. FDA and
ISSC officials stated that although they have not directly evaluated
the states' education efforts since 2004, their indirect measure of
the effectiveness of consumer education was whether they achieved
their 2008 60 percent illness rate reduction goal. They acknowledged,
however, that the goal was not achieved, and, therefore, presumably
consumer education alone would not achieve the goal.
Some state officials told us that it is very difficult to measure and
evaluate the direct impact that consumer education has on a relatively
rare event, such as V. vulnificus illness. One state official said
that his state did not have the expertise and financial resources to
conduct an evaluation of the effectiveness of its consumer education
programs. The same official added that it would be difficult to prove
that a specific case of V. vulnificus was prevented because of
consumer education efforts.
An ISSC official said that some members of the ISSC have concluded
that consumer education is not going to result in a significant
reduction in V. vulnificus illnesses. For example, one state official
said that the effectiveness of education is hampered by the fact that
some of those who are most vulnerable to V. vulnificus illness, such
as alcoholics with liver disease, are risk takers who refuse to change
their raw oyster consumption habits.
In our September 2005 report on managing for results, we noted that
federal agencies should regularly measure the effectiveness of their
programs to determine whether progress is being made toward
performance goals.[Footnote 14] Specifically, agencies should compare
their programs' results against their goals and determine where to
target program resources to improve performance. We recognize that it
is difficult to assess the effectiveness of consumer education
programs. Nonetheless, the absence of information on the effectiveness
of V. vulnificus consumer education programs limits the ability of the
ISSC and the states to identify and increase the use of consumer
education approaches that are working well and discontinue those that
have not been effective. Furthermore, without regular evaluations of
the effectiveness of consumer education, ISSC and state officials
cannot ensure that their resources are targeted strategically and are
not wasted on efforts that are ineffective.
FDA and the ISSC Have Not Directly Evaluated the Effectiveness of Time
and Temperature Controls in Reducing V. Vulnificus Consumption-Related
Illnesses:
Neither FDA nor the ISSC has directly evaluated the effectiveness of
the new time and temperature controls in reducing V. vulnificus
illnesses since they were implemented in May 2010. Instead, FDA and
the ISSC have relied on illness rate reduction as the overall measure
of effectiveness of all V. vulnificus illness reduction strategies
combined. Both FDA and ISSC officials acknowledge, however, that doing
so does not distinguish the effect of time and temperature controls
from that of other factors. Consequently, illness rate reduction does
not provide a direct indication of the effectiveness of the time and
temperature controls, implemented and enforced by the states, in
contributing to V. vulnificus illness reduction.
Senior FDA and ISSC officials told us that one way to more directly
evaluate the effectiveness of time and temperature controls is to
conduct studies to determine the level of V. vulnificus bacteria in
oysters prior to and following implementation of the
controls.[Footnote 15] FDA officials told us that such studies were
conducted in 1998-1999 and 2007, prior to the implementation of the
new time and temperature controls.[Footnote 16] Those studies surveyed
the level of V. vulnificus bacteria and other pathogens in oysters
collected from both retail and wholesale establishments.[Footnote 17]
The level of V. vulnificus bacteria found in 2007 was similar to that
found in 1998-1999. According to the 2007 study, the similarity was
not surprising given that time and temperature controls had not
changed since the 1998-1999 study and that the ISSC's efforts to
reduce V. vulnificus illnesses had focused on educating high-risk
consumers. FDA officials told us that data from those studies could be
compared against future study data to measure the effectiveness of new
controls, including time and temperature controls, aimed at reducing
exposure to V. vulnificus bacteria by consumers of raw oysters. A
senior ISSC official told us that he intends to promote the use of
such studies to evaluate time and temperature control effectiveness.
FDA officials told us that, although they would like to repeat the
1998-1999 and 2007 studies, FDA has no plans to do so given the
expense of the studies, competing priorities, and resource constraints.
To estimate the level of V. vulnificus illness rate reduction states
might expect to achieve from time and temperature controls, FDA and
the ISSC have relied on FDA's V. vulnificus risk calculator. Estimates
generated by the risk calculator indicated that the new time and
temperature controls implemented in May 2010 would help the states to
achieve the 60 percent illness rate reduction goal by the end of 2010.
To achieve the calculator's estimated illness rate reduction, oyster
industry members would have to fully comply with the time and
temperature controls. Our discussions with FDA, state officials, and
oyster industry representatives, however, suggest that while data
regarding compliance levels are unavailable, full compliance is highly
unlikely. In January 2011 FDA and the ISSC determined the goal still
had not been met.
To assess the precision of the risk calculator's estimates, we
replicated and modified a risk simulation model--developed by the
World Health Organization and the Food and Agriculture Organization
(WHO/FAO) of the United Nations in partnership with FDA--that FDA used
as a basis for developing the risk calculator. Our analysis indicates
that even with 100 percent compliance, the risk of V. vulnificus
illness under time and temperature controls may differ from the number
estimated by the risk calculator. For example, in Texas in the month
of August, FDA's risk calculator estimates that time and temperature
controls will lead to 2.84 illnesses per 100,000 raw oyster servings.
While this is accurate on average, the number of illnesses per 100,000
servings could be as low as 2.44 or as high as 3.63 (for a 90 percent
uncertainty interval), according to our analysis. We find a similar
range of uncertainty in the estimated number of V. vulnificus
illnesses for Florida and Louisiana. See appendix I for more details
about our analysis. While uncertainty is an inherent part of estimates
produced by all quantitative models, the risk calculator does not
report the amount of uncertainty associated with its estimates.
Extent of Industry Compliance with Time and Temperature Controls Is
Unknown:
Although under the shellfish safety guidelines, states are responsible
for enforcing oyster industry compliance with time and temperature
controls, senior officials in Florida, Louisiana, and Texas told us
they do not track compliance rates. A senior ISSC official confirmed
that these states do not systematically collect, analyze, and report
compliance information. Enforcement consists largely of periodic state
inspections of oyster-processing plants and on-the-water harvester
activities. The latter includes checking log sheets on which
harvesters record whether they are harvesting oysters for raw
consumption and, if so, whether they are complying with various
elements of the time and temperature controls. Enforcement personnel
in Louisiana, Florida, and Texas told us they do not inspect all
harvesting vessels and do not verify the accuracy of all of the
information recorded by the harvesters whose vessels they do inspect.
For example, a Louisiana official told us that Louisiana enforcement
personnel are to check the log sheet to ensure the harvester has
recorded the time harvesting began but has no way of verifying whether
the information is accurate. Figure 5 shows a sample log sheet used in
Louisiana.
Figure 5: Sample Log Sheet:
[Refer to PDF for image: illustration]
Harvester-Dealer Time/temperature:
Log Sheet:
Harvester Information:
Boat Name/number:
Harvester Name/License Number:
Harvester Signature:
Date:
Molluscan Shellfish Harvested For Other Than Raw (Half Shell)
Consumption:
Harvesting Area/lease No.:
Product Intended For Other Than Raw Consumption: Circle One:
Bedding;
Shucking;
Relaying;
Other (Explain):
Time Harvesting Begins:
Time Harvesting Ends:
Number Of Sacks Of Oysters Harvested:
Molluscan Shellfish Harvested For Raw (Half Shell) Consumption:
Harvesting Area/lease No.:
Time Harvesting Begins:
Number Of Sacks Of Oysters Harvested:
Certified Dealer Information:
Temperature Of Cooler When Unloading Oysters Begins:
Time When Last Oyster From Boat Are Placed In Cooler:
Temperature Of Cooler When Last Oysters From The Boat Are Placed In
Cooler:
Original Certified Dealer Signature (Or Authorized Representative):
Date:
Source: Louisiana Administrative Code, Title 51, Part IX, Chapter 3,
Section 345.
[End of figure]
FDA is responsible for evaluating states' enforcement of time and
temperature controls. However, FDA officials told us that FDA's
evaluations do not include assessments of the degree to which states
are ensuring industry compliance. Instead, FDA officials told us their
evaluations consist of checking states' V. vulnificus risk management
plans to ensure the plans include the time and temperature controls
outlined in the shellfish safety guidelines, accompanying state
officials on selected oyster-processing-plant inspections and on-the-
water patrols, and reviewing selected shellfish safety plans and
records.
A senior ISSC official told us the ISSC planned to evaluate the
effectiveness of the new time and temperature controls, in part, based
on the rate of oyster industry compliance and the level of states'
enforcement. However, because FDA, the ISSC, and states did not
collect any industry compliance or state enforcement data, when it
came time to conduct the evaluation in January 2011, the ISSC had to
rely on testimonial evidence from state officials regarding the extent
of industry compliance and state enforcement.
Although data are unavailable regarding oyster industry compliance
with time and temperature controls, our discussions with state
officials and oyster industry members suggest full compliance is
highly unlikely. During several discussions with state officials and
oyster industry members, we were told of instances of intentional
mislabeling, a form of seafood fraud.[Footnote 18] For example,
harvesters initially labeled oysters harvested without meeting the new
time and temperature controls for shucking or postharvest processing
only but later mislabeled them for raw consumption. Figure 6 shows
sample labels for oysters to be consumed raw and for oysters to be
shucked or postharvest processed.
Figure 6: Sample Oyster Labels:
[Refer to PDF for image: illustration]
Oyster label: green label;
Description: Green labels are to be attached to sacks of oysters that
have been harvested without meeting time and temperature controls and
that should not, therefore, be consumed raw and instead must be
shucked or postharvest processed.
Oyster label: Blue label;
Description: Blue labels are to be attached to sacks of oysters that
have been harvested without meeting time and temperature controls but
have been postharvest processed to reduce V. vulnificus bacteria to
nondetectable levels.
Oyster label: White label;
Description: White labels are to be attached to sacks of oysters that
have been harvested in compliance with time and temperature controls
and may be consumed raw.
Source: Louisiana Department of Wildlife and Fisheries.
[End of figure]
According to two large oyster processors we spoke with operating in
both Louisiana and Texas, mislabeling is widespread and is driven by a
considerable financial incentive to avoid the costs of complying with
the time and temperature controls and obtain the higher price accorded
raw oysters. A senior Florida regulatory official told us that
mislabeling was identified during a recent routine inspection of a
local oyster-processing plant and that he was aware of several
occasions where oysters were served raw that should have been shucked
or postharvest processed because they had not been harvested in
compliance with the time and temperature controls. In July 2010, the
ISSC sent a letter to member states informing them of deaths traced to
raw consumption of oysters that should have been shucked or
postharvest processed and requesting immediate action to ensure
accurate labeling. According to a senior Louisiana law enforcement
official, however, mislabeling is an easy practice to engage in and is
very difficult for regulatory and law enforcement personnel to detect.
During a January 2011 ISSC meeting, ISSC members acknowledged that
compliance with the time and temperature controls was not as good as
it should be. According to the meeting minutes, there have been
numerous complaints from oyster processors regarding instances of
noncompliance in Florida. At the January 2011 meeting, the ISSC passed
a motion encouraging increased enforcement of the time and temperature
controls by the Gulf Coast states. As of March 2011, however, the ISSC
was unable to tell us what specifically they meant by increased
enforcement or how the states planned to implement the motion. A
senior FDA official told us that this motion is unlikely to be
implemented in any meaningful way given limited state enforcement
capacity.
Extent of Illness Reduction from Time and Temperature Controls Depends
on Compliance Levels:
Given that compliance data are unavailable and that compliance rates
are likely to be less than 100 percent, according to FDA, state
officials, and oyster industry representatives, we used our
modification of the WHO/FAO risk simulation model to estimate the
effect of the compliance rate on the effectiveness of time and
temperature controls in reducing V. vulnificus illness. Specifically,
we estimated the number of illnesses during the summer months under
the baseline scenario--in which the new more stringent 2010 time and
temperature controls were not in effect--and under scenarios that
assumed various levels of compliance with the new time and temperature
controls. Our estimates show that the extent to which the new time and
temperature controls would reduce V. vulnificus illnesses varies
considerably with the level of compliance. For example, during a
typical August month in Louisiana, assuming that 100 percent of
oysters are harvested in compliance with time and temperature
controls, the risk calculator estimates the controls will reduce
illnesses by 41 percent on average, and our analysis estimates that
illness reduction could range from 30 percent to 47 percent.[Footnote
19] As shown in figure 7, at lower levels of compliance, the illness
reduction would be considerably smaller. If 80 percent of the oysters
are harvested in compliance with these controls--meaning that 20
percent would be harvested out of compliance--we estimate that time
and temperature controls would reduce illnesses by 15 percent to 23
percent. As a result, even assuming 80 percent compliance in the
summer months, it is unlikely that these controls will lead to the
level of illness reduction estimated by the risk calculator. We found
that noncompliance would have a similar effect in the other summer
months and in the other states. See appendix I for details.
Figure 7: Estimated Reduction in the Number of V. Vulnificus Illnesses
as a Result of Time and Temperature Controls for Louisiana in August:
[Refer to PDF for image: line graph]
Percentage reduction in number of illnesses from baseline:
Risk calculator estimate: 41%.
Estimated compliance rate: 70%;
Upper bound: 17%;
Lower bound: 11%;
Mean: 14%.
Estimated compliance rate: 80%;
Upper bound: 23%;
Lower bound: 15%;
Mean: 19%.
Estimated compliance rate: 90%;
Upper bound: 32%;
Lower bound: 21%;
Mean: 26%.
Estimated compliance rate: 100%;
Upper bound: 47%;
Lower bound: 30%;
Mean: 38%.
