U. S. Food and Drug Administration
Center for Food Safety and Applied Nutrition
FDA Prime Connection
M-I-96-1 - Milk Monitor. with Antimicro. Drug Screening Tests (1/25/96)
DHHS:PHS:FDA:CFSAN:OFP:DCP:MSB
200 "C" St., SW
Washington, DC 20204
M-I-96-1
January 25, 1996
TO: All Regional Food and Drug Directors
Attn.: Regional Milk Specialists
FROM: Chief, Milk Safety Branch, HFS-626
SUBJECT: MILK MONITORING WITH ANTIMICROBIAL DRUG
SCREENING TESTS
The FDA Center for Veterinary Medicine has issued a CVM Update
"Milk Monitoring with Antimicrobial Drug Screening Tests". A copy is
provided for your information and use in responding to questions about
drug screening test methods used for monitoring raw milk as required by
Appendix N of the Grade A Pasteurized Milk Ordinance.
Copies of this memorandum are provided for distribution to State Milk
Sanitation Regulatory Agencies, State Milk Sanitation Rating Officers,
State Laboratory Evaluation Officers and District Milk Specialists in your
region. It should be made readily available to the dairy industry and to
veterinarians in dairy practices.
Joseph M. Smucker
______________________________________________________________
M-I-96-1 January 25, 1996
CVM UPDATE
(301) 594-1735
FAX (301) 594-1831
Center for Veterinary Medicine
______________________________________________________
MILK MONITORING WITH ANTIMICROBIAL
DRUG SCREENING TESTS
Under the Public Health Service Act, the Food and Drug Administration
(FDA) and the States administer the Interstate Milk Shippers Program,
a voluntary Federal/State program established to ensure the safety and
wholesomeness of fresh milk in the United States. Under this program,
the FDA publishes the Grade A Pasteurized Milk Ordinance (PMO), a
model regulation used in voluntary, cooperative interstate milk safety
programs in which all 50 States, the District of Columbia, and Puerto
Rico participate. The PMO specifically requires that all bulk milk
pickup tankers be tested for the presence of beta-lactam drug residues.
Prior to 1991, the PMO recognized only one official test method for
detecting drug residues in milk, the Bacillus stearothermophilus Disc
Assay (BSDA). Changes to the PMO in 1991 required intensified testing
of milk for beta-lactam residues and created the need for additional
rapid, reliable screening tests that "have been evaluated through AOAC
and accepted by FDA." (AOAC International, formerly known as the
Association of Official Analytical Chemists, is a scientific organization
whose primary objective is to validate and improve analytical methods.)
As a result of this change to the PMO, 17 screening tests for beta-lactam
antibiotics, one test for chloramphenicol, and one test for sulfonamide
drugs have been evaluated and accepted by FDA. These tests are accepted
for the monitoring of truck tanker loads of raw, commingled, bovine milk
in accordance with Appendix N of the PMO and from bulk tank producer
samples in accordance with Section 6 of the PMO.
The reliability of these tests to monitor the nation's milk supply has
been questioned by some individuals. This CVM UPDATE addresses
these issues and clarifies the role of screening tests for monitoring raw
milk.
The accepted screening tests have met a standard for a low incidence of
false positive and false negative findings. Combined with these standards,
there are important principals of use which must be considered. These
are:
1. A positive result from a screening test is a presumptive
indication that an analyte is present in the milk sample.
2 The screening test does not necessarily identify the specific
analyte causing the test to be positive nor does it measure the
quantity.
3. All the accepted tests may produce a positive result when the
drug concentration is below the tolerance/safe level. This is a
false violative result, not a false positive result.
4. A chemical analysis is required to determine whether or not a
given milk sample contains antimicrobial drug residues above the
tolerance/safe level (violative).
5. Despite their limitations, the accepted tests represent the
"state of the art" in rapid detection of drug residues in milk.
The percentage of truck tankers found positive in 1994 (National Milk
Drug Residue Third Party Data Base) was 0.063 percent. This low
incidence of positive truck tankers supports our standard for selectivity
(false positive test result). Further, this low incidence also demonstrates
that the majority of the milk producers are using drugs in a responsible
manner. The FDA has found no evidence which would indicate that the
use of approved beta-lactam drugs in accordance with label directions will
cause a violative or non-violative residue in a truck tanker. The FDA has
concluded that despite the inherent limitations of screening tests, the issue
remains one of proper drug use. The FDA believes the use of the
accepted tests under the provisions of Appendix N, PMO, has reduced the
amount of positive milk entering the food supply.