Source: GAO.
[End of figure]
Adequate Capacity Does Not Exist to Use Postharvest Processing on Gulf
Coast Oysters, and Questions Exist about the Feasibility of Developing
Such Capacity:
According to a March 2011 FDA-commissioned report by RTI, the Gulf
Coast oyster industry does not currently have adequate capacity to use
postharvest processing on all Gulf Coast oysters intended for raw
consumption that are harvested during warmer months.[Footnote 20] The
report found that two key issues need to be addressed to develop
adequate capacity, including the construction of several central
postharvest processing facilities. The report concluded that it would
take at a minimum 2 to 3 years to develop the necessary capacity.
[Footnote 21] However, we identified six issues of concern regarding
the RTI report's economic analysis that call into question the
completeness of its cost and timeline estimates.
RTI Report Finds Adequate Capacity Does Not Exist for Postharvest
Processing without Additional or Expanded Facilities:
In October 2009, FDA announced its intent to begin requiring
postharvest processing, in part, because it believed that adequate
capacity existed. RTI's March 2011 FDA-commissioned report, however,
found that adequate capacity does not exist and identified two key
issues that must be addressed to ensure such capacity. First, about
five or six central postharvest processing facilities would be needed
to accommodate smaller Gulf Coast oyster processors that may be unable
to conduct postharvest processing at their current facilities due to
various limitations. For example, these smaller facilities generally
lack sufficient floor space for installing postharvest processing
equipment without undergoing costly plant expansion, and their owners
may lack the financial resources to expand their plants and purchase
postharvest processing equipment. In addition, the report described
several necessary steps in developing the central facilities,
including: (1) determining the legal and operating structure of the
facilities, (2) identifying the property where the newly constructed
facilities are to be located or existing buildings are to be modified,
and (3) securing the financing for developing the facilities. While
central facilities may allow some smaller oyster processors access to
postharvest processing facilities during the warmer months, other
challenges remain, such as the additional costs to transport oysters--
refrigerated--to and from the central facilities.
Second, technical and financial assistance to several processing
facilities would be needed to expand or alter their existing
facilities, and purchase and install additional postharvest processing
equipment. Again, the report describes several steps that must occur
before initiating the expansion of existing facilities, such as
developing plans for expanding the plant or altering the plant layout,
and securing financing for purchasing additional equipment and
constructing the expanded facility. Overall, the RTI report concluded
that it will take a minimum of 2 to 3 years and, depending on the
postharvest processing method used, about $6 million to $32 million in
initial investment costs (excluding land purchase and construction
costs for new centralized facilities) to develop the infrastructure
required to ensure the Gulf Coast oyster industry has adequate
capacity to use postharvest processing on all Gulf Coast oysters
intended for raw consumption that are harvested during warmer months.
In our July 2001 report on shellfish safety, we raised the concern
that if the 60 percent V. vulnificus illness rate reduction goal was
not achieved by 2008, postharvest processing capacity may not be
available because the ISSC did not have a detailed plan for ensuring
such capacity.[Footnote 22] Consequently, we recommended that FDA work
with the ISSC to prepare and implement a detailed plan for developing
adequate postharvest processing capacity to help achieve the ISSC's V.
vulnificus illness rate reduction goals. In its response, FDA agreed
with our recommendation, and the ISSC agreed that it did not have a
detailed plan to ensure postharvest processing capacity. At that time,
ISSC officials said that the matter was a high priority and would be
addressed at its upcoming July 2001 meeting. At the July 2001 meeting,
the ISSC proposed that the V. vulnificus risk management plans include
a process for implementing a required postharvest treatment capacity
for 50 percent of all oysters intended for the raw consumption market--
during the months of May through September--should the 40 percent
illness reduction goal not be achieved by December 31, 2006.
In 2003, the ISSC surveyed oyster dealers with postharvest processing
capabilities in Florida, Louisiana, and Texas and found that there was
sufficient capacity to use postharvest processing on 100 percent of
the oysters harvested from May through September that were intended
for raw consumption. According to FDA officials, until January 2011,
when RTI presented its preliminary results, they believed there was
sufficient capacity to use postharvest processing on all Gulf Coast
oysters harvested from May through September. However, according to an
ISSC official, the 2003 survey had major limitations such as quick
freezing as a postharvest processing option, not considering the
location of existing postharvest processing facilities, and not
addressing whether existing facilities would treat their competitors'
oysters. The RTI report indicated that quick freezing is not
appropriate for oysters harvested in warmer months because this option
substantially reduces their quality. The ISSC official said that in
hindsight, FDA and the ISSC did not adequately define capacity in 2001
when they began to discuss postharvest processing capacity goals.
Six Issues of Concern Raise Questions about the Feasibility of
Developing Adequate Postharvest Processing Capacity:
FDA stated in October 2009 that postharvest processing should be
required beginning in May 2011, in part, because it believed that
adequate capacity existed. When the ISSC raised concerns, FDA tasked
RTI with analyzing the feasibility and economic impacts of such a
requirement. Although we believe that the overall method RTI used for
its analysis is credible, its conclusion--that postharvest processing
capacity to treat all Gulf Coast oysters intended for raw consumption
that are harvested from April through October can be developed in a
minimum of 2 to 3 years--is questionable due to six issues of concern
we identified in RTI's economic analysis. We recognize that some of
the issues we identified are the result of constraints faced by RTI,
such as not being within the scope of the FDA-approved RTI report work
plan, a lack of data, and the associated contractual report due dates
(i.e., FDA needed the report completed before the 2011 summer oyster
harvest season to help inform policy decisions). The six issues of
concern are as follows.
* Baseline data may not be representative of the industry. The RTI
report relied on 2008 data--such as oyster harvest volumes, oyster
prices, and the number of Gulf Coast oyster processors--as a
representative baseline to estimate economic impacts of a postharvest
processing requirement. We believe the 2008 data are not necessarily
representative of the current state of the Gulf Coast oyster industry
due to the events that occurred in 2010--the Deepwater Horizon oil
spill and the implementation, on May 1, 2010, of the new, more
stringent time and temperature controls. The lead author of the RTI
report explained that using the 2008 data as a baseline was
appropriate because 2008 was the most recent and complete year of
data. Furthermore, the lead author said that it could take several
years for the oyster industry to adjust to the 2010 events and that
waiting for this adjustment to occur would not necessarily change the
overall report's conclusions regarding the economic impacts of
postharvest processing. However, the lead author acknowledged that
using 2008 data was a limitation of the study and that using a
baseline after the 2010 events would allow for more refined estimates.
We believe the estimates in the report may be of limited use for
determining how the market would respond to a postharvest processing
requirement because the estimates are premised on the oyster
industry's structure prior to the 2010 oil spill and implementation of
the new time and temperature controls, which may not reflect the Gulf
Coast oyster industry of the future. For example, oyster production
was severely curtailed in 2010 compared with the baseline production
in 2008. According to the Louisiana Department of Wildlife and
Fisheries, the Louisiana oyster harvest was down by 50 percent in
2010. Given the baseline used, the results of the economic impact
analysis may not provide a valid basis for the oyster-processing
industry to make investment decisions if a postharvest processing
requirement is implemented.
* Key costs are excluded. Certain key costs are excluded from the
report's economic analysis. For example, the report does not include
information on costs associated with purchasing land needed to expand
existing postharvest processing facilities or construct new
centralized facilities. The lead author of the RTI report said that
land costs vary significantly by location. According to the lead
author, a detailed analysis of such costs was not within the scope of
the FDA-approved RTI report work plan. Although we agree that such
costs are highly variable across regions, we believe that including a
mean or median land cost would be better than omitting land costs
altogether, as such costs may account for a large portion of the total
costs to expand existing or construct new facilities. In addition,
other significant costs are excluded from the economic analysis, such
as construction costs for the new centralized facilities, insurance
coverage for additional processing plant space and postharvest
processing equipment, and costs for transporting oysters to and from
the central postharvest processing facilities. The report acknowledges
that insurance coverage may be a significant expense, especially in
areas prone to severe weather and flooding. Furthermore, according to
the report, processors will incur transportation costs if they are
unable to install processing equipment at their facilities and instead
have to rely on centralized facilities. Transportation costs would
include either paying for trucking services or purchasing and
operating a refrigerated truck. According to the lead author, these
costs were not included because a detailed analysis of such costs was
not within the scope of the FDA-approved report work plan. If key
costs are not analyzed and included in the cost estimates, the full
scope of the financial resources needed to ensure the Gulf Coast
oyster industry has sufficient capacity to use postharvest processing
on oysters harvested during the warmer months will not be known.
* Who would pay to expand processing capacity is not clear. The RTI
report does not clearly address who would pay for postharvest
processing, which includes purchasing and installing the equipment, as
well as transporting harvested oysters to and from postharvest
processing facilities. The lead author of the RTI report said
inquiring about possible financial sources available for subsidizing
the expansion of postharvest processing capacity was beyond the scope
of the report. However, she suggested that expansion could be
subsidized by an entity within state government or by an oyster
industry cooperative that was established to develop a financing
mechanism. In addition, she said the ISSC could take the lead in
coordinating the development of the financing mechanisms needed to
expand postharvest processing facilities. We believe that identifying
financial support is a major issue in assessing the feasibility of
requiring postharvest processing, particularly considering state
government budget constraints and the financial losses the oyster
industry incurred as a result of the 2010 Deepwater Horizon oil spill
and the ongoing effects of the recent economic recession. Difficulties
in obtaining financing could impact the time frame for postharvest
processing to become operational, and therefore the minimum 2 to 3
year estimate for increasing capacity might not be reliable.
* Limited support exists for estimated time frame for increasing
postharvest processing capacity. According to the RTI report, existing
processors would need a minimum of 2 years to increase their
postharvest processing capacity; however, the report does not describe
in detail the basis for the 2-year estimate. In addition, according to
the report, it will take at least 3 years to develop the centralized
postharvest processing facilities. The lead author of the report said
that the estimates were based, in part, on information obtained from
surveying Gulf Coast processors. However, the lead author acknowledged
that few processors contributed cost information associated with
purchasing, installing, and operating postharvest processing equipment
because this type of information is proprietary. We recognize the
proprietary nature of the cost data, but we believe the basis for
RTI's time frame estimates could be more transparent. For example, the
report could provide specific time frames associated with the steps
the report says are required to increase postharvest processing
capacity. Absent such transparency, it is difficult to know whether
the estimate is well supported and likely to be accurate.
* Assumptions about postharvest processing for oysters shipped within
state borders are likely inaccurate. The RTI report's economic impact
analysis assumes that three Gulf Coast states--Florida, Texas, and
Louisiana--would require postharvest processing for oysters harvested
in the warmer months that are intended for raw consumption and sold
within the state's borders. However, statutes passed in 2011 in both
Louisiana and Texas state that federal regulations that prohibit the
interstate sale of oysters without postharvest processing do not apply
to oysters harvested and sold within the state.[Footnote 23] By not
incurring the added cost of postharvest processing, these oysters
would affect overall oyster prices. The lead author of the RTI report
agreed that the availability of cheaper nonpostharvest processed raw
oysters might significantly constrain the ability of retailers and
restaurateurs, for example, to sell the higher-priced postharvest
processed oysters. Although the RTI report's analysis includes a range
of assumptions on the likely proportion of oysters sold within or
outside of a state's borders, these assumptions are not incorporated
in the economic impact model. Incorporating them is important because
they provide oyster processors with important information on whether
they should make investments in postharvest processing equipment. For
instance, if some state regulations allow the sale, within state
borders, of oysters intended for raw consumption without postharvest
processing, certain processors may decide not to sell oysters outside
their state to avoid the cost of postharvest processing equipment,
which would place competitive pressure on all oyster prices. We
believe that without including a range of assumptions about the
proportion of oysters likely to be sold both within and outside of a
state's borders, the overall economic impacts, including the
likelihood of oyster processors investing in postharvest processing
capacity, will not be fully known.
* Postharvest processing costs may not be able to be passed on to
consumers. The RTI report also assumes that oyster processors can pass
on some of their postharvest processing costs to consumers. The
studies cited in the report indicate there is no clear consensus on
whether any of the postharvest processing costs could be passed on to
consumers or, if the costs could be passed on, what the amount would
be. These studies generally found that consumers preferred raw
unprocessed oysters to postharvest processed oysters, and although
some were willing to accept postharvest processed oysters, they were
not necessarily willing to pay a higher price for them. The lead
author of the RTI report agreed that the report's assumption that
oyster processors can pass on some of their postharvest processing
costs to consumers is uncertain. Without the ability to pass on their
higher costs to consumers, many of the current oyster-processing
establishments could face closure because, with the addition of
postharvest processing costs, their total costs may exceed their
returns. Also, oyster harvesters who depend on these processors may
have to stop harvesting during the warmer months or quit harvesting
altogether. Without an analysis that provides a range of estimates for
the price increase that could be passed on to consumers, the Gulf
Coast oyster industry will not have sufficient information to help
determine whether postharvest processing is economically feasible.