The FDA recognizes the economic losses to the milk producer which
would result from false violative and false positive test results. With this
issue in mind, the NCIMS and the FDA agreed to retest all original
positive truck tanker samples using the same test when the initial test is
conducted by an industry analyst. Only after the results from retesting
indicate a positive finding is a truck tanker rejected. Retesting increases
the probability of acceptance of a non-violative milk tanker and decreases
the number of non-violative truck tanker rejections. The FDA must also
be concerned with the incidence of false negative results to ensure the
public health.
The low incidence of positive truck tanker results do not appear to be
caused by unreliable tests. Based on reports from the States, the FDA
has concluded that misuse of animal drugs is the cause of most positive
test results from truck tanker testing even when residue concentrations are
below the tolerance/safe level. The FDA has found no evidence which
indicates that treating lactating cows in accordance with labeled directions
will cause a positive truck tanker. The follow-up by the State regulatory
agencies on positive truck tankers indicates that the positive test results are
primarily the result of misuse of animal drugs.
It has been suggested by some individuals that testing under the
provisions of Appendix N, PMO, with the FDA and NCIMS accepted
screening tests be discontinued. The FDA takes the position that
discontinuing tanker truck testing is not consistent with a commitment to a
safe milk supply, and therefore would be unacceptable to the Agency.
Prior to the implementation of Appendix N, PMO, the Government
accounting Office concluded that there was no comprehensive strategy to
ensure the safety of the milk supply. The State regulatory agencies and
FDA are committed to maintaining a safe milk supply and have
developed a comprehensive strategy for ensuring a safe milk supply. The
strategy adopted by FDA and NCIMS includes monitoring of truck tankers
in accordance with Appendix N, PMO, monitoring producer bulk tanks in
accordance with Section 6, PMO, participation in the ten point Milk and
Dairy Beef Quality Assurance Program in the event of a violation,
monitoring the use and labeling of drugs through the PMO Farm
Inspection Program, and individual cow testing.
Although research indicates that some screening tests may produce false
positive test results in milk from individual cows, the FDA is not aware of
any data which supports the conclusion that unique factors in the milk
from individual cows produce false positive findings in truck tanker milk
samples. The FDA maintains the view that the misuse of animal drugs
causes a majority of screening test positives at the truck tanker. Despite
the inherent limitations of screening tests, the issue remains one of proper
drug use.
No screening test has been evaluated by the Center for Veterinary
Medicine or the AOAC International Research Institute for use on milk
from individual cows. Nine of the currently accepted tests for testing
truck tanker milk are being evaluated for this use.
FDA has prepared a document entitled "Evaluation and Use of Milk and
Antimicrobial Drug Screening Tests" which provides a detailed discussion
of the evaluation and use of the screening tests as well as related issues
regarding the monitoring for animal drug residues in milk. Copies of this
document are available by calling CVM's Communications and Education
Branch on (301) 594-1755 or by writing to Communications and
Education Branch, HFV-12, CVM/FDA, 7500 Standish Place, Rockville,
MD 20855. Comments or questions on this issue may be addressed to
Dr. Norris E. Alderson, HFV-500, CVM/FDA, 7500 Standish Place,
Rockville, MD 20855. Phone 301-594-1702; email
alderson@a1.cvm.fda.gov.
FOOD AND DRUG ADMINISTRATION
EVALUATION AND USE OF MILK
ANTIMICROBIAL DRUG SCREENING TESTS
I. BACKGROUND
The use, in accordance with label directions, of the approximately 40
animal drugs approved for use in lactating dairy cows presents
consumers with a negligible risk from residues in milk 1. The eleven
approved drugs used to treat mastitis and respiratory infections --
penicillin, ampicillin, cephapirin, hetacillin 2, amoxicillin,
ceftiofur, oxytetracycline, chlortetracycline, novobiocin,
erythromycin, and sulfadimethoxine -- are among the most widely used
drugs to treat dairy cows. The first five can cause hypersensitivity
reactions in some people. Because of this, residues of beta-lactam
type antibiotics are of greatest concern from the standpoint of human
food safety.
Prior to 1991, the only required monitoring of raw milk for animal
drug residues was described in Section 6 of the Grade A Pasteurized
Milk Ordinance (PMO). It states, "...drug tests on each producer's
milk shall be conducted by State regulatory agencies at least four
times during any consecutive 6 months." The tests accepted at that
time by the National Conference on Interstate Milk Shipments (NCIMS)
for Section 6, PMO, testing were the Difco Bacillus stearothermophilus
Disc Assay (BSDA), the generic BSDA described in Appendix G of the
PMO, the Charm I and II Liquid Tests, and the Delvo Test P.
II. DEFINITIONS
The following definitions are important to the understanding of the
evaluation and use of milk antimicrobial drug screening tests:
________________________________________
1 Additional animal drugs may be given to lactating cattle by veterinarians under the
Center for Veterinary Medicine's Extra Label Drug Use Policy.
2 Residues of hetacillin are almost immediately and completely metabolized to
ampicillin following administration to animals, and for this reason, are not normally
considered in drug residue monitoring programs.
A. A truly positive test result is a positive test result on a milk
sample in which the actual drug concentration is at or above the
established tolerance or safe level.
B. A false violative test result is a positive test result on a
milk sample in which the actual drug concentration is at or above
the detectable concentration, but below the established tolerance
or safe level.
C. A false positive test result is a positive test result on a milk
sample which actually contains no drug residue at a detectable
concentration.
D. A false negative test result is a negative test result on a milk
sample which contains drug residue at a detectable concentration.
E. The sensitivity rate is the percentage of truly positive samples
that are found by the test to be positive.
F. The selectivity rate is the percentage of truly negative samples
that are found by the test to be negative.
G. The estimated minimum ninety percent sensitivity level is the
estimated lowest drug concentration which gives, with 95 percent
confidence, a positive result with the test at least 90 percent
of the time on samples which are truly positive.
III. APPENDIX N, PASTEURIZED MILK ORDINANCE
At their 1991 biennial meeting, the National Conference on Interstate
Milk Shipments (NCIMS) passed a resolution modifying the PMO to
require the contents of every bulk milk pick-up truck tanker to be
tested for beta-lactam residues prior to entering the food supply.
This change in the PMO, which was identified as Appendix N, created a
need for rapid, reliable screening tests that could detect residues of
beta-lactam antibiotics in milk when present at or above the
concentrations established by the U.S. Food and Drug Administration
(FDA) to be safe.
In part because Appendix N, PMO, required methods used in monitoring
programs to "have been evaluated through AOAC and accepted by FDA,"
the Agency began the evaluation program in cooperation with the AOAC
International-Research Institute (AOACI-RI) and the A. D. Little, Inc.
(A. D. Little, Inc. is no longer participating in the evaluation
program). These firms served as independent third parties in the
initial evaluation.
A protocol was developed for the evaluation of the tests for detecting
antibiotics in raw, bulk, commingled milk. This protocol was
developed with cooperation and input of representatives of all
interests of the milk industry. Since the NCIMS (Appendix N, PMO)
required monitoring at the truck tanker, the protocol was only for
raw, bulk, commingled milk and not for milk from individual cows.
The selectivity and sensitivity data developed by the test sponsors
were verified in an independent laboratory. The selectivity standard
used requires tests to correctly identify, with 95 percent confidence,
milk containing no drug residues in 90 percent of the samples (90/95).
The initial sensitivity standard required the tests to detect milk
containing residues of claimed drugs at their tolerance/safe levels 90
percent of the time with 95 percent confidence. A test could meet
these standards by correctly identifying 30 of 30 zero control samples
and 30 of 30 samples containing each claimed drug at its
tolerance/safe level. With this selectivity standard for acceptance,
the probability is low that an accepted screening test would produce a
positive result on a milk sample that does not contain any of the drug
it is designed to detect.
Although the only standards FDA established for acceptance of tests
were those for selectivity rate and sensitivity rate, other parameters
of test performance were also evaluated in the joint program, and the
sensitivity was validated with incurred residues in milk. Parameters
studied include: effect of the presence in milk of PMO-permitted
levels of bacteria and somatic cells; possible interference of other
classes of animal drugs which might be present as residues in milk;
the percent of positive results to be expected from a test when milk
containing residue levels below a drug's tolerance are assayed (dose
response)(false violative test results); and ruggedness testing
provided for the evaluation of variation in times, volumes,
temperatures, etc. in the directions for operating the test that would
impact the performance of each test. In requiring the incurred
residue parameter, milk from animals receiving the different drugs are
evaluated using the tests. This requires that the tests detect the
incurred residues at the tolerance/safe level with the concentration
in the milk being quantitatively determined by a chemical test. The
objective of including these parameters of test performance is to
insure that manufacturers of FDA accepted tests provide potential
users with as much information as possible about the performance of a
given test. This enables individual users to make informed decisions
about which test is the best one for them to use and also to encourage
test manufacturers to develop improved tests.