Conclusions:
It has been nearly 2 years since FDA informed the ISSC that the
current V. vulnificus illness rate reduction goal does not
sufficiently protect public health. However, since then, FDA and the
ISSC have not come to agreement on what an appropriate V. vulnificus
illness reduction goal should be or on the best strategy to achieve
such a goal. In the absence of such agreement, it will be very
difficult for FDA and the ISSC to make progress in reducing the number
of V. vulnificus illnesses. In addition, the approach FDA and the ISSC
use to measure progress in reducing V. vulnificus illnesses has three
main limitations that undermine its credibility. For example, the
approach is based on data from only four states, including California,
which has had nearly zero consumption-related V. vulnificus illnesses
since it began requiring postharvest processing of Gulf Coast oysters
in 2003. Consequently, the FDA and ISSC measurement approach does not
provide a credible representation of the Gulf Coast or national impact
of V. vulnificus illnesses or the real status of their efforts to
reduce them. Since 2001, FDA, the ISSC, and the Gulf Coast states have
relied on consumer education and time and temperature controls to
reduce V. vulnificus illnesses, but neither FDA nor the ISSC has
routinely evaluated whether these strategies have been effective in
reducing V. vulnificus illness. Our analysis shows that the extent to
which the new time and temperature controls will reduce V. vulnificus
illnesses varies considerably with the level of compliance. Without
regular evaluations of these illness reduction strategies, FDA, the
ISSC, and state officials and policymakers have no way of knowing
whether either strategy has been successful and should be continued or
is ineffective and should be stopped, which can result in wasted
resources and a failure to reach policy goals. Finally, FDA has
concluded that because consumer education and time and temperature
controls have not resulted in achievement of the 60 percent illness
rate reduction goal, Gulf Coast oysters harvested during the warmer
months and intended for raw consumption should be postharvest
processed to reduce V. vulnificus to nondetectable levels. However,
the 2011 FDA-commissioned RTI report found that adequate capacity to
use postharvest processing on all Gulf Coast oysters harvested from
April through October that are intended for raw consumption does not
currently exist and is at best 2 to 3 years away. Furthermore, our
review of the report's economic analysis found several issues that the
report did not thoroughly address, which could significantly impact
the feasibility of developing adequate postharvest processing capacity
specified in the FDA-commissioned report.
Recommendations for Executive Action:
To better ensure the safety of oysters from the Gulf of Mexico that
are sold for raw consumption, we recommend that the Secretary of
Health and Human Services (HHS) direct the Commissioner of FDA to work
with the ISSC to take the following four actions:
* Agree on a nationwide goal for reducing the number of V. vulnificus
illnesses caused by the consumption of Gulf Coast raw oysters and
develop strategies to achieve that goal, recognizing that consumer
education and time and temperature controls have not resulted in
achievement of the 60 percent V. vulnificus illness rate reduction
goal and that the capacity to use postharvest processing on Gulf Coast
oysters harvested from April through October that are intended for raw
consumption does not currently exist.
* Correct the limitations in the current approach to measuring
progress toward the 60 percent V. vulnificus illness rate reduction
goal or design and implement a new approach without these limitations.
* Regularly evaluate the effectiveness of V. vulnificus illness
reduction strategies, such as consumer education and time and
temperature controls, to determine whether they are successful and
should be continued or are ineffective and should be stopped.
* Conduct further study of the six issues of concern we identified
regarding the RTI report's economic analysis to ensure a more accurate
assessment of the feasibility of developing adequate capacity and
before FDA and the ISSC move forward with revising the National
Shellfish Sanitation Program's shellfish safety guidelines to provide
postharvest processing for oysters harvested from Gulf Coast waters
during warmer months and intended for raw consumption.
Agency Comments and Our Evaluation:
We provided a draft of this report to HHS and the ISSC for review and
comment. In written comments, which are included in appendix II, HHS
provided FDA responses, which generally agreed with the report's four
recommendations. Specifically, FDA agreed with our first and second
recommendations. Regarding our third recommendation, FDA agreed that
the approach used to evaluate the effectiveness of illness reduction
strategies has limitations that undermine its credibility. FDA also
said that assessing the effectiveness of existing controls on illness
reduction is extremely difficult. In an effort to better monitor
compliance with time and temperature controls, FDA intends to take a
number of steps, including conducting annual on-site checks at oyster
landing sites and processing plants to examine compliance with V.
vulnificus Hazard Analysis and Critical Control Point controls,
harvester records, time and temperature logs, and actual product
temperature. We recognize that assessing the effectiveness of V.
vulnificus illness reduction strategies is difficult, but continue to
believe it would useful for FDA and the ISSC to attempt to do so,
because without such evaluations it is difficult to determine whether
the strategies are successful and should be continued or are
ineffective and should be stopped.
Concerning our fourth recommendation, which identified six issues of
concern in the FDA-commissioned report on postharvest processing
capacity, FDA agreed to conduct further study or take other actions to
address our concerns on four issues--key costs are excluded; who would
pay to expand processing capacity is unclear; support for estimated
time frame for increasing postharvest processing capacity is limited;
and assumptions about postharvest processing for oysters shipped
within state borders are likely inaccurate--but disagreed with one
issue and neither agreed nor disagreed with the other issue. FDA
disagreed with our assessment that the 2008 baseline data used in the
study may not be representative of the Gulf Coast oyster industry.
Furthermore, FDA said that use of 2010 data would not have represented
a typical harvest year because the Deepwater Horizon oil spill
resulted in closures of many Gulf Coast oyster harvest areas, thereby
reducing oyster harvest levels. However, we did not suggest that 2010
data be used for the baseline; instead we believe it is preferable to
use a baseline from either an average of several years or a
sensitivity analysis of alternative baselines, including one that
incorporates data for 2010, the year of the Deepwater Horizon oil
spill and the implementation of the new, more stringent time and
temperature controls. FDA did not agree or disagree with our
assessment that postharvest processing costs may not be able to be
passed on to consumers. Instead, FDA stated that there are many
uncertainties regarding whether the cost of postharvest processed
oysters can be passed on to consumers. We believe the FDA-
commissioned report's analysis could be improved by providing a range
of postharvest processing cost estimates that can passed on to
consumers, which would help the oyster industry determine the extent
to which postharvest processing is economically feasible. FDA also
provided technical comments, which we incorporated as appropriate.
The ISSC stated in its written comments--which are included in
appendix III--that it generally agreed with the recommendations in the
report. The ISSC also provided additional information on the FDA and
ISSC efforts to address V. vulnificus illnesses and the circumstances
that led to the implementation of the current V. vulnificus illness
reduction strategies. Also, the ISSC commented that the goal of the V.
vulnificus risk management plans was to reduce V. vulnificus illnesses
nationally and that four states--California, Florida, Louisiana, and
Texas--were used to measure effectiveness. However, even though the
ISSC states that the 60 percent illness rate reduction goal is a
national goal, it determined achievement toward a national goal by
calculating the rate of illness for those four states.
As agreed with your office, unless you publicly announce the contents
of this report earlier, we plan no further distribution until 30 days
from the report date. At that time, we will send copies to the
appropriate congressional committees, the Secretary of Health and
Human Services, the Executive Director of the Interstate Shellfish
Sanitation Conference, and other interested parties. In addition, the
report will be available at no charge on the GAO Web site at
[hyperlink, http://www.gao.gov].
If you or your staff have any questions about this report, please
contact me at (202) 512-3841 or shamesl@gao.gov. Contact points for
our Offices of Congressional Relations and Public Affairs may be found
on the last page of this report. Key contributors to this report are
listed in appendix IV.
Sincerely yours,
Signed by:
Lisa Shames:
Director, Natural Resources and Environment:
[End of section]
Appendix I: Estimating the Impact of Time and Temperature Controls on
the Number of Illnesses from V. Vulnificus:
To estimate the impact of time and temperature controls on the number
of illnesses from Vibrio vulnificus (V. vulnificus), we took several
steps. We replicated a model developed by the World Health
Organization and the Food and Agriculture Organization (WHO/FAO) that
simulates the risk of illness from V. vulnificus based on several
factors, such as water temperature and the number of hours that
harvested oysters are left unrefrigerated.[Footnote 24] We then
modified the model to simulate the impact of the time and temperature
controls implemented in Florida, Louisiana, and Texas in May 2010.
Specifically, we examined the amount of uncertainty in the model's
estimates of the risk of illness and the impacts of various levels of
compliance with time and temperature controls on the estimated number
of V. vulnificus illnesses.
The Structure of the WHO/FAO Risk Simulation Model:
The WHO/FAO, in partnership with the Food and Drug Administration
(FDA), developed a risk simulation model that estimates levels of V.
vulnificus in raw oysters and the subsequent impact of these levels on
the risk of illness.[Footnote 25] The risk simulation model was
presented in a WHO/FAO report on the assessment of risk of V.
vulnificus in raw oysters.[Footnote 26] To estimate the impact of time
and temperature controls on the risk of V. vulnificus illnesses, we
replicated this model and modified it to account for potential changes
to harvesting and storage practices in response to the imposition of
time and temperature controls and to analyze various rates of
compliance with these controls.
The WHO/FAO risk simulation model is a Monte Carlo simulation, a type
of numerical analysis that produces a range of estimates to account
for the natural variability in the model's data inputs and the
statistical uncertainty in the parameters of the model's equations.
Data inputs, such as water temperature, vary naturally from day to day
within a month and from year to year for a given month. Similarly, the
parameters in the model's equations that estimate V. vulnificus levels
based on data inputs and that predict the risk of illness based on
estimated V. vulnificus levels, while based on scientific studies, are
subject to statistical uncertainty. To account for the variability in
data inputs and the uncertainty in the parameters of the model's
equations, the WHO/FAO risk simulation model calculates a range in
possible estimates of risk, each using slightly different values of
the data inputs and slightly different values of the parameters. To
produce a single estimate of the risk of illness, the model first
estimates levels of V. vulnificus at each of four stages in the
production process--from harvest to first refrigeration to cooldown to
consumption. The model then estimates the overall risk of illness
based on the estimated levels of V. vulnificus at the four stages.
Figure 8 illustrates each of the stages of the model and the factors
that influence them. In this figure, the light gray boxes represent
the input factors, the black boxes represent calculations based on
those factors, and the arrows indicate which factor influences which
calculation. For example, water temperature, air temperature and the
number of hours that oysters are left unrefrigerated are input factors
that influence the level of V. vulnificus in oysters at the time they
are first refrigerated. The dark gray box and the dotted arrows
represent our modification of the WHO/FAO risk simulation model.
Specifically, we modified the way in which the model determines the
number of hours that oysters are left unrefrigerated, which will then
impact the level of V. vulnificus in oysters at the time of cooldown.
Finally, because V. vulnificus has been found to stop growing and to
begin dying off when refrigerated at 55 degrees Fahrenheit or below,
the number of days oysters are refrigerated affects the V. vulnificus
level at the time oysters are consumed, which affects the number of V.
vulnificus illnesses that are likely to occur.
Figure 8: Key Components of WHO/FAO Risk Simulation Model and GAO
Modifications:
[Refer to PDF for image: illustration]
1) Water temperature (Data inputs used in WHO/FAO model):
Relationships in WHO/FAO model:
V. vulnificus level at time of harvest;
Air temperature.
GAO modification to WHO/FAO model:
Time and temperature controls.
2) Air temperature (Data inputs used in WHO/FAO model):
Relationships in WHO/FAO model:
V. vulnificus level at time of refrigeration;
V. vulnificus level at time of cooldown.
3) V. vulnificus level at time of harvest (Estimates made by WHO/FAO
model):
Relationships in WHO/FAO model:
V. vulnificus level at time of refrigeration.
4) V. vulnificus level at time of refrigeration (Estimates made by
WHO/FAO model):
Relationships in WHO/FAO model:
V. vulnificus level at time of cooldown.
Data inputs used in WHO/FAO model: Number of hours unrefrigerated.
5) V. vulnificus level at time of cooldown (Estimates made by WHO/FAO
model):
Relationships in WHO/FAO model:
V. vulnificus level at time of consumption.
Data inputs used in WHO/FAO model: Number of hours to cooldown.
6) V. vulnificus level at time of consumption (Estimates made by
WHO/FAO model):
Relationships in WHO/FAO model:
Illnesses per 100,000 servings (Estimates made by WHO/FAO model):
Data inputs used in WHO/FAO model: Number of days in refrigeration.
7) Time and temperature controls (GAO modification to WHO/FAO model):
GAO modification to WHO/FAO model:
Number of hours unrefrigerated (Data inputs used in WHO/FAO model);
Number of hours to cooldown (Data inputs used in WHO/FAO model).
8) Number of days in refrigeration (Data inputs used in WHO/FAO model):
V. vulnificus level at time of consumption.
Source: GAO replication of and modification to WHO/FAO risk simulation
model.
Note: The arrow from water temperature to time and temperature
controls indicates that time and temperature controls vary from month
to month--the controls are stricter during warmer months and less
strict during cooler months.
[End of figure]
Our Replication and Modification of the WHO/FAO Risk Simulation Model:
To replicate the WHO/FAO risk simulation model, we took several steps.
We reviewed the WHO/FAO risk assessment and documented the model's key
data inputs, assumptions, and equations. We asked FDA modelers, who
led the development of the WHO/FAO risk simulation model, to review
our documentation, and we revised our version of the model based on
their comments. We programmed the model in Statistical Analysis
Software, generated preliminary estimates, and asked FDA modelers to
review these estimates, and we compared these estimates to identify
remaining differences between our version of the model and the version
used in the WHO/FAO risk assessment. We used the same data inputs as
reported in the WHO/FAO risk assessment, including the statistical
distributions of water temperature, the difference between water
temperature and air temperature, the number of hours that oysters are
unrefrigerated, the number of hours until oysters cool down to 55
degrees Fahrenheit, and the number of days that oysters remain in
refrigeration. We also used the same values as the WHO/FAO risk
assessment for parameters that convert these data inputs into the
model's estimates, such as the parameters that define the relationship
between water temperature and V. vulnificus levels at harvest and the
parameters that define the relationship between V. vulnificus levels
at consumption and the risk of illness.[Footnote 27]
To verify that we correctly replicated the WHO/FAO risk simulation
model, we compared our estimates to the estimates reported in the WHO/
FAO risk assessment for each of the four seasons. Using the same data
inputs, model parameters, and assumptions, our estimates of the risk
of illness differ from the estimates reported in the WHO/FAO risk
assessment by less than 1 percent in the spring, less than 1 percent
in the summer, 4 percent in the fall, and 42 percent in the winter.