To date, 17 screening tests for beta-lactam residues, one test for
chloramphenicol, and one test for sulfonamide residues in milk have
been accepted by FDA, certified as "performance tested" by AOACI-RI,
and/or accepted by A.D. Little, Inc. These tests are also accepted by
the NCIMS for Appendix N, PMO, testing. The attached Chart contains
the tolerance/safe levels for approved beta-lactam drugs and the 90/95
percent sensitivity levels for accepted tests.3
IV. BACILLUS STEAROTHERMOPHILUS DISC ASSAY
(BSDA)
Appendix G of the PMO presently recognizes a generic BSDA. The
BSDA is a microbiological test based on growth inhibition to detect
antibiotic presence. Performance of the BSDA was evaluated in 1982
in an AOAC collaborative study using only penicillin. A 16 mm zone
was accepted as the standard for a violative result. A positive test
result with this test indicates that an inhibitor is present in the
milk. It does not provide the identity nor quantity of the inhibitor.
In addition to penicillin, this BSDA procedure has been demonstrated
to detect residues of ampicillin, cephapirin, and amoxicillin at or
below safe levels. The test also detects residues of ceftiofur at
slightly above its level of concern for misuse.
Charm Sciences and Difco Laboratories market a complete BSDA kit.
The Charm Sciences' test has been evaluated and accepted for Appendix
N monitoring. Difco recently decided to have its test evaluated. At
this time, the Difco BSDA is not accepted for Appendix N, PMO,
testing.
At the 1995 meeting of the NCIMS, all the tests accepted for milk
monitoring under the provisions of Appendix N, PMO, were also accepted
___________________________________________________
3 The BSDA described in Appendix G and discussed above as having been evaluated
earlier by AOAC, was widely used in the years prior to 1991. The test was
available as a complete kit from Difco Laboratories, but some testing facilities used
a generic BSDA made in-house from agar, spores, standards, etc. obtained from
Difco or from other sources. The BSDA listed in the attached Chart is a complete
kit manufactured by CharmSciences, Inc. Neither the Difco BSDA nor any generic
BSDA was submitted for inclusion in the Agency's initial evaluation of the BSDA
test. Under the provisions of Appendix N, any version of the BSDA that has not
been accepted by FDA should no longer be used to monitor tanker truck milk.
for Section 6, PMO, testing (producer bulk tank). The Difco BSDA, the
Delvo Test P 5 Pack, the generic BSDA described in Appendix G, PMO,
and the Charm I and II Liquid tests are accepted for use on farm bulk tank
samples but not Appendix N, PMO, tanker truck samples. These tests
have not passed the performance standards of the Appendix N, PMO,
tests, and their use in regulatory decisions is an issue that is expected to
be an item for consideration by the NCIMS. (Note: The Charm I and II
Liquid tests are no longer manufactured).
V. SENSITIVITY STANDARD
Since the screening tests give positive results at drug concentrations below
the tolerance/safe level, there have been a number of questions regarding
the acceptance of the tests. These comments have inferred that tests
should not have been accepted by the NCIMS as they were to have been
evaluated exactly "at the safe level or tolerance" (Appendix N, PMO).
The original evaluation required a determination of a 90/95 percent
concentration at or below the tolerance/safe level. All the evaluations
have included the tolerance/safe level. Had FDA not accepted tests which
give a positive result at drug concentrations below the tolerance/safe level,
none of the tests would have been accepted (including the current Section
6, PMO, tests). The current state of screening test technology is
inadequate for this level of precision.
All the tests have a characteristic sigmoid response curve. This means
that as drug concentration increases, there is also a corresponding increase
in percent positives until you reach a concentration plateau after which all
samples will be positive. In the original evaluations, only a minimal
number of assays were required at the lower concentrations. The Food
and Drug Administration originally established a sensitivity rate standard
requiring a test to give 90 percent positive results with 95 percent
confidence on milk samples containing drug concentrations at or below
the tolerance/safe level. All the tests met this standard for each drug
claimed on the label. When the data were presented to the NCIMS, they
requested that the test sponsors provide data which define the dose
response at the lower drug concentrations with equal confidence as the
90/95 percent level. These data provide the means to determine the
estimated minimal 90 percent level, with 95 percent confidence. The
development of these data is nearing completion and will be reflected on
revised labeling for each test. The FDA remains convinced that the dose
response information is important to the user in deciding on the test to
use. This information is on the label of each accepted test.
Ideally, the accepted tests should give no positive results when there are
drug residues below the tolerance/safe level (false violative test result).
Equally desired is for the tests to give a positive result 100 percent of the
time when the drug concentration is exactly at or above the tolerance/safe
level (no false negative test results). The screening tests evaluated to date
do not perform in this manner. They do not "turn on and off" precisely at
the tolerance/safe level.