[Footnote 28] We report only our estimates from the summer months
because the risk of V. vulnificus illness is greatest during these
months and because estimates from our model are most similar to the
estimates from the WHO/FAO model during these months. See table 2 for
a comparison between the WHO/FAO and GAO estimates for key stages of
the model for the summer months.
Table 2: Comparison of Estimates between WHO/FAO Risk Simulation Model
and GAO Replication of the Model for July, August, and September:
Variables: Log10 V. vulnificus per gram at time of harvest;
WHO/FAO risk: simulation model: Mean: 3.27;
WHO/FAO risk: Standard deviation: 0.64;
[Empty];
GAO replication of risk simulation model: Mean: 3.27;
GAO replication of risk simulation model: Standard deviation: 0.65;
[Empty];
Percentage difference[A]: Mean: 0.00;
Percentage difference[A]: Standard deviation: 1.50.
Variables: Log10 V. vulnificus per gram growth prior to refrigeration;
WHO/FAO risk: simulation model:
Mean: 0.68;
Standard deviation: 0.35;
GAO replication of risk simulation model:
Mean: 0.69;
Standard deviation: 0.34;
Percentage difference[A]:
Mean: 1.45%;
Standard deviation: -2.86%.
Variables: Log10 V. vulnificus per gram growth during cooldown;
WHO/FAO risk: simulation model:
Mean: 0.50;
Standard deviation: 0.22;
GAO replication of risk simulation model:
Mean: 0.50;
Standard deviation: 0.22;
Percentage difference[A]:
Mean: 0.00;
Standard deviation: 0.00.
Variables: Log10 V. vulnificus per gram die-off during
refrigeration[B];
WHO/FAO risk: simulation model:
Mean: 0.31;
Standard deviation: n/a[D];
GAO replication of risk simulation model:
Mean: 0.32;
Standard deviation: 0.15;
Percentage difference[A]:
Mean: 3.13%;
Standard deviation: n/a[D].
Variables: Number of V. vulnificus Illnesses per 100,000 Servings;
WHO/FAO risk: simulation model:
Mean: 4.28;
Standard deviation: 90 percent uncertainty interval (3.69-4.97)[C];
GAO replication of risk simulation model:
Mean: 4.30;
Standard deviation: 90 percent uncertainty interval (3.62-5.19)[C];
Percentage difference[A]:
Mean: 0.47;
Standard deviation: n/a[E].
Source: WHO/FAO risk assessment and GAO replication of WHO/FAO risk
simulation model.
[A] Estimates reported in the WHO/FAO risk assessment were rounded to
two numbers past the decimal point. As a result, the calculations of
percentage difference between the WHO/FAO estimates and the GAO
estimates may be subject to rounding errors.
[B] V. vulnificus per gram die-off implies a reduction in bacterial
levels during refrigeration. Therefore, although the estimates are
presented as positive numbers, they refer to a decrease in V.
vulnificus levels.
[C] To be consistent with the estimates in the WHO/FAO risk
assessment, we present the 90 percent uncertainty interval for the
number of V. vulnificus illnesses per 100,000 servings, rather than
the standard deviation.
[D] The standard deviation of the Log10 V. vulnificus per gram die-off
rate during refrigeration was not reported in the WHO/FAO risk
assessment.
[E] For simplicity, the percentage differences between the upper and
lower bound of the 90 percent uncertainty intervals are omitted,
although the information provided in the table allows these percentage
differences to be calculated.
[End of table]
After verifying that we replicated the WHO/FAO risk simulation model,
we modified it to simulate the impact of time and temperature controls
in Florida, Louisiana, and Texas. Since time and temperature controls
are specific to each state and each month, we modified the model to
provide the estimated risk of V. vulnificus illness for each state and
each month. In particular, we used monthly, rather than seasonal,
parameters that were reported in the WHO/FAO risk assessment to
estimate water temperature and the difference between water
temperature and air temperature. In addition, the WHO/FAO risk
assessment reported parameters for the distribution of the number of
hours that oysters are left unrefrigerated separately for Louisiana
and for the rest of the Gulf Coast states. We applied these parameters
to our simulation, using one set of parameters for Louisiana and
another set of parameters for Florida and Texas. The WHO/FAO risk
assessment model's estimates are based on the average of 100 samples
of 10,000 observations each. To provide more reliable uncertainty
intervals for these estimates, our modification of the WHO/FAO model
uses 1,000 samples.
Time and Temperature Controls for Florida, Louisiana, and Texas and
Our Simulations of Various Compliance Rates with the Controls:
Effective May 1, 2010, Florida, Louisiana, and Texas implemented new,
more stringent time and temperature controls that specify (1) the
maximum number of hours that oysters are allowed to be unrefrigerated
after being harvested and (2) the maximum number of hours before
refrigerated oysters must cool down to 55 degrees Fahrenheit. The new
more stringent controls established by the three states for 2010 and
incorporated in their risk management plans are presented in table 3
for each state and each month. These controls are stricter during the
warmer months when V. vulnificus bacteria multiply more quickly. In
August, for example, Louisiana and Texas have the most restrictive
controls for the time oysters could remain unrefrigerated, allowing 1
hour from harvest until refrigeration, and Florida has the most
restrictive controls for the time until refrigerated oysters must cool
down, allowing 2 hours from when they are first refrigerated until
they reach 55 degrees Fahrenheit.
Table 3: Time and Temperature Controls for 2010, by State and Month:
Maximum number of hours:
Month: January;
From harvest to refrigeration, Florida: 18;
From harvest to refrigeration, Louisiana: 36;
From harvest to refrigeration, Texas: 18;
From refrigeration to cooldown to 55 degrees Fahrenheit, Florida: 10;
From refrigeration to cooldown to 55 degrees Fahrenheit, Louisiana: 10;
From refrigeration to cooldown to 55 degrees Fahrenheit, Texas: 10.
Month: February;
From harvest to refrigeration, Florida: 18;
From harvest to refrigeration, Louisiana: 36;
From harvest to refrigeration, Texas: 18;
From refrigeration to cooldown to 55 degrees Fahrenheit, Florida: 10;
From refrigeration to cooldown to 55 degrees Fahrenheit, Louisiana: 10;
From refrigeration to cooldown to 55 degrees Fahrenheit, Texas: 10.
Month: March;
From harvest to refrigeration, Florida: 18;
From harvest to refrigeration, Louisiana: 8;
From harvest to refrigeration, Texas: 18;
From refrigeration to cooldown to 55 degrees Fahrenheit, Florida: 10;
From refrigeration to cooldown to 55 degrees Fahrenheit, Louisiana: 10;
From refrigeration to cooldown to 55 degrees Fahrenheit, Texas: 10.
Month: April;
From harvest to refrigeration, Florida: 12;
From harvest to refrigeration, Louisiana: 8;
From harvest to refrigeration, Texas: 10;
From refrigeration to cooldown to 55 degrees Fahrenheit, Florida: 10;
From refrigeration to cooldown to 55 degrees Fahrenheit, Louisiana: 10;
From refrigeration to cooldown to 55 degrees Fahrenheit, Texas: 10.
Month: May;
From harvest to refrigeration, Florida: 6;
From harvest to refrigeration, Louisiana: 1;
From harvest to refrigeration, Texas: 4;
From refrigeration to cooldown to 55 degrees Fahrenheit, Florida: 2;
From refrigeration to cooldown to 55 degrees Fahrenheit, Louisiana: 6;
From refrigeration to cooldown to 55 degrees Fahrenheit, Texas: 6.
Month: June;
From harvest to refrigeration, Florida: 6;
From harvest to refrigeration, Louisiana: 1;
From harvest to refrigeration, Texas: 3;
From refrigeration to cooldown to 55 degrees Fahrenheit, Florida: 2;
From refrigeration to cooldown to 55 degrees Fahrenheit, Louisiana: 6;
From refrigeration to cooldown to 55 degrees Fahrenheit, Texas: 6.
Month: July;
From harvest to refrigeration, Florida: 6;
From harvest to refrigeration, Louisiana: 1;
From harvest to refrigeration, Texas: 1;
From refrigeration to cooldown to 55 degrees Fahrenheit, Florida: 2;
From refrigeration to cooldown to 55 degrees Fahrenheit, Louisiana: 6;
From refrigeration to cooldown to 55 degrees Fahrenheit, Texas: 6.
Month: August;
From harvest to refrigeration, Florida: 6;
From harvest to refrigeration, Louisiana: 1;
From harvest to refrigeration, Texas: 1;
From refrigeration to cooldown to 55 degrees Fahrenheit, Florida: 2;
From refrigeration to cooldown to 55 degrees Fahrenheit, Louisiana: 6;
From refrigeration to cooldown to 55 degrees Fahrenheit, Texas: 6.
Month: September;
From harvest to refrigeration, Florida: 6;
From harvest to refrigeration, Louisiana: 1;
From harvest to refrigeration, Texas: 3;
From refrigeration to cooldown to 55 degrees Fahrenheit, Florida: 2;
From refrigeration to cooldown to 55 degrees Fahrenheit, Louisiana: 6;
From refrigeration to cooldown to 55 degrees Fahrenheit, Texas: 6.
Month: October;
From harvest to refrigeration, Florida: 6;
From harvest to refrigeration, Louisiana: 1;
From harvest to refrigeration, Texas: 4;
From refrigeration to cooldown to 55 degrees Fahrenheit, Florida: 2;
From refrigeration to cooldown to 55 degrees Fahrenheit, Louisiana: 6;
From refrigeration to cooldown to 55 degrees Fahrenheit, Texas: 6.
Month: November;
From harvest to refrigeration, Florida: 18;
From harvest to refrigeration, Louisiana: 8;
From harvest to refrigeration, Texas: 10;
From refrigeration to cooldown to 55 degrees Fahrenheit, Florida: 10;
From refrigeration to cooldown to 55 degrees Fahrenheit, Louisiana: 10;
From refrigeration to cooldown to 55 degrees Fahrenheit, Texas: 10.
Month: December;
From harvest to refrigeration, Florida: 18;
From harvest to refrigeration, Louisiana: 36;
From harvest to refrigeration, Texas: 18;
From refrigeration to cooldown to 55 degrees Fahrenheit, Florida: 10;
From refrigeration to cooldown to 55 degrees Fahrenheit, Louisiana: 10;
From refrigeration to cooldown to 55 degrees Fahrenheit, Texas: 10.
Source: GAO analysis of Florida, Louisiana, and Texas 2010 risk
management plans.
[A] The number of hours to cooldown for Florida refers to the rapid
cooling control option, one of four time and temperature control
options authorized in Florida's V. vulnificus risk management plan and
the only option used during the summer of 2010 by the Florida oyster
industry, according to a senior Florida regulatory official.
[End of table]
For the purpose of this analysis, we define the baseline scenario as
the risk of illness in the absence of the new, more stringent time and
temperature controls. Under the baseline scenario, the WHO/FAO risk
simulation model assumes a certain statistical distribution in the
number of hours that oysters ordinarily would be left unrefrigerated.
The values in the statistical distribution, which is based on
assumptions in the risk simulation model, range from 1 hour to 10 or
more hours, depending upon the state and the month, and specifies the
percentage of oysters that ordinarily would be left unrefrigerated for
any given number of hours within this range. Based on this statistical
distribution, we estimated the percentage of oysters that ordinarily--
that is, in the absence of the new, more stringent time and
temperature controls--would be refrigerated within the maximum number
of hours established by time and temperature controls for each state
and each month. In states and months with the least stringent
controls, a majority of oysters ordinarily would be refrigerated
within these time limits, even in the absence of these time and
temperature controls. For example, in Florida during the three summer
months, according to the assumed statistical distribution,
approximately 85 percent of oysters harvested ordinarily would be
refrigerated within the 6-hour limit established by the new, more
stringent time and temperature controls. By contrast, in states and
months with the most stringent controls, fewer oysters harvested would
ordinarily be refrigerated within these limits. For example, in
Louisiana and Texas in August, virtually none of the harvested oysters
ordinarily would be refrigerated within the 1-hour limit established
by these time and temperature controls, according to the assumed
statistical distribution.
To simulate compliance with the maximum number of hours that oysters
are allowed to be unrefrigerated under applicable time and temperature
controls, we make the following assumptions about the behavior of
oyster harvesters. First, harvesters that ordinarily--that is, under
the baseline scenario--would leave oysters unrefrigerated for less
than the maximum number of hours would continue to leave them
unrefrigerated for the same number of hours that they ordinarily would
have. Second, harvesters who ordinarily would leave oysters
unrefrigerated for more than the maximum number of hours, and who
decide to change to comply with time and temperature controls, would
leave oysters unrefrigerated for no more than the maximum allowed
number of hours. Third, harvesters who ordinarily would leave oysters
unrefrigerated for more than the maximum number of hours, but who
decide not to change to comply with time and temperature controls,
would continue to leave oysters unrefrigerated for the same number of
hours that they ordinarily would have. Similarly, to model the impact
of compliance on the number of hours until oysters reach the desired
55 degrees Fahrenheit, we assumed that producers would facilitate more
rapid cooling so that oysters would take no longer than the maximum
number of hours to cooldown.
Using these assumptions, we developed 10 compliance scenarios for each
state and each month. These scenarios correspond to estimated
compliance rates of 10 percent through 100 percent in increments of
10. Under these scenarios, the model first estimates the percentage of
oysters that ordinarily would be refrigerated within the maximum
number of hours established by time and temperature controls for each
state and each month. To obtain a given compliance rate, the model
calculates the additional percentage of oysters that would need to be
refrigerated within the maximum number of hours by time and
temperature controls to reach a given rate of compliance with regard
to the maximum time allowed to be unrefrigerated. For this additional
percentage of oysters, the model assumes that oysters would be
refrigerated within the maximum number of hours allowed by the
controls for that state and month. In Florida during the three summer
months, for example, 85 percent of oysters are assumed to be
refrigerated within the 6-hour limit in the absence of time and
temperature controls, based on the assumed statistical distribution.
To attain a 90 percent compliance rate, the model would select the
additional 5 percent of oysters, from among the 15 percent that exceed
the limit, and would assume that these oysters would be unrefrigerated
for no longer than 6 hours. Since actual compliance rates are unknown,
these calculations allow us to estimate the number of hours that
oysters would be unrefrigerated assuming various compliance rates.
Estimates of the Number of V. Vulnificus Illnesses Using FDA's Risk
Calculator:
The three states used FDA's risk calculator and their own input data,
including water temperature and air temperature for each month, to
establish the specific limits for time and temperature controls. The
risk calculator, which was developed by FDA, is a simplified version
of the WHO/FAO risk simulation model and operates in a computer
spreadsheet. It allows the user to estimate the risk of illness from
V. vulnificus under various scenarios, such as different limits for
the maximum number of hours that oysters can left be unrefrigerated.
To determine the estimated number of illnesses per 100,000 servings of
raw oysters (i.e., risk of illness) consumed by the susceptible
population for each state and month under time and temperature
controls, we used FDA's risk calculator. To make the results of our
analysis comparable across the states and consistent with the
assumptions of the baseline scenario, we used the input data for water
temperature and air temperature from the WHO/FAO risk simulation
model, rather than the data used by the states. As a result, our
estimates of the risk of illness differ somewhat from the estimates
that the states made in using the risk calculator to develop their
risk management plans.
Unlike the states, we estimated the number of illnesses per 100,000
servings of raw oysters consumed by the susceptible population, rather
than the total number of illnesses, because (1) time and temperature
controls are designed to affect the risk of illness per serving, not
the total number of raw oyster servings consumed and (2) complete
state-by-state and month-by-month data on the number of raw oyster
servings consumed were not available. The estimated number of
illnesses per 100,000 servings for each state and each month, as
computed by the risk calculator, is presented in table 4. These
estimates represent the number of illnesses that states would expect,
based on the risk calculator, as a result of time and temperature
controls. We compared these estimated numbers of illnesses, for each
state and each month, to the estimates of our modification of the risk
simulation model, which accounts for uncertainty in the estimates and
for various compliance rates.
Effect of Uncertainty and Compliance Rates on the Likely Impact of
Time and Temperature Controls on the Number of V. Vulnificus Illnesses:
FDA's risk calculator estimates the same number of V. vulnificus
illnesses as the WHO/FAO risk simulation model, on average. Unlike the
risk simulation model, however, the risk calculator does not provide
uncertainty distributions associated with these estimates. Using our
modification of the WHO/FAO risk simulation model, we computed the
amount of uncertainty associated with estimates made by FDA's risk
calculator. In any given month and in any given state, uncertainty in
model assumptions may cause the actual number of illnesses to differ
from the number estimated by the risk calculator. For example, in
Texas in the month of August, FDA's risk calculator estimates that
time and temperature controls will lead to 2.84 illnesses per 100,000
raw oyster servings consumed by the susceptible population. While this
is true on average, the number of illnesses per 100,000 servings could
vary from the lower bound of 2.44 to the upper bound 3.63 (for a 90
percent uncertainty interval), according to our analysis. We find a
similar range of uncertainty in the estimated number of V. vulnificus
illnesses for Florida and Louisiana. Table 4 presents our estimates
compared with estimates made by the risk calculator, assuming 100
percent compliance with time and temperature controls for the three
states during the summer months.
Table 4: Comparison of FDA's Risk Calculator and GAO's Modification of
WHO/FAO Risk Simulation Model Concerning Estimated Number of V.
Vulnificus Illnesses per 100,000 Servings of Oysters at Baseline:
Florida:
Month: July;
FDA risk calculator--mean[A]: 3.46;
GAO's modification of WHO/FAO risk simulation model--lower bound[B]:
2.87;
GAO's modification of WHO/FAO risk simulation model--mean: 3.47;
GAO's modification of WHO/FAO risk simulation model--upper bound[C]:
4.23.
Month: August;
FDA risk calculator--mean[A]: 3.45;
GAO's modification of WHO/FAO risk simulation model--lower bound[B]:
3.00;
GAO's modification of WHO/FAO risk simulation model--mean: 3.53;
GAO's modification of WHO/FAO risk simulation model--upper bound[C]:
4.30.
Month: September;
FDA risk calculator--mean[A]: 3.05;
GAO's modification of WHO/FAO risk simulation model--lower bound[B]:
2.51;
GAO's modification of WHO/FAO risk simulation model--mean: 3.05;
GAO's modification of WHO/FAO risk simulation model--upper bound[C]:
3.69.
Louisiana:
Month: July;
FDA risk calculator--mean[A]: 2.84;
GAO's modification of WHO/FAO risk simulation model--lower bound[B]:
2.33;
GAO's modification of WHO/FAO risk simulation model--mean: 2.89;
GAO's modification of WHO/FAO risk simulation model--upper bound[C]:
3.53.
Month: August;
FDA risk calculator--mean[A]: 2.84;
GAO's modification of WHO/FAO risk simulation model--lower bound[B]:
2.44;
GAO's modification of WHO/FAO risk simulation model--mean: 2.95;
GAO's modification of WHO/FAO risk simulation model--upper bound[C]:
3.63.
Month: September;
FDA risk calculator--mean[A]: 2.52;
GAO's modification of WHO/FAO risk simulation model--lower bound[B]:
2.02;
GAO's modification of WHO/FAO risk simulation model--mean: 2.54;
GAO's modification of WHO/FAO risk simulation model--upper bound[C]:
3.11.
Texas:
Month: July;
FDA risk calculator--mean[A]: 2.84;
GAO's modification of WHO/FAO risk simulation model--lower bound[B]:
2.33;
GAO's modification of WHO/FAO risk simulation model--mean: 2.89;
GAO's modification of WHO/FAO risk simulation model--upper bound[C]:
3.53.
Month: August;
FDA risk calculator--mean[A]: 2.84;
GAO's modification of WHO/FAO risk simulation model--lower bound[B]:
2.44;
GAO's modification of WHO/FAO risk simulation model--mean: 2.95;
GAO's modification of WHO/FAO risk simulation model--upper bound[C]:
3.63.
Month: September;
FDA risk calculator--mean[A]: 2.85;
GAO's modification of WHO/FAO risk simulation model--lower bound[B]:
2.33;
GAO's modification of WHO/FAO risk simulation model--mean: 2.86;
GAO's modification of WHO/FAO risk simulation model--upper bound[C]:
3.48.
Source: GAO analysis with FDA's risk calculator and GAO's modification
of the WHO/FAO risk simulation model.
Notes: These estimates assume 100 percent compliance with state time
and temperature controls.
[A] Estimates made with the FDA's risk calculator do not produce
uncertainty intervals, so we only report a mean estimate, rather than
a mean plus a lower bound and an upper bound.
[B] The lower bound is based on the 5th percentile of the distribution
of risk calculations from our modification of the WHO/FAO risk
simulation model.
[C] The upper bound is based on the 95th percentile of the
distribution of risk calculations from our modification of the WHO/FAO
risk simulation model.
[End of table]
Furthermore, we estimate that time and temperature controls would
result in a smaller reduction in the number of V. vulnificus illnesses
if compliance rates are less than 100 percent. Table 5 shows the
estimated reduction in V. vulnificus illnesses as a result of time and
temperature controls for various compliance rates for each of the
three states during the summer months, based on our modification of
the WHO/FAO risk simulation model. For example, during August,
assuming that 100 percent of oysters are harvested in compliance with
time and temperature controls, our analysis estimates these controls
will reduce illnesses by between 16 percent and 27 percent in Florida,
between 30 percent and 47 percent in Louisiana, and between 26 percent
and 43 percent in Texas. If compliance is less than 100 percent,
however, we estimate that these controls will lead to a much smaller
reduction in illnesses. As can be seen in table 5, if 90 percent of
oysters are harvested in compliance with time and temperature
controls--meaning a noncompliance rate of 10 percent--the illness
reduction is smaller than the illness reduction under the assumption
of 100 percent compliance. For example, in the month of August, we
estimate that illnesses would be reduced between 11 percent and 18
percent in Florida, between 21 percent and 32 percent in Louisiana,
and between 19 percent and 31 percent in Texas, assuming 90 percent
compliance. Furthermore, if 80 percent of oysters are harvested in
compliance with these controls--meaning that noncompliance rates are
20 percent--the estimated illness reduction is smaller still. In
particular, we estimate that illnesses would be reduced between 8
percent and 14 percent in Florida, between 15 percent and 23 percent
in Louisiana, and between 14 percent and 22 percent in Texas. At lower
levels of compliance, an even smaller reduction in the number of V.
vulnificus illnesses is likely.
Table 5: Estimated Percentage Reduction in the Number of V. Vulnificus
Illnesses as a Result of Time and Temperature Controls, by State and
Month:
Florida:
July;
100 percent compliance:
Lower bound[A]: 17;
Mean: 22;
100 percent compliance:
90 percent compliance:
Lower bound[A]: 11;
Mean: 15;
Upper bound[B]: 18;
80 percent compliance[C]:
Lower bound[A]: 8;
Mean: 11;
Upper bound[B]: 14;
70 percent compliance[C]:
Lower bound[A]: 7;
Mean: 9;
Upper bound[B]: 11.
August;
100 percent compliance:
Lower bound[A]: 16;
Mean: 21;
Upper bound[B]: 27;
90 percent compliance:
Lower bound[A]: 11;
Mean: 14;
Upper bound[B]: 18;
80 percent compliance[C]:
Lower bound[A]: 8;
Mean: 11;
Upper bound[B]: 14;
70 percent compliance[C]:
Lower bound[A]: 7;
Mean: 9;
Upper bound[B]: 11.
September;
100 percent compliance:
Lower bound[A]: 17;
Mean: 22;
Upper bound[B]: 29;
90 percent compliance:
Lower bound[A]: 12;
Mean: 16;
Upper bound[B]: 20;
80 percent compliance[C]:
Lower bound[A]: 9;
Mean: 11;
Upper bound[B]: 15;
70 percent compliance[C]:
Lower bound[A]: 7;
Mean: 9;
Upper bound[B]: 12.
Louisiana:
July;
100 percent compliance:
Lower bound[A]: 31;
Mean: 39;
Upper bound[B]: 48;
90 percent compliance:
Lower bound[A]: 21;
Mean: 26;
Upper bound[B]: 32;
80 percent compliance[C]:
Lower bound[A]: 15;
Mean: 19;
Upper bound[B]: 23;
70 percent compliance[C]:
Lower bound[A]: 11;
Mean: 14;
Upper bound[B]: 17.
August;
100 percent compliance:
Lower bound[A]: 30;
Mean: 38;
Upper bound[B]: 47;
90 percent compliance:
Lower bound[A]: 21;
Mean: 26;
Upper bound[B]: 32;
80 percent compliance[C]:
Lower bound[A]: 15;
Mean: 19;
Upper bound[B]: 24;
70 percent compliance[C]:
Lower bound[A]: 11;
Mean: 14;
Upper bound[B]: 17.
September;
100 percent compliance:
Lower bound[A]: 31;
Mean: 39;
Upper bound[B]: 49;
90 percent compliance:
Lower bound[A]: 23;
Mean: 28;
Upper bound[B]: 35;
80 percent compliance[C]:
Lower bound[A]: 16;
Mean: 20;
Upper bound[B]: 25;
70 percent compliance[C]:
Lower bound[A]: 11;
Mean: 14;
Upper bound[B]: 18.
Texas:
July;
100 percent compliance:
Lower bound[A]: 27;
Mean: 35;
Upper bound[B]: 43;
90 percent compliance:
Lower bound[A]: 19;
Mean: 24;
Upper bound[B]: 30;
80 percent compliance[C]:
Lower bound[A]: 14;
Mean: 18;
Upper bound[B]: 22;
70 percent compliance[C]:
Lower bound[A]: 10;
Mean: 13;
Upper bound[B]: 15.
August;
100 percent compliance:
Lower bound[A]: 26;
Mean: 34;
Upper bound[B]: 43;
90 percent compliance:
Lower bound[A]: 19;
Mean: 25;
Upper bound[B]: 31;
80 percent compliance[C]:
Lower bound[A]: 14;
Mean: 18;
Upper bound[B]: 22;
70 percent compliance[C]:
Lower bound[A]: 10;
Mean: 13;
Upper bound[B]: 16.