VI. QUESTIONS/ISSUES ON PERFORMANCE OF THE
CURRENTLY ACCEPTED TESTS
There is no ideal screening test currently available for detecting
antimicrobial drug residues in milk (see attached Chart). None of the
beta-lactam tests detect all the beta-lactam drugs. Some have the 90/95
percent sensitivity level for specific drugs below the tolerance/safe level,
and all give false violative test results. These multi-residue tests are
developed to be sensitive to at least four of the beta-lactam drugs at their
tolerance/safe level. Each multi-residue test, including the BSDA,
detects one or more drugs at levels below their tolerance, and each test
fails to detect one or more approved beta-lactam drugs at their tolerances.
There is a consumer concern when a test is used which does not detect all
the beta-lactam drugs because milk might contain residue that has not been
shown to be safe. The above is also true for the sulfonamide test.
The rejection of safe milk is an economic issue for the milk producer.
Currently the NCIMS accepted screening tests, including the BSDA,
produce positive results on a certain proportion of samples which contain
below tolerance/safe level concentrations of residues (false violative test
results). For example, a test that demonstrates a 90/95 percent sensitivity
of 10 ppb for residues of ampicillin, might produce a positive result on 40
percent of the samples with 4 ppb, etc. Of course, the number of truck
tankers that would be rejected because of this detection characteristic of
screening tests is a function of the number of truck tankers that arrive at a
processor containing 4 ppb of ampicillin. Our best estimates of this
number is that it is low. To help test users select the most appropriate test
for their situation and minimize the number of false violative test results,
the Agency requires concentration-response data to be placed on the labels
of all accepted screening tests.
Furthermore, it is obvious from Chart that different tests demonstrate
different sensitivities toward the six approved beta-lactams. It is not
surprising therefore, that the action of replacing the single official BSDA
(Appendix G, PMO) with 17 different, more reliable, and some more
sensitive, accepted beta-lactam tests has resulted in some confusion. For
example, the same sample of milk could produce different results when
tested by different receiving stations, processors, or State regulatory
agencies, that do not employ the same test.
Recognizing the economic importance of false violative test results to the
milk producer, the Agency and the NCIMS agreed to a procedure whereby
all truck tankers found to have a positive test result in their initial
screening test must be confirmed using a procedure to reduce the
incidence of false positive and false violative test results. This is actually
a retesting of the positive sample using the same test when the initial test
is conducted by the industry analyst. This retesting includes the
following:
1. Positive control sample -- Ensures that truly positive milk
will test positive.
2. Negative control sample -- Ensures that truly negative milk
will test negative.
3. Two truck tanker samples -- For the truck tanker to be
rejected, one of the retests must be positive.
In the example of the test described above, the follow-up testing procedure
would reduce the chance of finding a positive truck tanker with milk
containing 4 ppb of ampicillin from 40 to 26 percent. Some comments
have criticized this procedure because of use of the term "confirm." The
purpose of the retesting is to reduce the number of false violatives.
The following are other issues for consideration in forming an opinion
regarding the use of milk residue screening tests:
1. Milk producers rarely, if ever, treat their entire herd with
an antibiotic. Rather, in most instances, a few cows are
treated at the same time. Following the milk discard time,
the milk from a few treated cows is commingled in the farm
bulk tank with a greater amount of milk from untreated
animals. Farm bulk tank milk is normally further diluted
with milk from several farms while being transported to a
milk processing plant where it is tested. The effect of this
dilution virtually eliminates the possibility that truly positive
truck tanker milk will occur when drugs are used in an
approved manner. For a test to give a positive result, even
if it is a false violative test result, the drug generally must
have been misused in the cow(s) that contributed the milk
containing residues at a high enough concentration to
contaminate the truck tanker. A positive test result on a
truck tanker sample is an indication that the producer whose
milk caused the positive truck tanker test offered milk for
sale which was above the recommended tolerance/safe level
(regulatory action level). Follow-up testing on producer
samples and on-farm investigations confirm this conclusion.
The proper use of approved animal drugs in individual
dairy cows does not cause the accepted screening tests to
produce a truly positive truck tanker result as some may
suggest.
In accordance with the provisions of Appendix N, PMO, all
positive test results must be investigated. As a false
violative test result is a possible outcome from the
monitoring of milk with screening tests and based on a case
by case review of each investigation, additional testing
and/or investigation may be justified.