September;
100 percent compliance:
Lower bound[A]: 21;
Mean: 27;
Upper bound[B]: 34;
90 percent compliance:
Lower bound[A]: 15;
Mean: 19;
Upper bound[B]: 24;
80 percent compliance[C]:
Lower bound[A]: 10;
Mean: 13;
Upper bound[B]: 17;
70 percent compliance[C]:
Lower bound[A]: 6;
Mean: 8;
Upper bound[B]: 11.
Source: GAO modification of the WHO/FAO risk simulation model.
[A] The lower bound is based on the 5th percentile of the distribution
of calculations.
[B] The upper bound is based on the 95th percentile of the
distribution of calculations.
[C] Under the baseline scenario, approximately 85 percent of oysters
in Florida in July, August, and September are assumed to be
refrigerated within the maximum number of hours allowed by time and
temperature controls. As a result, we assume that compliance rates for
Florida during the summer will be no lower than 85 percent for time-to-
refrigeration controls. Therefore, the 80 percent and 70 percent
compliance scenarios for Florida refer only to compliance with
refrigeration to cooldown controls.
[End of table]
Because time and temperature controls are less effective at lower
compliance rates, the probability that these controls will lead to the
illness reduction estimated by the risk calculator is also lower when
compliance rates are lower. Table 6 shows the probability that time
and temperature controls will reduce V. vulnificus illnesses to the
number estimated by FDA's risk calculator or lower for various
compliance rates for each of the three states, based on our risk
simulation model. During the summer, assuming that 100 percent of
oysters are harvested in compliance with time and temperature
controls, as would be expected with the risk calculator's design,
there is between a 43 percent chance and a 55 percent chance that
these controls will reduce illnesses to the number estimated by the
risk calculator or lower, depending on the state and the month. If
compliance is less than 100 percent, however, our analysis shows that
it is unlikely that time and temperature controls will reduce
illnesses to the number estimated by the risk calculator or lower. As
can be seen in table 5, if 90 percent of oysters are harvested in
compliance with time and temperature controls--meaning a noncompliance
rate of 10 percent--the chances that these controls will reduce
illnesses to the estimated number or lower drop substantially when
compared with the chances under the assumption of 100 percent
compliance. In particular, for the month of August, we estimate that
the probability drops from 48 percent to 18 percent in Florida, from
43 percent to 2 percent in Louisiana, and from 43 percent to 4 percent
in Texas. Furthermore, if 80 percent of oysters are harvested in
compliance with these controls--meaning that noncompliance rates are
20 percent--the likelihood of success is smaller still. In particular,
we estimate that the probability that these controls will reduce
illnesses to the number estimated by the risk calculator or lower
drops to 7 percent in Florida, less than 1 percent in Louisiana, and
1 percent in Texas.
Table 6: Probability That the Time and Temperature Controls Will
Reduce V. Vulnificus Illnesses to the Number Estimated by FDA's Risk
Calculator or Lower, by Estimated Compliance Rate[A]:
Florida:
Month: July;
Estimated compliance rate[B]:
100 percent: 0.54;
Estimated compliance rate[B]:
90 percent: 0.22;
80 percent: 0.11[C];
70 percent: 0.08[C].
Month: August;
Estimated compliance rate[B]:
100 percent: 0.48;
90 percent: 0.18;
80 percent: 0.07[C];
70 percent: 0.05.
Month: September;
Estimated compliance rate[B]:
100 percent: 0.55;
90 percent: 0.24;
80 percent: 0.10[C];
70 percent: 0.07[C].
Louisiana:
Month: July;
Estimated compliance rate[B]:
100 percent: 0.47;
90 percent: 0.03;
80 percent: 0.00;
70 percent: 0.00.
Month: August;
Estimated compliance rate[B]:
100 percent: 0.43;
90 percent: 0.02;
80 percent: 0.00;
70 percent: 0.00.
Month: September;
Estimated compliance rate[B]:
100 percent: 0.50;
90 percent: 0.06;
80 percent: 0.00;
70 percent: 0.00.
Texas:
Month: July;
Estimated compliance rate[B]:
100 percent: 0.47;
90 percent: 0.07;
80 percent: 0.01;
70 percent: 0.00.
Month: August;
Estimated compliance rate[B]:
100 percent: 0.43;
90 percent: 0.04;
80 percent: 0.01;
70 percent: 0.00.
Month: September;
Estimated compliance rate[B]:
100 percent: 0.52;
90 percent: 0.17;
80 percent: 0.04;
70 percent: 0.01.
Source: GAO modification of the WHO/FAO risk simulation model.
[A] Probabilities are expressed between 0.00 and 1.00, with values
close to 0.00 being highly unlikely and values close to 1.00 being
highly likely.
[B] Compliance rates are estimated based on the distribution of time
to refrigeration and time to cooldown, which is used in both the
WHO/FAO risk simulation and in GAO's modification of the WHO/FAO risk
simulation model, in combination with the thresholds established by
time and temperature controls.
[C] Under the baseline scenario, approximately 85 percent of oysters
in Florida in July, August, and September are assumed to be
refrigerated within the maximum number of hours allowed by time and
temperature controls. As a result, we assume that compliance rates for
Florida during the summer will be no lower than 85 percent for time-to-
refrigeration controls. Therefore, the 80 percent and 70 percent
compliance scenarios for Florida refer only to compliance with
refrigeration to cooldown controls.
[End of table]
Limitations to Our Analysis:
Like all quantitative models, our analysis is subject to certain
limitations. First, our analysis is subject to all of the limitations
to which the WHO/FAO risk simulation model is subject. Though the WHO/
FAO model is based on credible scientific studies, uses a valid and
reliable methodology, and predicts actual illnesses rates with
reasonable accuracy, it is subject to limitations just as all
quantitative models are. For example, the model assumes that V.
vulnificus levels at harvest are only determined by water temperature,
that all strains of V. vulnificus are equally virulent, and that the
risk of infection is identical for all members of the susceptible
population, though these are simplifications. Furthermore, the exact
relationship between levels of V. vulnificus and the observed number
of illness is not known, and there are no precise estimates of the
size of the susceptible population. Second, our simulations of
compliance rates are based on certain assumptions about handling of
oysters under the baseline scenario--including the number of hours
that oysters would be unrefrigerated and the number of hours until
oysters cool down--and on certain assumptions about how producers
might respond to time and temperature controls under various
compliance scenarios. Since we do not have direct data on actual
compliance rates, however, our estimates are only an approximation and
cannot be validated against observed data. Third, our estimates of the
probability that time and temperature controls will lead to the levels
of illness estimated by the risk calculator are approximations and are
a function of the data inputs, assumptions, and equations in the risk
simulation model. In spite of these limitations, however, we believe
our estimates are sufficiently reliable to demonstrate that there is a
substantial chance that time and temperature controls will not lead to
the number of V. vulnificus illnesses estimated by the risk calculator
or lower, especially with less than perfect compliance rates.
[End of section]
Appendix II: Comments from the Department of Health and Human Services:
Department Of Health & Human Services:
Office Of The Secretary:
Assistant Secretary for Legislation:
Washington, DC 20201:
August 15, 2011:
Lisa Shames:
Director, Natural Resources and Environment:
U.S. Government Accountability Office:
441 G Street N.W.
Washington, DC 20548:
Dear Mr. Kohn:
Attached are comments on the U.S. Government Accountability Office's
(GAO) draft report entitled: "Food Safety: FDA Needs to Reassess Its
Approach to Reducing an Illness Caused by Eating Raw Oysters" (GA0-11-
607).
The Department appreciates the opportunity to review this report
before its publication.
Sincerely,
Signed by:
Jim R. Esquea:
Assistant Secretary for Legislation:
Attachment:
[End of letter]
General Comments Of The Department Of Health And Human Services (HHS)
On The Government Accountability Office'S (GAO) Draft Report Entitled.
"Food Safety: FDA Needs To Reassess Its Approach To Reducing An
Illness Caused By Eating Raw Oysters" (GAO-11-607):
The Department appreciates the opportunity to review and comment on
this draft report. Vibrio Vulnificus (V. vulnificus) is a naturally
occurring bacterium that can cause a severe and life threatening
illness that is fatal about 50 % of the time, generally causing about
15 deaths per year. V. vulnificus is associated with the consumption
of raw oysters and characterized by fever and chills, decreased blood
pressure (septic shock), and blistering skin lesions. At greatest risk
are individuals whose immune systems have been compromised or who have
certain health conditions, such as liver, stomach, or blood disorders;
cancer; AIDS; diabetes; kidney disease; and chronic alcohol abuse.
Effective technologies have been developed that can largely eliminate
the hazard of V. vulnificus while producing oysters that retain the
sensory qualities of untreated product. These technologies, known as
Post Harvest Processing (PHP), include individual quick freezing (IQF)
with extended frozen storage, high hydrostatic pressure, mild heat,
and low dose gamma irradiation. PHP technologies have proven to be
effective in eradicating V. vulnificus associated illness. For
example, in 2003, the State of California prohibited Gulf Coast
oysters from entering the state during the season of greatest risk
unless they had undergone PHP. Once PHP was required in California,
the number of deaths in the state fell from 40, between 1991 to 2001,
to nearly zero since then. California's PHP requirement has virtually
eliminated the state's V. vulnificus-related deaths and illness from
consuming raw oysters.
The Food and Drug Administration (FDA) has collaborated with the
Interstate Shellfish Sanitation Conference (ISSC) for years to reduce
V. vulnificus illness through improving consumer education and
refrigeration practices, but these practices have failed to achieve
measurable reductions of V. vulnificus illnesses nationally. FDA has
proposed the implementation of PHP, or other equivalent controls, to
substantially reduce V. vulnificus illness, but the Gulf Coast
industry, state officials, and elected representatives have raised
concerns about implementing PHP controls. FDA has considered these
concerns and recognizes the need to further examine the timing and
processes for oyster harvesters to gain access to PHP facilities or
equivalent controls. To that end, FDA commissioned an independent
study to assess how PHP or other equivalent controls can be
implemented in a safe, efficient, and economic manner and will be
addressing the concerns related to that study raised by GAO in an
addendum to that study. FDA will continue to collaborate and dialog
with industry, state officials, and the ISSC to explore reasonable and
workable approaches to substantially reduce V. vulnificus illness and
protect the American people from this painful, deadly and preventable
disease.
FDA's responses to GAO's recommendation are set forth below:
GAO Recommendations:
To better ensure the safety of oysters from the Gulf of Mexico that
are sold for raw consumption, we recommend that the Commissioner of
FDA work with the Executive Board of the ISSC to take the following
four actions:
* agree on a nationwide goal for reducing the number of V. vulnificus
illnesses caused by the consumption of Gulf Coast raw oysters and
develop strategies to achieve that goal, recognizing that consumer
education and time and temperature controls have not resulted in
achievement of the 60 percent V. vulnificus illness rate reduction
goal and that the capacity to use post-harvest processing (PHP) on
Gulf Coast oysters harvested from April through October that are
intended for raw consumption does not currently exist;
FDA Response:
The ISSC has attempted to achieve the 60% illness reduction goal that
had been established in 2001 through improved refrigeration practices,
limited PHP and consumer education, but these efforts have not
succeeded. FDA recognizes the efforts that went into these
undertakings and will continue to collaborate with the ISSC to find
strategies and explore approaches to establishing reasonable and
workable goals for reducing V. vulnificus illness and protecting
Americans from this deadly and preventable disease. As FDA continues
in these efforts the agency remains mindful that effective
technologies have been developed that can largely eliminate the hazard
of V. vulnificus while producing oysters that retain the sensory
qualities of untreated product.
* correct the limitations in the current approach to measuring
progress toward the 60 percent V. vulnificus illness rate reduction
goal or design and implement a new approach that does not have the
limitations of the current one;
FDA Response:
FDA agrees that the current approach used by the ISSC to count V.
vulnificus illnesses and assess illness rate reduction is defective
and should be corrected. The evaluation of success of existing control
measures is based on counting illnesses reported in four "core" states
(CA, LA, TX, FL). For a number of years, FDA has advised the ISSC of
concerns with that approach. While the ISSC has claimed some success
in its effort to reduce V. vulnificus illnesses, using numbers for the
four "core" states, the rate of illness at the national level has
remained relatively static. Much of the success claimed by the ISSC is
directly attributable to the 2003 California ban on raw, untreated
Gulf oysters. That ban virtually eliminated oyster associated V.
vulnificus illnesses in California, which previously reported 5 to 6
annually. Continued use of California as a "core" state in the ISSC's
illness counting system biases the calculated illness reduction rate.
Even if the ISSC's 60% goal had been achieved, it is unlikely that a
measurable reduction in the rate of illness nationally would have been
realized. This presented itself as a significant factor in FDA's
announcement of its intent to revise its policy and issue guidance
regarding PHP. FDA wishes to continue working with the ISSC to develop
a counting formula that accounts for illness nationally and that
realistically defines how effective V. vulnificus control measures
are, whatever they include.
* regularly evaluate the effectiveness of V. vulnificus illness
reduction strategies, such as consumer education and time and
temperature controls, to determine whether they are successful and
should be continued or are ineffective and should be stopped;
FDA Response:
FDA agrees with GAO that the approach that has been used to evaluate
the effectiveness of illness reduction strategies has limitations that
undermine its credibility, including the limited number of states used
in determining V. vulnificus illness reduction, and the overstatement
of the effectiveness of the primary V. vulnificus illness reduction
strategies, consumer education and time and temperature controls”-by
including V. vulnificus illness data from California.
Historically FDA, ISSC and the States have devoted significant
resources to conducting V. vulnificus education campaigns. Directed at
the consuming public, these activities have been aimed at informing at-
risk consumers about the risks of consuming raw molluscan shellfish.
Campaign efforts have also targeted health professionals who provide
care to at risk individuals, including those with underlying medical
conditions, such as liver disease and chronic alcohol abuse While FDA
has not undertaken a study to specifically examine the impact of
educational programs, there is no indication that they have resulted
in any substantial reduction in the occurrence of V. vulnificus
illnesses, as evidenced by the relatively static level of illnesses
and deaths occurring each year nationally. A survey commissioned by
the ISSC in 2004 does not suggest a reduction in the number of at-risk
consumers who are consuming raw oysters and there is no evidence of
illness reduction at the national level. Furthermore, even though the
independent impact of education on the rate of illness cannot be
measured, the impact appears to be marginal at best given that the
current illness reduction rate (based on 2009 and 2010 data) is only
38.8% in the "core' states. That rate of reduction is significantly
skewed by the use of California as a counting state. FDA has concluded
that additional efforts to educate will have little if any beneficial
outcome.
With regard to assessing the effectiveness of existing controls on
illness reduction, it is extremely difficult, and perhaps impossible,
to tease out the contribution of one control measure versus another.
For that reason, ISSC goals have relied on illness counting to
determine their success. Unfortunately, the counting strategy employed
by the ISSC is flawed, for reasons previously discussed and pointed
out by GAO. Studies to compare V. vulnificus levels in retail oysters
subsequent to states' implementation of time and temperature controls
to levels found in previous retail studies may help identify levels of
consumer exposure. However, FDA has no plans to conduct additional
studies given existing budgetary and competing priority
considerations. One thing that remains clear is implementation of
strict time and temperature controls by states has not achieved the
ISSC 60% illness rate reduction goal. Nor have these controls resulted
in any illness reduction at the national level. Arguments have been
put forth suggesting that industry compliance is problematic and that
increased effort by states and FDA to enforce compliance is needed.
Toward that end, FDA is moving from biennial to annual evaluation of
V. vulnificus control plans being used by states and industry. As part
of its increased compliance evaluation, FDA will conduct annual onsite
checks at oyster landing sites and processing plants to examine
compliance with V. vulnificus HACCP controls, harvester records,
time/temperature logs, and actual product temperatures. Such efforts
will help address concerns that the goal has not been met due to
inadequate implementation and enforcement of controls.
* conduct further study of the six issues of concern that we
identified regarding the RTI report's economic analysis to ensure a
more accurate assessment of the feasibility of developing adequate
capacity and before FDA and the ISSC move forward with revising the
National Shellfish Sanitation Program's shellfish safety guidelines to
provide postharvest processing for oysters harvested from Gulf Coast
waters during warmer months and intended for raw consumption.
FDA Response:
The 6 issues of concern identified by GAO are as follows:
* Baseline data may not be representative of industry;
FDA disagrees with the argument that baseline data, upon which the
study is premised, is not representative. Data for 2008 are
representative of a typical year, in which natural or manmade
disasters are not of impact. As such, 2008 serves well as a baseline
for what a "normal" year in the Gulf historically represents. Use of
data for 2010 would not have represented a typical harvest year due to
the Gulf oil spill disaster that reduced harvest levels due to
closures in many Gulf Coast harvest areas. This circumstance would
have skewed the results, possibly underestimating the impact and cost
associated with PHP. Furthermore, to have waited until more recent
data was available, and for what would be representative of a "normal"
year, would have delayed efforts by FDA to examine the feasibility of
PHP. Moreover, according to RTI, the overall conclusions of their
study likely would not have changed.
* Key costs are excluded;
FDA recognizes that exclusion of certain costs can and have affected
final cost outcomes presented in the RTI report. In an effort to
better assess how costs associated with needs such as land purchase,
new facility construction, transportation, and insurance, FDA has
commissioned additional work to address these cost considerations.
* Who should pay to expand processing capacity is not clear;
FDA recognizes the importance to industry of identifying financing
opportunities to consider and tap to partially defray the costs of
implementing the PHP. FDA has commissioned additional analysis to be
performed by Research Triangle Institute to develop information to
fill this gap. FDA does not consider identification of funding
opportunities to be principally the responsibility of the Agency.
* Limited support exists for estimated time frame for increasing post-
harvest processing capacity;
The report presents what are considered minimum time frames for
meeting the needed PHP capacity and its implementation. As a baseline
minimum, it provides FDA with guidance on what the general time frame
for full implementation may be. FDA recognizes that there may be
additional time needs and constraints. The Agency stands ready to
engage the industry and states in dialog regarding time frames.
* Assumptions about post-harvest processing for oysters shipped within
state borders are likely inaccurate;
FDA and RTI recognize that the study did not consider the possibility
of Gulf States allowing for the intrastate sale of untreated oysters.
FDA has commissioned additional analysis to be performed by Research
Triangle Institute to address this concern. It may be possible that
analysis could be done to account for intrastate shipment and sale of
oysters for raw half-shell consumption that have not undergone PHP. It
may also be possible that costs could be recalculated assuming that
private processors would only post-harvest process interstate half-
shell oyster shipments. In addition, the economic impact model used to
assess the price and quantity effects of PHP requirements could be
altered to assume that only interstate shipments of oysters intended
for raw half-shell consumption would be post-harvest processed. It has
been pointed out however, that to make these alterations to the model
would require development of assumptions regarding numerous values in
the model given the lack of data (e.g., estimates of the degree to
which consumers in each of the Gulf states would substitute between
oysters that have and have not been post-harvest processed).
* Post-harvest processing costs may not be able to be passed on to
consumers.
PHP Gulf oysters are currently marketed at premium prices, according
to the report. However, if PHP becomes the standard for Gulf oysters,
the ability to gain premium prices to offset PHP processing costs
becomes less likely. There are many uncertainties around the question
of price. RTI indicated to FDA that, "If it is indeed the case that
none of the costs of PHP could be passed along to consumers, an
economic impact model is not an appropriate tool for assessing effects
of the PHP requirements because the main purpose of this type of model
is to determine the extent to which prices in the market would adjust
to a change. In this case, the results of the economic impact model
(Section 5.2 of the report) should simply be disregarded, and the
closure analysis (Section 5.1 of the report), which includes estimates
of the total costs of complying with PHP requirements, should be the
focus of the economic analysis."
[End of section]
Appendix III: Comments from the Interstate Shellfish Sanitation
Conference:
ISSC:
Interstate Shellfish Sanitation Conference:
209-2 Dawson Road:
Columbia, SC 29223:
Phone: 803-788-7559:
Fax: 803-788-7576:
Email: issc@issc.org:
August 22, 2011:
Mr. Steve Secrist, Assistant Director:
United States Government Accountability Office:
Natural Resources & Environment:
Western Region, San Francisco Office:
301 Howard Street, Suite 1200:
San Francisco, CA 94105:
Dear Mr. Secrist:
Thank you for providing the Interstate Shellfish Sanitation Conference
(ISSC) an opportunity to review and comment on your draft report
entitled, Food Safety: FDA Needs to Reassess Its Approach to Reducing
an Illness Caused by Eating Raw Oysters (GAO-11-607).
The Executive Board of the ISSC has reviewed the report and their
comments have been incorporated into the attached document. FDA has a
representative on the Executive Board; however the agency did not
participate in this ISSC review. The report focuses on ISSC efforts to
reduce Vibrio vulnificus (Vv) related illnesses and deaths. The
comments are formatted consistent with the draft report.
The ISSC is in general agreement with the recommendations of your
report. However, the scope of your investigation did not allow for a
review of the history of involvement by ISSC and FDA on this issue.
The scope did not allow for a full explanation of the many issues
associated with Vv that makes this problem very unique. Regardless, we
will continue to work with FDA to develop risk-based, cost effective
ways to improve the safety of raw molluscan shellfish. We continue to
be committed to reducing illness associated with Vv and will continue
our efforts to explore cost effective appropriate measures which can
be implemented to address illnesses associated with this naturally
occurring Vibrio.
The ISSC Executive Board and membership appreciates your efforts in
preparation and communication in the development of this report. Your
efforts were thorough and the depth of knowledge obtained by your
staff is to be commended. Should you have any questions on comments
regarding this response, please contact Ken B. Moore, ISSC Executive
Director or me at (508) 990-2860 extension 122.
Sincerely,
Signed by:
J. Michael Hickey, Chairman:
ISSC Executive Board:
cc: ISSC Executive Board Members:
Ken B. Moore, Executive Director:
ISSC Vibrio Management Committee Members:
[End of letter]
Interstate Shellfish Sanitation Conference Comments On The Government
Accountability Office Draft Report:
Food Safety: FDA Needs to Reassess Its Approach to Reducing an Illness
Caused by Eating Raw Oysters GA0-11-607:
The Interstate Shellfish Sanitation Conference (ISSC) welcomes and
appreciates the opportunity to review and comment on the Government
Accountability Office's (GAO) draft report. The ISSC is in general
agreement with the four (4) recommendations of the report. Provided
below are general comments and specific comments to the report.
Background of ISSC:
The National Shellfish Sanitation Program (NSSP) was developed in 1925
when the U. S. Public Health Service responded to a request for
assistance from local and state public health officials in controlling
typhoid fever and other bacterial diseases associated with the
consumption of raw molluscan shellfish (oysters, clams, and mussels).
The public health control procedures established by the Public Health
Service were dependent on the cooperative and voluntary efforts of
State regulatory agencies. These efforts were augmented by the
assistance and advice of the Public Health Service (now the Food and
Drug Administration [FDA]) and the voluntary participation of the
shellfish industry. These three parties combined to form a tripartite
cooperative program. The guidelines of the program have evolved into
the NSSP Guide for the Control of Molluscan Shellfish which is managed
and updated by the ISSC. The cooperative nature of the NSSP allows FDA
to administer a domestic and international program with a relatively
small federal commitment.
In the many years since its establishment, the program has proven to
be effective in minimizing the reoccurrence of illness associated with
bacterial pathogens originating from human waste. The NSSP has also
responded and essentially eliminated the occurrence of illness from
natural toxins associated with harmful algae blooms. The ISSC, NSSP,
and FDA continue to face new challenges in assuring that molluscan
shellfish are safe for human consumption. Naturally occurring
pathogens, particularly Vibrio parahaemolyticus (Vp) and Vibrio
vulnificus (Vv) is one of those challenges we must address. Our
commitment has not changed since 1925. The ISSC Vibrio Management
Committee is aggressively pursuing effective and appropriate
strategies that will address this food safety concern.
General Comments:
The ISSC applauds the effort of the GAO to examine the Vv problem.
However, the scope of your investigation did not allow for a broader
explanation of the uniqueness of the Vv issue. An understanding of the
uniqueness is critical for a full understanding of the present
controls that exist for addressing Vv illnesses.
The controls which have been incorporated into the NSSP since 1987 to
address Vv were developed by ISSC and supported by FDA. FDA was fully
engaged in the development of many of the approaches. Together, we
recognize the limited success of several of our programs. In the late
1980s we agreed that with the small number of illnesses that physician
and consumer education was more prudent than regulation. We now know
that while education has benefits it will not significantly reduce
national Vv illnesses. In 2001 we agreed that if the industry was
allowed to process oysters to reduce Vv to non detectable levels and
label the product safe that consumers would demand the safer product
and the market place would encourage the industry to Post Harvest
Process (PHP) oysters. This has not been the case. Consumer demand for
PHP product has not created the financial incentive to encourage the
majority of the industry to pursue PHP.
Your report provides an accurate estimate of the prevalence of Vv
illnesses which is approximately 32 illnesses per year. When compared
to other food borne illnesses, this number is very small. This number
has remained virtually unchanged since the early 1990s. During this
period the number of reporting states has nearly doubled. The Centers
for Disease Control and Prevention (CDC) reports that due to the
severity of the illness, practically all cases are reported. The State
Voting Delegates of ISSC, responsible for implementing controls in
their respective States, have struggled to identify controls to
address a naturally occurring organism that affects only 32
individuals annually. It is also important to note that the general
population is not at risk.
Vv poses a risk to immuno-compromised individuals. Approximately 7% of
the US population is immuno-compromised. Only a small number of that
7% is affected. Most food safety concerns place all consumers at risk.
States prioritize resources to address food safety issues which pose
the greatest threat of illness to consumers. Implementing regulatory
controls which cause industry financial hardship make regulating this
problem very problematic.
The inability to identify other food safety issues with similar
illness burdens that have been regulated with similar costs to an
industry has made consensus on this issue difficult. The cost benefit
debate on Vv has always been an obstacle for the ISSC in agreeing on
controls. Yet the ISSC has continued to be proactive in its efforts to
reduce Vv illnesses. The report recommends that the ISSC and FDA agree
on an appropriate Vv illness reduction goal. To accomplish this, ISSC
and FDA must address the two broad questions: (1) what should be the
goal of a public funded regulatory program for addressing a food
safety issue which affects 32 persons annually; and (2) to what extent
should a program of this type impose economic hardship to the industry.
Your report outlines several areas of disagreement between the FDA and
ISSC. There is agreement between FDA and ISSC in several areas that
provide a foundation for identifying approach for addressing the
problem. The FDA and ISSC agree that Vv illnesses pose a health risk
which requires public health intervention. There is agreement on the
scope of the problem and the ability of known controls to reduce risk.