2. An objective of the National Drug Residue Milk Monitoring
Program (NDRMMP) is to estimate the incidence of
non-beta-lactam drug residues. This objective is being
accomplished through use of screening tests for other animal
drugs which have neither been evaluated nor accepted for
use in Appendix N type monitoring programs.
Some have misinterpreted data from FDA's NDRMMP to
conclude that the screening tests used in this program
produced an 85.6 percent false positive rate. This is a
misrepresentation of the data. These quarterly reports
include any screening test samples submitted by the states to
FDA for confirmation as well as the NDRMMP samples.
Those samples previously found to be positive by screening
tests are reported as violative as determined by chemical
tests only when the drug concentration is found to be at or
above the tolerance/safe level. The majority of the
screening test positive samples (presumptive positive result)
have been found to contain detectable concentrations of
drug; these are false violative test results. This misleading
report of a 85.6 percent false positive results includes those
samples found to contain drug concentrations below the
tolerance/safe level but not reported as violative. Except
for the beta/lactam samples, all confirmed positive test
results are based on a chemical test.
3. The FDA funds a contract under which State milk
regulatory agencies report the results of milk monitoring in
their states. The proportion of reported positive loads
throughout the country has been very low. In 1994, there
was 3.2 million truck tanker loads tested with 0.063 percent
found positive (63 of each 100,000 truck tankers). The
milk rejected for human food totaled approximately 68
million pounds which is approximately 0.044 percent of the
milk supply. These numbers represent the maximum
percent of positive test results at or above the tolerance/safe
level plus false violative and false positive test results.
4. Several reports in the scientific literature describe the
presence of "natural defense secretions" in milk from
mastitic cows that will produce positive screening test
results on milk from cows that have not been treated with
any animal drug. These data indicate that certain screening
tests produce false positive test results in milk from
individual cows secreting these natural inhibitors. Some
authors have attempted to apply the false positive rates
obtained in milk from individual cows to samples from
truck tankers containing commingled milk derived from
many cows. The FDA believes these are invalid
comparisons as the data reported would not support a
conclusion that false positive milk from one or several cows
produces a false positive test result in truck tanker milk.
Using the level of somatic cells as an indicator of the
presence of natural inhibitors, none of the research and
test evaluations conducted at FDA has revealed an effect on
test performance from natural inhibitors unless the milk has
a somatic cell count that exceeds several million - many
times the level of these cells permitted under the PMO.
VII. TRULY POSITIVE, FALSE POSITIVE, OR FALSE
VIOLATIVE TEST RESULT
Prior to the implementation of milk testing in accordance with the
provision of Appendix N, PMO, the milk processors, who carry out the
testing, acknowledged that some positives could result from residue
concentrations that are below the tolerance/safe level. However, since
they advised FDA that they do not want any antibiotic residues in their
products, they indicated they were prepared to accept some false violative
results.
Based on the levels of drug residues expected to be in milk from
individual cows following recommended milk discard periods for approved
drugs, the use of animal drugs in accordance with label directions does not
cause a truly violative result in truck tanker samples when using an
accepted screening test. There is the small possibility that when using an
approved beta-lactam drug in accordance with label directions on an entire
herd with that herd being the only one represented on a given truck
tanker, a positive result is possible. The FDA believes the probability of
this scenario happening is remote as herd treatment for mastitis is a rare
event. There is also the possibility of a positive screening test result for
chlortetracycline when administered via the feed on a herd basis.
It is important to understand that a truly false positive, false negative,
or false violative test result can only be determined by a chemical
analysis of the milk sample. A determination of the true status of a
positive sample, before the milk enters the food chain, requires that:
1. Each positive tanker be held for further chemical analysis to
determine the drug and concentration. This would prove to
be impractical because of the time and cost involved.
2. Screening tests of increased capability be accepted for
screening truck tankers. Current state of the art for
practical, inexpensive tests, does not permit an identification
of the drug nor determination of concentration. In the
meantime, the NCIMS screening tests are the best
technology available.
VIII. MILK DISCARD TIME ON APPROVED DRUG
LABELS
Milk discard time is the number of hours after treatment that are required
for residues in milk to reach the tolerance/safe level. The milk discard
time is not the point at which residues can no longer be detected with the
chemical tests or some screening tests. Currently, milk discard times are
established using a chemical test; the accepted milk screening tests do not
have the required analytical characteristics to establish official milk
discard times.
IX. EVALUATION OF SCREENING TESTS FOR USE IN
MILK FROM INDIVIDUAL COWS
The FDA announced in 1993 its interest in developing a protocol for the
evaluation of screening tests for milk from individual cows. A number of
meetings were held involving representatives of the industry, including the
National Mastitis Council. A protocol was designed using the same
administrative procedures as for the accepted tests for commingled milk
with the test manufacturers paying for the evaluation. As there is no
regulatory requirement for the testing of milk from individual cows, the
manufacturers saw no economic incentive to have their tests evaluated;
and consequently, no tests have been evaluated under this protocol.
Given FDA's view of the importance of having reliable, independently
evaluated tests available for use by the producer/practitioner in milk from
individual cows, the Agency decided to fund an evaluation of the tests.
Recognizing the screening tests which had already been evaluated for
commingled milk and the parameters included in that evaluation, FDA
developed a different protocol using the selectivity and mastitis concerns
as parameters for additional evaluation. This revised protocol was also
reviewed by representative of the National Mastitis Council. The protocol
addresses the issues of false positive test results in milk from individual
cows which are healthy and have not been treated with drugs and also a
mastitis model to address the issue of cows recovering from mastitis. In
the mastitis model study, the tests must give the correct result on a milk
sample from healthy cows and also on visually normal milk following
recovery from an endotoxin challenge in the udder. The tests must also
give the correct response on the visually normal milk when claimed drugs
are added. The Agricultural Research Service at the United States
Department of Agriculture, Beltsville, MD is conducting the studies
under an Interagency Agreement with the Center for Veterinary Medicine.
Test manufacturers requested that nine of the accepted beta-lactam tests be
evaluated. The selectivity study section of the protocol has been
completed. The mastitis model study was completed in 1995. The FDA
expects to issue a report on these studies in the spring of 1996.
As screening tests can detect drug concentrations below the tolerance/safe
level, a positive test result is possible with a screening test on a milk
sample collected from cows after the labeled milk discard time. For this
reason, FDA does not recommend the use of screening tests on cows
which have been individually treated in accordance with label directions.
It should be noted, however, that these lower concentrations will not cause
a violative or non-violative positive truck tanker except in those exceptions
described previously.
If it is desirable to have no detectable residue in milk from individual
cows, then testing with a screening test until the sample is negative is a
viable option. To preclude the possibility of a truck tanker positive result,
screening tests may also be used in milk from individual cows treated in
an extra-label manner.
X. STATE OF THE ART
The FDA recognizes that screening tests for detecting antimicrobial drugs
in milk are neither drug specific nor quantitative in their performance. A
positive test result means that something in the milk caused this outcome.
The evaluation protocol used provides ample assurance that a positive test
will result when the drug claimed on the label is in the sample at
concentrations at or above the tolerance/safe level. Similarly, FDA is
confident that the test result will be negative when there is no drug in the
sample. The residue concentration between no detectable concentration
and the 90/95 percent concentration (false violative test result) is a
concern, particularly for the producer, as these concentrations may cause a
positive test result and the truck tanker of milk deemed unacceptable for
human food. From a human safety perspective, FDA is also concerned
for the false negative potential of the tests.
It is important to understand that under Section 6 of the PMO, producer
bulk tank milk has been monitored for a number of years, before
implementation of Appendix N, PMO, using some of the accepted
Appendix N, PMO, tests. These screening tests have the same limitations
whether used as a Section 6, PMO, test or an Appendix N, PMO, test.
The difference is that now the PMO requires the monitoring of each
tanker truck for beta-lactam residues before entry into the food chain.
XI. SUMMARY
Some have inaccurately reported results of milk monitoring to advance
the proposition that the use of the accepted milk residue screening tests be
discontinued until more accurate tests are available. The percentage of
truck tankers found positive in 1994 (National Milk Drug Residue Third
Party Data Base) was 0.063 percent. This low incidence of positive truck
tankers supports our standard for selectivity (false positive test result).
Further, this low incidence of positive truck tankers also demonstrates that
the majority of the milk producers are using drugs in a responsible
manner. The FDA has found no evidence which indicates that the use of
approved drugs in accordance with label directions will cause a positive
test result in a truck tanker with the possible exception when
chlortetracycline is administered in feed on a herd basis. The FDA has
concluded that despite the inherent limitations of screening tests, the issue
remains one of proper drug use. The FDA believes the use of the
accepted tests under the provisions of Appendix N, PMO, has reduced the
amount of milk containing antimicrobial drugs entering the food supply.
The low incidence of positive truck tankers and the rejection of those
truck tanker loads for human food is recognition of the milk industry's
focus on and commitment to the human safety of milk. This low
incidence and Appendix N, PMO, monitoring is a significant factor
for the milk industry in maintaining consumer confidence in the safety
of milk. The reliability of the accepted tests to detect drug residues adds
credibility to the testing program, and the small percentage of positive
tankers confirms the validity of the acceptance standards for the tests.
The FDA recognizes the limitations of the screening tests. These are:
1. The tests cannot identify the drug residue nor the
concentration of the residue. A safety determination cannot
be determined from a positive result.
2. The screening tests produce some positive results when the
drug concentration is below the tolerance/safe level. Under
the provisions of Appendix N, PMO, this limitation
provides for a few positive truck tanker loads of milk to be
rejected for human consumption when the milk is safe.
While these accepted tests have their limitations in drug identification and
quantification, they work well in detecting violative drug residues in milk.
The low probability of a false positive and a false negative finding is
inherent in the acceptability standards of the evaluation. The FDA has
found no evidence which would indicate that the accepted tests are not
performing in this manner.
The FDA recognizes the economic losses to the milk producer which
would result from false violative and false positive test results. With this
issue in mind, the NCIMS and the FDA agreed to retest all original
positive truck tanker samples using the same test when tested by an
industry analyst. Only after the results from retesting indicate a positive
finding is a truck tanker rejected. Retesting increases the probability of
acceptance of a non-violative milk tanker and decreases the number of
non-violative truck tanker rejections. The FDA must also be concerned
with the incidence of false negative results to ensure the public health.
Based on reports from the states, the FDA has concluded that misuse of
animal drugs is the cause of most positive test results from truck tanker
testing even when residue concentrations are below the tolerance/safe
level. The FDA has found no evidence which indicates that treating
lactating cows in accordance with labeled directions will cause a positive
truck tanker. The low incidence of positive truck tanker results do not
appear to be caused by unreliable tests. The follow-up by the State
regulatory agencies on positive truck tankers indicates that the positive test
results are primarily the result of misuse of animal drugs.
Office of Science, HFV-500
CVM/FDA
7500 Standish Place
Rockville, MD 20855
January, 1996
MILK DRUG RESIDUE TEST DETECTION LEVELSa
A. SCREENING TESTS FOR BETA-LACTAM DRUGS
DRUG PEN AMP AMOX CLOX CEPH CEFT
____________________________________________________________________________________
Tolerance/Safe
Level PPB 5 10 10 10 20 50
TEST
Charm I/Cowside
II Test for NDb NDb NDb 10 NDb NDb
Cloxacillin
Charm II Test for
Cloxacillin NDb NDb NDb 10 NDb NDb
Charm II Tablet
Competitive Assay 4.8 9 10 70c 4.5 25
Charm Farm Test 5 10 10 40 20 25
Charm II Tablet
Sequential Assay 4.8 8 10 50c 4.5 23
Charm II Tablet
Transit Test 4.8 9 10 80c 4.5 13
Charm Rapid
Inhibition Test 3 4.5 4.5 25c 16 50
Charm I/Charm II
Tablet Test 4.8 10 10 50c 8 40
Charm II Tablet
Quantitative Assay 4.8 8 10 10 4.5 23
Charm B.
Stearothermophilus 5 6.5 10 48c 11 75c
Disk Assay
Delvotest P 3 10 8 30c 8 50
Delvo-X-PRESS 5 10 10 50 10 10
Lactek B-L 5 10 8 8 16 NDb
Lactek CEF NDb NDb NDb NDb NDb 50
Penzyme III Test 5 10 8 80c 8 80
Penzyme Milk Test 5 10 8 80 8 80
SNAP Test 5 10 10 50c 8 50
B. SCREENING TESTS FOR OTHER DRUGS
DRUGS CAP SMZ SDZ SDM STZ
________________________________________________________________________________
Tolerance/Safe
Level PPB 1 10 10 10 10
TEST
Charm II
Chlororamphenicol 1 NDb NDb NDb NDb
Test
Charm II Sulfa
Drug Test NDb 9.6 5.8 3.6 7.5
a - PPB which can be detected by test 90 percent of time with 95 percent confidence.
b - ND indicates not detected at 100 ppb.
c - Precise 90/95% levels were not determined for these sensitivities that are significantly above the
tolerance/safe level.
ABBREVIATIONS
PEN = Penicillin
AMP = Ampicillin
AMOX = Amoxicillin
CLOX = Cloxacillin
CEPH = Cephapirin
CEFT = Ceftiofur
CAP = Chloramphenicol
SMZ = Sulfamethazine
SDM = Sulfadimethoxine
STZ = Sulfathiazole
Go back to listing of M-I memos
Return to FDA Prime Connection Milk Safety references
Go BACK to CFSAN/FDA food and consumer information pages
Hypertext updated by ear, 11/19/96