The only major disagreement is the extent of public health
interventions that will appropriately address the problem. The extent
of the interventions dictates the financial impact to the shellfish
industry. The present controls adopted by the ISSC recognize a risk at
harvest and are intended to minimize any increase in risk as a result
of post harvest growth. Although these controls pose significant fiscal
challenges for the industry, states have imposed these controls. The
FDA is proposing an approach requiring PHP, which would reduce the
levels of Vv post harvest and further reduce the risk. While this
approach seems plausible it can not be implemented without financial
devastation to the industry (see Research Triangle Institute (RTI)
report).
The FDA announced in 2009, intentions to reformulate policy to require
post harvest processing or equivalent controls. This FDA announcement
exacerbated the controversy associated with Vv controls. The
cooperative nature of the NSSP requires support from all participants.
The announcement of FDA was unilateral and has alienated the industry
and states. Since the announcement FDA has been reluctant to engage in
discussions regarding Vv goals and strategies to achieve those goals.
It appears that all interested parties are fully engaged in the Vv
issue but as you indicate in your report, we must work together to
find agreement. The major challenge for FDA and ISSC is to find a
middle ground approach that can be supported by all interested
parties. It seems apparent from Congressional involvement on this
issue that the members of Congress have a desire for mutual agreement
on an acceptable risk for at-risk consumers choosing to eat raw
molluscan shellfish. For that reason the ISSC firmly supports the
recommendations of GAO.
Specific Comments to the GAO Report:
Page 2:
"The shellfish safety guidelines also included goals for reducing the
rate of illness for four reporting states."
ISSC Comments:
The goal of the ISSC Vv Management Plan was to reduce illnesses
nationally. The four (4) states of California, Florida, Louisiana, and
Texas were used to measure effectiveness. These states were chosen
because of their history in reporting Vv cases.
Page 24:
"A senior FDA official told us that this motion is unlikely to be
implemented in any meaningful way given limited state enforcement
capacity."
ISSC Comments:
The FDA is responsible for ensuring compliance. The FDA should not
have concurred with ISSC adoption of time temperature controls if
there were concerns regarding implementation and compliance. The ISSC
expects that the 2011 focused efforts of FDA to evaluate State
compliance will result in effective implementation.
[End of section]
Appendix IV: GAO Contact and Staff Acknowledgments:
GAO Contact:
Lisa Shames, (202) 512-3841 or shamesl@gao.gov:
Staff Acknowledgments:
In addition to the contact named above, Stephen D. Secrist, Assistant
Director; Leo G. Acosta, Analyst in Charge; Kevin Bray; Mark A. Braza;
Allen T. Chan; Nancy L. Crothers; Barbara J. El Osta; Lorraine R. Ettaro;
Mitchell B. Karpman; Anthony R. Padilla; Emmy L. Rhine; Anne O. Stevens;
Kiki Theodoropoulos; and Nimish D. Verma made key contributions to this
report. Also contributing to this report were Michael D. Derr,
Katherine M. Raheb, and Jena Y. Sinkfield.
[End of section]
Footnotes:
[1] The National Shellfish Sanitation Program is a federal/state
cooperative program recognized by FDA and the ISSC for the sanitary
control of shellfish (i.e., oysters, clams, mussels, and scallops)
produced and sold for human consumption. States agree to have adequate
laws and regulations to provide a legal basis within the state for
sanitary control of interstate phases of the shellfish industry and
certify that shippers of shellfish meet National Shellfish Sanitation
Program standards.
[2] Mississippi submitted its V. vulnificus risk management plan in
2009 after it was determined that two or more confirmed cases of V.
vulnificus consumption-related illnesses could be traced to the
state's waters.
[3] According to a senior ISSC official, the ISSC selected the four
states to report V. vulnificus illness data because of the quality of
their reporting systems--each had been consistently reporting V.
vulnificus illnesses for the longest time period--and because most
other states were not reporting V. vulnificus illnesses.
[4] The ISSC calculates illness rate reduction as the change in the
number of V. vulnificus illnesses per unit of population from baseline
years 1995 through 1999. In those years, in the four states, there
were 19.6 V. vulnificus illnesses and a population of 70,637,188, on
average, resulting in an illness rate of 0.28 V. vulnificus illnesses
per million people, according to ISSC data. In 2009 and 2010, there
was a population of 85,419,577, on average, in the four states.
Therefore, for example, to achieve the 60 percent illness rate
reduction from the baseline years, the number of illnesses in the four
states in 2009 and 2010 could not exceed 9.4 illnesses, on average.
[5] We selected interviewees from attendees at a November 2010
conference convening leading Vibrio researchers who had submitted
research abstracts regarding the effect of temperature on V.
vulnificus in Gulf Coast oysters and from referrals by FDA and state
officials.
[6] RTI is a leading research institute whose largest single field of
study is health research.
[7] Hemochromatosis causes the body to absorb too much iron from food.
The excess iron is stored in organs, especially the liver, heart and
pancreas, leading to life-threatening conditions such as cancer, heart
problems, and liver disease.
[8] Postharvest processing significantly reduces exposure to V.
vulnificus bacteria by reducing it to nondetectable levels, which, as
currently defined in the shellfish safety guidelines, would result in
a reduction from the current average of 32 cases per year to
approximately 1.2 cases per year, according to a 2005 report by the
World Health Organization and Food and Agriculture Organization of the
United Nations. Oyster harvest area closures essentially eliminate
exposure to V. vulnificus bacteria because raw oysters could not be
legally harvested from the closed areas. Shucking essentially
eliminates exposure to V. vulnificus bacteria because shucked oysters
are to be cooked before consumption, which destroys the V. vulnificus
bacteria.
[9] The V. vulnificus risk calculator is a tool that was designed by
FDA for the states to use to predict the level of V. vulnificus
illness rate reduction they could achieve as a result of different
combinations of time and temperature controls.
[10] RTI International, Analysis of How Post-harvest Processing
Technologies for Controlling Vibrio vulnificus Can Be Implemented
(Research Triangle Park, North Carolina: March 2011).
[11] FDA's Hazard Analysis Critical Control Point system focuses on
identifying and preventing hazards that could cause food-borne
illnesses rather than relying on spot checks of manufacturing
processes of finished seafood products to ensure safety. Processors
are required to determine whether there are food safety hazards that
are likely to occur and identify the preventive measures that can
control those hazards by preventing, eliminating, or reducing them to
an acceptable level.
[12] According to the shellfish safety guidelines, the number of V.
vulnificus illnesses in what the ISSC considers the four core
reporting states--California, Florida, Louisiana, and Texas--is used
to calculate illness rate reduction to determine if the 60 percent
goal has been achieved.
[13] GAO, Results-Oriented Government: Practices That Can Help Enhance
and Sustain Collaboration among Federal Agencies, [hyperlink,
http://www.gao.gov/products/GAO-06-15] (Washington, D.C.: Oct. 21,
2005).
[14] GAO, Managing For Results: Enhancing Agency Use of Performance
Information for Management Decision Making, [hyperlink,
http://www.gao.gov/products/GAO-05-927] (Washington, D.C.: Sept. 9,
2005).
[15] In our July 2001 report on shellfish safety, we recommended that
FDA gather baseline data--such as the results of regular shellfish
microbial tests--to facilitate evaluations of the effectiveness of
shellfish safety programs. In response, FDA said it would review, in
consultation with the ISSC, whether testing of shellfish meats should
be added to the program. FDA decided not to implement the
recommendation because it did not believe that its program could be
improved by testing shellfish meats for contaminants. See GAO, Food
Safety: Federal Oversight of Shellfish Safety Needs Improvement,
[hyperlink, http://www.gao.gov/products/GAO-01-702] (Washington, D.C.:
July 9, 2001).
[16] See Cook et al., Vibrio vulnificus and Vibrio parahaemolyticus in
U.S. Retail Shell Oysters: A National Survey from June 1998 to July
1999, J. Food Prot. 65:79-87 (2002); and DePaola et al., Bacterial and
Viral Pathogens in Live Oysters: 2007 United States Market Survey,
Appl. Environ. Microbiol. 76:2754-2768 (2010).
[17] The studies grouped the establishments into three types: (1)
restaurants and "raw bars" where shellfish are opened on site for raw
consumption; (2) seafood markets, including grocery stores, that offer
oysters in the shell for retail sale; and (3) wholesale dealers
selling oysters in the shell to the above establishments. The 2007
study noted that these data near the point of consumption provide
considerable insight on the level of protection provided by the
current U.S. system of shellfish safety controls. It cautioned,
however, that such data do not identify which aspects of the system
(e.g., harvest, transportation, and retail) account for bacterial
growth or die-off.
[18] As noted in our 2009 seafood fraud report, seafood fraud can
include a variety of illegal activities undertaken for economic gain.
Such fraud typically involves mislabeling the seafood product by, for
example, providing incorrect information about it. See GAO, Seafood
Fraud: FDA Program Changes and Better Collaboration among Key Federal
Agencies Could Improve Detection and Prevention, GAO-09-258
(Washington, D.C.: Feb. 19, 2009).
[19] This is for a 90 percent uncertainty interval.
[20] In its report, RTI identified April through October as the warmer
months.
[21] According to the RTI report, it will take a minimum of 2 years to
increase postharvest processing capacity at existing processing
facilities and 3 years to develop centralized postharvest processing
facilities for use by smaller oyster processors.
[22] [hyperlink, http://www.gao.gov/products/GAO-01-702].
[23] La. Rev. Stat. § 56:437; Tex. Parks & Wild. Code. § 76.303.
[24] Both WHO and FAO are located within the United Nations system.
WHO is the directing and coordinating authority for health and is
responsible, among other things, for providing leadership on global
health matters, shaping the health research agenda, setting norms and
standards, articulating evidence-based policy options, providing
technical support to countries, and monitoring and assessing health
trends. FAO serves both developed and developing countries by acting
as a neutral forum where all nations meet as equals to negotiate
agreements and debate policy and helps developing countries and
countries in transition modernize and improve agriculture, forestry,
and fisheries practices and ensure good nutrition for all.
[25] The risk of illness represents the probability that an individual
in the susceptible population will become ill from eating a serving of
raw oysters. The estimated number of illnesses per 100,000 servings
consumed by the susceptible population can be obtained by multiplying
100,000 by the risk of illness per serving. Following the WHO/FAO risk
assessment, the susceptible population includes people with certain
medical conditions, including chronic liver disease, cancer, kidney
disease, diabetes, and human immunodeficiency virus/acquired immune
deficiency syndrome, who are at risk of developing a potentially fatal
bloodstream infection known as septicemia.
[26] World Health Organization and Food and Agriculture Organization
of the United Nations, Risk Assessment of Vibrio Vulnificus in Raw
Oysters: Interpretive Summary and Technical Report, Microbial Risk
Assessment Series, 8 (Geneva, Switzerland and Rome, Italy: 2005).
[27] For these two parameters in particular, we obtained from the FDA
additional information that was consistent with but more detailed than
the information reported in the WHO/FAO risk assessment.
[28] Spring is defined as April, May, and June; summer is July,
August, and September; fall is October, November, and December; and
winter is January, February, and March. Although the difference in our
estimates for winter is 42 percent, this represents a difference of
about only one illness per 1 million oyster servings in absolute terms.
[End of section]
GAO's Mission:
The Government Accountability Office, the audit, evaluation and
investigative arm of Congress, exists to support Congress in meeting
its constitutional responsibilities and to help improve the performance
and accountability of the federal government for the American people.
GAO examines the use of public funds; evaluates federal programs and
policies; and provides analyses, recommendations, and other assistance
to help Congress make informed oversight, policy, and funding
decisions. GAO's commitment to good government is reflected in its core
values of accountability, integrity, and reliability.
Obtaining Copies of GAO Reports and Testimony:
The fastest and easiest way to obtain copies of GAO documents at no
cost is through GAO's Web site [hyperlink, http://www.gao.gov]. Each
weekday, GAO posts newly released reports, testimony, and
correspondence on its Web site. To have GAO e-mail you a list of newly
posted products every afternoon, go to [hyperlink, http://www.gao.gov]
and select "E-mail Updates."
Order by Phone:
The price of each GAO publication reflects GAO‘s actual cost of
production and distribution and depends on the number of pages in the
publication and whether the publication is printed in color or black and
white. Pricing and ordering information is posted on GAO‘s Web site,
[hyperlink, http://www.gao.gov/ordering.htm].
Place orders by calling (202) 512-6000, toll free (866) 801-7077, or
TDD (202) 512-2537.
Orders may be paid for using American Express, Discover Card,
MasterCard, Visa, check, or money order. Call for additional
information.
To Report Fraud, Waste, and Abuse in Federal Programs:
Contact:
Web site: [hyperlink, http://www.gao.gov/fraudnet/fraudnet.htm]:
E-mail: fraudnet@gao.gov:
Automated answering system: (800) 424-5454 or (202) 512-7470:
Congressional Relations:
Ralph Dawn, Managing Director, dawnr@gao.gov:
(202) 512-4400:
U.S. Government Accountability Office:
441 G Street NW, Room 7125:
Washington, D.C. 20548:
Public Affairs:
Chuck Young, Managing Director, youngc1@gao.gov:
(202) 512-4800:
U.S. Government Accountability Office:
441 G Street NW, Room 7149:
Washington, D.C. 20548: