[Federal Register Volume 59, Number 31 (Tuesday, February 15, 1994)]
[Unknown Section]
[Page 0]
From the Federal Register Online via the Government Publishing Office [www.gpo.gov]
[FR Doc No: 94-2030]


[[Page Unknown]]

[Federal Register: February 15, 1994]


_______________________________________________________________________

Part III





Department of Transportation





_______________________________________________________________________



Office of the Secretary



National Highway Traffic Safety Administration



_______________________________________________________________________



49 CFR Part 40




Procedures for Transportation Workplace Drug and Alcohol Testing 
Programs and Proposed Model Specifications for Screening Devices To 
Measure Alcohol in Bodily Fluids; Final Rule, Proposed Rule, Notice
DEPARTMENT OF TRANSPORTATION

Office of the Secretary

49 CFR Part 40

[Docket 48513]
RIN 2105-AB95

 
Procedures for Transportation Workplace Drug and Alcohol Testing 
Programs

AGENCY: Office of the Secretary, DOT.

ACTION: Final rule.

-----------------------------------------------------------------------

SUMMARY: Under the Omnibus Transportation Employee Testing Act of 1991, 
the Department of Transportation is required to implement alcohol 
testing programs in various transportation industries. This rule 
establishes uniform testing procedures that would be used by all 
Department of Transportation operating administrations conducting 
alcohol testing programs under the Act or conducting alcohol testing 
programs modeled on those required by the Act. This rule also 
implements changes required by the statute in the Department's drug 
testing procedures.

DATES: Effective Dates: This rule is effective March 17, 1994, except 
Sec. 40.25(f)(10)(i)(B), which is effective August 15, 1994. Compliance 
Date: Compliance with Sec. 40.25(f)(10)(i)(B) is authorized beginning 
March 17, 1994.

FOR FURTHER INFORMATION CONTACT: Donna Smith, Acting Director, 
Department of Transportation Office of Drug Enforcement and Program 
Compliance, 400 7th Street, SW., Washington DC, 20590, room 9404, 202-
366-3784; or Robert C. Ashby, Deputy Assistant General Counsel for 
Regulation and Enforcement, 400 7th Street, SW., room 10424. 202-366-
9306.

SUPPLEMENTARY INFORMATION:

Background

    The Omnibus Transportation Employee Testing Act of 1991, enacted 
October 28, 1991, directed significant changes in the Department of 
Transportation's substance abuse-related programs for most 
transportation industries that the Department regulates. These changes 
are discussed in detail in the Common Preamble published in today's 
Federal Register. With respect to drug testing procedures, the Act 
added a requirement for using the ``split sample'' approach to testing, 
which Congress believed would provide an additional safeguard for 
employees. The Act also imposes a variety of requirements for alcohol 
testing procedures, which this regulation also implements. The Coast 
Guard is not amending its existing alcohol testing regulations (33 CFR 
part 95 and 46 CFR part 4), and will continue to use separate 
procedures for that testing.
    The Department's drug testing procedures, 49 CFR part 40, have 
governed drug testing under all six operating administration drug 
testing rules since 1988. Likewise, this rule governs alcohol testing 
procedures for the five modes affected (the Coast Guard is not covered 
by the alcohol testing procedures of this part). Under the rule, the 
existing drug testing procedures become a separate subpart of the 
regulation, and we are adding new subpart containing the alcohol 
testing procedures.
    Having all the Department's uniform drug and alcohol testing 
procedures in a single regulation will simplify compliance for covered 
parties and avoid confusion by permitting all parties to look to one 
source for information on these issues. This should be particularly 
helpful to those employers who have employees covered by more than one 
DOT operating administration. However, employers regulated solely by 
the Coast Guard should continue to refer to 33 CFR part 95 and 46 CFR 
part 4 for alcohol testing requirements and procedures.
    The Department published the Notice of Proposed Rulemaking (NPRM) 
for this rule on December 15, 1992, at the same time as the operating 
administrations (OAs) published their proposed alcohol and, in some 
cases, drug testing rules. We received over 250 comments to the part 40 
docket. In addition, the OAs' dockets received some comments on the 
testing procedure issues raised by the part 40 NPRM. The Department 
considered all these comments.

Comments and Responses

Split Sample Procedures for Drug Testing

    This discussion concerns how we will carry out a statutory 
requirement to use the ``split sample'' method for collecting and 
analyzing urine samples for purposes of the Department's drug testing 
program. The Act requires split samples to be used for testing under 
the Federal Highway Administration (FHWA), Federal Aviation 
Administration (FAA), Federal Transit Administration (FTA), and Federal 
Railroad Administration (FRA) rules.
Mandatory Use of Split Sample Method
    The NPRM proposed to implement the statutory requirement for split 
samples in drug testing by making mandatory the optional split sample 
procedure in the existing part 40. The procedure would remain optional 
under the Research and Special Programs Administration (RSPA) and Coast 
Guard drug testing rules, which are not affected by the Act. Several 
commenters wanted the split sample procedure to remain optional in all 
modes. Because the statute requires the use of split samples in the 
four OAs mentioned above, the Department cannot adopt this comment. In 
order to give employers time to prepare to use the split sample 
collection method, the rule does not require affected employers to 
begin using this method until 6 months from the date of this rule's 
publication. Employers, who under the existing rule have the option of 
using this approach, may begin using the split sample method at any 
time.
Sample Volume
    The NPRM proposed that the total amount of urine collected be 45 ml 
(30 ml for the primary specimen and 15 ml for the split specimen). The 
existing rule calls for a 60 ml collection; the Department believed 
that this was a greater quantity than is needed. Eighteen comments 
supported the NPRM proposal; two commenters opposed the proposal, one 
of whom supported collecting 60 ml each for the primary and split 
specimens. Based on information about laboratory testing needs gained 
over the course of four years of implementing a drug testing program, 
the Department is persuaded that 45 ml (30 ml for the primary specimen 
and 15 ml for the split specimen) is sufficient. This reduction from 
the current 60 ml minimum should also reduce ``shy bladder'' situations 
in which a test is canceled for lack of sufficient specimen volume.
Time Period for Requesting Test of Split Specimen
    Another subject of interest to commenters was the time frame in 
which employees could request a test of a split specimen. The NPRM 
proposed a 72-hour period, following the employee's being informed of a 
verified positive test, during which he or she could request a test of 
the split specimen. Twenty commenters favored this approach, saying 
that this period was sufficient to allow an employee to make a choice 
about whether to request the test of the split specimen. Some of these 
commenters also asserted that allowing the much longer times permitted 
under some OA regulations (e.g., 60 days) could lead to tests of 
deteriorated samples and unreasonably postpone employer disciplinary 
actions. Seven commenters suggested a longer time frame (e.g., a week, 
20 days, 30 days, or 60 days). One of these comments asserted that 
employees needed a longer time to become aware of their rights, study 
their options, and seek representation. Three commenters favored a 
uniform time frame applicable to all OA rules, while one favored 
allowing each OA to set its own time frame. One commenter asked whether 
medical review officers (MROs) were required to inform employees of the 
time period available to request a test of a split specimen.
    The Department will adopt, on a uniform basis, the 72-hour time 
period. The Act requires the Department's procedures to provide for a 
test of the split specimen ``if the individual requests the independent 
test within 3 days of being advised of the results of the confirmation 
test.'' To comply with the statute, the Department is not required to 
provide a time period longer than 72 hours.
    Moreover, the Department has not seen a persuasive rationale for 
permitting a longer time period. Nothing prevents an employee who is 
told of a verified positive test from deciding in a very short time to 
seek a test of the split specimen. For example, some employees testing 
positive admit that they used drugs. Such employees may well not 
believe that testing the split specimen is necessary. If the employee 
concedes that the test was accurate, but contends that the MRO should 
have verified the test negative based on information concerning 
legitimate use of a drug, the employee is likely to seek redress other 
than a test of the split specimen. If, on the other hand, the employee 
is adamant that he or she never used a prohibited substance, or 
believes that the laboratory erred, the employee may well seek a test 
of the split specimen. None of these decisions on the employee's part 
need take more than 72 hours. Decisions concerning legal options, 
representation etc. can be made in the time frames appropriate to the 
processes involved: the decision on whether to seek a test of a split 
specimen need not wait on a decision about whether or how to make use 
of a grievance procedure, for example.
    By saying that the 72-hour time period for requesting a test of the 
split specimen is a uniform requirement, we mean that any time an 
employee makes a request for a split specimen test within 72 hours of 
being informed of a verified positive test, the split specimen must be 
tested. Except in the limited circumstances discussed below, employers 
or MROs are not required by part 40 to provide for a test of a split 
specimen if the employee makes the request more than 72 hours after 
being informed of a verified positive test. There is no information in 
the rulemaking record to support the need of employees in any 
particular industry for a longer time period. Nothing in this provision 
prohibits an employer from voluntarily (e.g., as part of a labor-
management agreement) honoring a request for a test of a split specimen 
made after 72 hours.
    The suggestion that MROs inform employees of this time period is a 
good one. To make the 72-hour period for making a choice on testing a 
split specimen meaningful, it is necessary to ensure that the employee 
knows about the timeframe. For this reason, we have added to the final 
rule a requirement that the MRO notify each employee about this choice. 
We have inserted parallel language concerning requests for the 
reanalysis of the primary specimen in situations (i.e., under the Coast 
Guard and RSPA drug rules) where the split sample collection method is 
not used.
    Under the final rule, when the MRO tells the employee that he or 
she has a confirmed positive test, the MRO must also tell the employee 
that he or she will have 72 hours following notice of a verified 
positive test in which to request a test of the split specimen. This 
notification is required in all cases of confirmed positive laboratory 
results, except in those situations in which an employee has 
effectively waived the opportunity to talk to the MRO. The 72-hour 
clock does not start to run until the time when the employee is 
notified, whether by the MRO or the employer, that the test result is a 
verified positive.
    The employee is not required to wait until after a verified 
positive test in order to request an analysis of the split specimen. An 
employee could, if he or she chose, ask the MRO at the time of the 
notification of a confirmed positive test to initiate the test of the 
split specimen. The MRO would satisfy this request. The verification 
process would continue, and the MRO would notify the employer of the 
verified result in the usual way. The verification and notification 
processes would not be on hold pending the result of the analysis of 
the split specimen. Such a delay in removing from performance of a 
safety-sensitive function an individual with a verified positive test 
could not be justified on safety grounds. Once a test is verified as 
positive, the employee must be removed from safety-sensitive functions. 
The employee may not again perform safety-sensitive duties until he or 
she has met the conditions of the applicable operating administration 
rule for return to duty, pending the result of the test of the split 
specimen.
    In any situation in which the MRO does not personally notify the 
employee of a verified positive test, we advise the MRO, upon receipt 
of a request from an employee to test the split specimen, to contact 
the employer or other party for verification of the time the employee 
was notified of the verified positive test. This should help to avoid 
potential questions about whether the employee has made a timely 
request.
    In addition, to ensure that employees are not unfairly deprived of 
the opportunity to request a test of the split specimen, the Department 
is adding a provision to allow an employee who fails to request this 
test within 72 hours to present information to the MRO that the failure 
to make a timely request was caused by circumstances beyond the 
employee's control. This provision is similar to one in the existing 
rule concerning an employee's opportunity to convince the MRO that 
there was a good reason for the employee's failure to contact the MRO 
for verification purposes (see Sec. 40.33(c)(6)). If the employee 
persuades the MRO, the MRO would initiate a test of split specimen, 
even though the employee's request had been made after the 72-hour 
period ended.
Number of Collection Containers
    With respect to the collection itself, the NPRM proposed that the 
employee provide the specimen into a collection container, which would, 
in most cases, be subdivided and poured into two separate specimen 
bottles. One commenter favored the proposed approach; six others said 
that a two-container, rather than three-container approach, made more 
sense. That is, in all situations--not just unusual situations, as the 
NPRM proposed--the employee should urinate into a specimen bottle, 
which would become one specimen. The collection site person would then 
pour an amount of the urine from that bottle into a second bottle, 
which would become the other specimen. Commenters said this approach 
would save time and money.
    The Department believes that these comments have merit, and the 
final rule permits either approach. The employer could use a collection 
container with the specimen subdivided and poured into two specimen 
bottles. Alternatively, the employer could use a specimen bottle 
capable of holding at least 60 ml, into which the employer would 
urinate. The specimen would then be subdivided, with 30 ml being poured 
into a second specimen bottle, which becomes the primary specimen for 
testing purposes. The original specimen bottle, into which the employee 
had urinated, would become the split specimen.
    This latter point may seem counter-intuitive, but there is a reason 
for it. We want to make sure that there is a 30 ml primary specimen. 
Pouring 30 ml of the void into the second specimen bottle ensures that 
this will be the case. If the instructions were to pour 15 ml of the 
void into the second bottle, to be used for the split specimen, the 
primary specimen might wind up with less than 30 ml of urine if the 
collection site person overpoured. Laboratories have informed the 
Department that they intend to provide only 60 ml bottles to collection 
sites, because of the economies of mass producing a single size 
container and to avoid confusion by collection site personnel. For this 
reason, the final rule's procedure should not result in extra costs.
Storage of Split Specimens
    Three commenters recommended that employers be authorized to store 
split specimens at the collection site rather than send them to the 
laboratory, in order to reduce shipping costs. The Department is not 
adopting this suggestion. Generally, laboratories have better, more 
secure storage facilities than many collection sites. The chances of 
loss, deterioration, tampering, etc. of a specimen are likely to 
increase in non-laboratory locations. A uniform procedure for storage 
and re-shipment of split specimens is likely to reduce opportunities 
for error in the system. The rule also addresses the issue of how long 
the split specimen should remain in storage. As noted above, the 
employee must notify the MRO within 72 hours of being informed of a 
verified positive test to trigger a requirement for a test of the split 
specimen. Consequently, it is not necessary for the laboratory to 
retain the split specimen for a prolonged period. In the Department's 
view, it is sufficient to require the split specimen to be stored 60 
days from the date it arrives at the laboratory, if a request for 
testing it has not been received. (The primary specimen would remain in 
storage for one year, as under the existing rule.)

Choice of Alcohol Testing Methods and Devices

NPRM Proposal
    The NPRM for alcohol testing procedures proposed that both the 
initial and confirmation tests would be done on an evidential breath 
testing device (EBT). An EBT is a breath testing device that is on the 
National Highway Traffic Safety Administration's (NHTSA) Conforming 
Products List (CPL), a list of breath testing devices that NHTSA has 
approved for use by law enforcement agencies in drunk driving cases. In 
addition, the EBTs would have to print out results and assign a 
sequential number to tests, to ensure that test results were preserved 
in a way that minimized the chances for human error or collusion (e.g., 
the disregarding of an initial positive test by an employer who did not 
want to lose an employee's services).
    The NPRM also proposed training requirements for breath alcohol 
technicians (BATs), who would administer the tests, and maintenance and 
calibration requirements for EBTs. In requiring EBTs for all testing, 
DOT proposed that other testing methods--blood, saliva, urine, non-
evidential breath, performance testing--could not be used for either 
screening or confirmation tests. In summary, the Department made this 
proposal because EBTs are a well-established, reliable, and accurate 
testing method; EBTs are minimally intrusive; EBTs can provide an on-
the-spot result that allows employers to take action that prevents 
potential safety risks; and EBTs can produce a printed record of the 
test result that will prevent disputes about the accuracy and integrity 
of the testing process.

Comments

Overview

    This proposal generated more comments than any other feature of the 
NPRM. Approximately 190 of the comments to part 40 addressed some 
aspect of testing methodology. These comments came from a variety of 
sources, including employers in all the industries covered by the 
proposed regulations, unions, laboratories, manufacturers of testing 
equipment and products, and consortia and third-party testing service 
providers. The most consistent theme among comments on this subject was 
a desire for greater flexibility in the choice of testing methodology 
than the NPRM proposed.

Support for NPRM Proposal

    Twenty-six comments, representing employers in several industries, 
unions, third-party testing services, manufacturers of breath testing 
equipment, state police agencies, and the National Transportation 
Safety Board, supported the NPRM proposal. They cited as reasons for 
their support the non-invasiveness of breath testing, its long 
acceptance by courts and employees, its provision of a quantitative 
readout, simplicity compared to blood or urine testing, and the 
relatively low operating costs involved. Some of these commenters 
qualified their support of the NPRM proposal by saying that breath 
testing, while a good method, should be one of an array of options 
available to employers, or required only for certain types of testing 
(e.g., pre-employment and random) where the employer has control over 
the time and place of testing.

Concerns About Cost of NPRM Proposal

    Eighty commenters, representing principally employers in all the 
regulated industries, third-party testing service providers, and 
manufacturers of other testing devices that compete with EBTs, said 
using EBTs for both screening and confirmation tests was too expensive. 
They quoted capital costs per EBT between $2-10 thousand (some EBT 
manufacturers who commented agreed with the lower end of this range). 
This cost would be multiplied, they believe, by a need to obtain EBTs 
for all the locations in which employers operate. For example, a 
trucking association cited a motor carrier that would have to buy an 
EBT for each of its 600 locations, at an estimated cost of $1.2 
million. In addition, there would be BAT training, maintenance, and 
calibration costs. Commenters who talked in cost per test terms cited 
estimates of between $20-100 per test, which they said was much higher 
than for competing methods. Railroad industry employers (who now use 
breath testing for alcohol) said that, to reduce capital costs, EBTs 
should not be required to have the sequential numbering and printout 
capabilities proposed in the NPRM (which they said would add $1500 to 
the cost of an EBT).

Concerns About Difficulty in Implementing NPRM Proposal

    Some commenters feared that there would be insufficient numbers of 
EBTs, BATs, and testing sites available to implement the proposal. 
There would be a rapid expansion of the need for EBTs (one commenter 
estimated a 3000-4000 percent increase in the market) that 
manufacturers may be unable to fulfill, as well as a rapid training 
need for thousands of BATs that would take substantial time to meet. 
Seventeen commenters (including a number of third-party service 
providers and employers) said that the cost of obtaining EBTs and 
training BATs, the unfamiliarity of many third-party testing sites with 
breath testing, and liability concerns would deter many potential 
third-party service providers from participating. This would 
particularly be a problem in small towns and rural areas, where the low 
volume of testing would make the needed investment too costly.

Concern About Confrontations

    Twenty-eight commenters (principally third-party service providers 
and employers) expressed concern about the possibility of 
confrontations between BATs and employees. These confrontations would 
occur, commenters said, because the BAT--not an employer representative 
with supervisory authority over the employee--would be the messenger of 
bad news about a test result. Several commenters cited the image of a 
90-pound female BAT having to deal with an angry (and perhaps 
intoxicated) 300-pound truck driver who had just been told he had 
failed an alcohol test.

Other Comments About NPRM Proposal

    Commenters expressed other concerns about the EBT-EBT approach. 
Some found the process too time-consuming. Others pointed out that the 
collection site is commonly recognized as the weak point of the drug 
testing process, and that conducting the alcohol testing process there 
increased the chance of error. Other comments said that there were too 
many opportunities for human and mechanical error in the breath testing 
process, which, together with what they regarded as the unreliability 
of EBTs at low alcohol concentrations, created numerous opportunities 
for litigation. Some commenters also said that, if all screening and 
confirmation testing were done on EBTs, the two tests should be run on 
different machines.

Legal Issues

    Several commenters raised legal challenges to the proposal. Nine 
commenters (primarily manufacturers of competing devices and unions) 
said that the statute requires split samples (i.e., the subdivision and 
retention of a portion of a sample for an additional test at a 
laboratory as a safeguard for the accuracy of the process) in all 
cases. Generally, EBTs do not retain breath samples. Therefore, these 
comments said, methods that permitted split samples (e.g., blood, 
urine, saliva) must be used. Thirty-one comments said that the statute 
contemplated the use of different methods for the screening and 
confirmation test, respectively. Eleven comments said that, since the 
results of EBT tests would be used to refer persons for rehabilitation 
or treatment, they would be considered medical devices subject to 
Department of Health and Human Services (DHHS) regulation. Since DHHS 
had not approved EBTs as medical devices, their use could be blocked.

Desire for More Flexibility

    Seventy-five commenters (representing a wide variety of equipment 
manufacturers, employers, and third-party service providers) favored 
allowing employers to choose the best testing method for them. In 
addition to the virtue of flexibility, this approach would permit each 
employer to choose the most cost-effective method of compliance in its 
own circumstances.
    Most of these commenters appeared to favor testing methods that 
would use two different testing methods (e.g., non-evidential breath or 
saliva screening test, blood test for confirmation). Ten commenters 
disagreed on this point, saying that non-evidential screening tests 
should never be permitted. Their primary concern was about the accuracy 
of these testing methods. Several commenters who favored using non-
evidential screening tests conceded that it would probably be necessary 
to suspend an employee's performance of safety sensitive functions 
pending a confirmation test of a positive non-evidential screening 
test. Most commenters who addressed confirmation procedures in a two-
method system said that confirmation tests (of whatever body fluid) 
should be done on GC (gas chromatography, the same highly accurate 
method used for confirmation tests under the drug testing program).

Specific Comments on Other Testing Methods

Non-Evidential Breath Testing Devices
(e.g., tubes filled with materials that turn a certain color when 
alcohol-laden breath is blown into them or small, hand-held electronic 
devices that register the presence or absence of alcohol concentration 
in breath)

    Twenty-nine commenters, including a variety of employers and 
manufacturers of the devices, supported using non-evidential breath 
testing devices. Most commenters cited cost (estimated at between $90-
550 for various models of non-evidential breath testing machines, and 
about $2-4 each for disposable devices) and convenience as reasons. A 
few opponents of non-evidential breath testing devices said their 
accuracy was questionable, both with respect to false positives and 
false negatives.
Saliva Testing
(i.e., a device which registers a particular alcohol concentration when 
a swab with saliva from the employee's mouth is inserted into it)

    Forty-five commenters favored the use of saliva testing. These 
commenters included a variety of employers, third-party service 
providers, equipment manufacturers, and others. Commenters claimed 
several advantages for use of screening saliva tests: modest cost 
(estimated at between $5-20 per test); simplicity of use, little need 
for training; existing ``approvals'' from NHTSA and Food and Drug 
Administration (FDA) for some devices (though in contexts other than a 
workplace testing program); non-invasive nature of the devices; 
sufficient accuracy for screening tests. Two commenters also said that, 
while it was most typical to use blood testing for confirmation after a 
saliva screen, saliva specimens could also be used for confirmation, as 
laboratories could run a gas chromatography analysis on saliva.
    A few commenters expressed concerns about saliva testing devices. A 
union provided data that it said showed that saliva devices had a mixed 
record for accuracy. Other commenters said saliva remained an unproven 
method, that saliva devices were not ethanol-specific, and that saliva 
alcohol and blood alcohol results may differ. Proponents of saliva 
testing devices conceded that chain of custody forms would be needed 
and that there was no method of automatically generating permanent 
records of test results that positively identified a particular 
employee with a particular result. They said that keeping paper records 
was adequate for this purpose, however.
Blood Testing
    Forty-eight commenters (again representing a variety of employers, 
plus third-party providers, laboratories and others) favored allowing 
the use of blood testing as a confirmation test method. The advantages 
cited for this method included well-established scientific and legal 
acceptance for accuracy, the availability almost anywhere of 
technicians trained in drawing blood, and utility for post-accident 
testing on employees who are unconscious. Some of these commenters said 
that, while blood testing is admittedly more invasive than other 
methods, employees accept it because of its reputation for accuracy. 
Also, they said, the low expected positive rates on screening tests 
will mean that few blood confirmation tests would have to be performed. 
Commenters estimated costs to be in the $20-60 range per test.
    Seven commenters opposed the use of blood testing, primarily on the 
ground that it is too invasive. In addition, a few commenters said that 
DHHS or DOT would have to develop laboratory certification standards 
for blood testing. Some comments said that employees might have to be 
required to ``stand down'' during the interval between the blood 
collection and the return of the test result from the laboratory.
Urine Testing
    Eight commenters favored allowing the use of urine testing, 
including some employers who now use this approach to their 
satisfaction and laboratories that do urine testing. One advantage 
cited for this approach is that alcohol could simply be added to the 
list of substances for which urine samples taken for drug testing are 
tested, at a low incremental cost. Commenters said that DOT or DHHS 
should develop laboratory certification procedures and cutoff levels. 
Some commenters also noted that detailed collection procedures would 
have to be developed, since urine testing for alcohol is more 
complicated than urine testing for drugs (e.g., two voids, twenty 
minutes apart, are recommended to measure alcohol concentration in 
urine).
Performance Testing
    Five commenters, most of whom were manufacturers of the devices, 
supported the use of performance tests for the screening or screening 
test. (A performance test does not measure alcohol concentration; it 
measures deviations from a personal norm of reaction time, motor 
coordination, etc.) One commenter opposed performance testing devices 
as inappropriate for this program.

Responses to Comments on Testing Methods

Legal Issues

    The Act provides, with respect to confirmation testing, that all 
tests * * * shall be confirmed by a scientifically recognized method of 
testing capable of providing quantitative data regarding alcohol * * *

Some comments asserted that this provision requires that a different 
testing method be used for the screen and confirmation tests, 
respectively. The statute says no such thing, stating only that the 
confirmation test must use a ``scientifically recognized'' method that 
can provide ``quantitative data'' regarding alcohol. As long as the 
method of confirmation meets these criteria, the statutory requirement 
is satisfied. Breath testing is scientifically and legally recognized 
as a method for accurately testing alcohol concentration, and devices 
meeting the Department's requirements provide quantitative data. (Blood 
testing, of course, also meets the statutory criteria.)
    The ability of a method of confirmation testing to pass these 
statutory tests is not dependent on the choice of a method of screening 
testing. Testing of breath for confirmation, as provided in this rule, 
is equally valid under the statute whether evidential breath testing, 
non-evidential breath testing, or saliva is used for the screening 
test. Testing of blood for confirmation is equally valid under the 
statute whether blood, breath, saliva or urine is used for the 
screening test. All that matters is that the confirmation testing 
method meet the statutory criteria in its own right.
    With respect to split samples, the Act requires the Department's 
regulations to provide that each specimen sample be subdivided * * * 
and that a portion thereof be retained in a secure manner to prevent 
the possibility of tampering, so that in the event the individual's 
confirmation tests results are positive the individual has an 
opportunity to have the retained portion assayed by a confirmation test 
done independently at a second certified laboratory if the individual 
requests the independent test within 3 days after being advised of the 
result of the confirmation test * * *

Some commenters asserted that this language should be read to require 
that split samples be used in all alcohol testing, with the implication 
that a method that did not permit the use of split samples could not be 
used. Since most EBTs--including those proposed by the Department in 
the NPRM--do not retain a sample that could theoretically be subdivided 
and preserved for testing of a split specimen, some of these commenters 
asserted not only that blood or other liquid-based testing methods were 
required, but that breath testing was prohibited.
    This interpretation is flatly contrary to the statute, which 
specifically contemplates the use of breath testing (see, e.g., sec. 
3(a) of the Act, adding section 614(d)(6) to the Federal Aviation Act). 
Breath testing is a well-recognized form of alcohol testing, and there 
is no evidence that Congress had any intention of prohibiting its use, 
either indirectly by requiring split samples or otherwise. The 
legislative history makes clear that the Senate sponsors of the 
legislation intended that breath testing be used and that split samples 
were not mandated for breath testing. In the floor debate, during a 
colloquy between Senators Danforth and Hollings, Senator Hollings 
stated

    [t]here are also requirements for split samples, primarily 
included in the legislation to allow urine samples to be retested. 
DOT would have the authority to determine that blood samples should 
be similarly handled. This specific requirement is not relevant in 
the case of breath testing for alcohol, but DOT is directed by this 
legislation to provide necessary safeguards in this area to ensure 
the validity of test results.

137 Cong. Rec S 14764, 14770.

    There is also internal evidence in the wording of the statutory 
provision that supports the reasonable interpretation that the split 
sample requirement is intended to apply to liquid body fluids like 
urine and blood, but not to breath. The statute uses the word 
``samples'' in ways that refer primarily to samples of liquid body 
fluids. For example, section 614(d)((1) of the amended Federal Aviation 
Act refers to the need for ``privacy in the collection of specimen 
samples.'' Privacy is very important with respect to collection of 
urine samples for drug testing. Because elimination functions are not 
involved, privacy is not as important in breath collections. In 
paragraph (d)(6) of the same section, the statute refers to detecting 
and quantifying ``alcohol in breath and body fluid samples, including 
urine and blood.'' In this language, the phrase ``including urine and 
blood'' is best understood as modifying ``body fluid samples,'' as 
opposed to ``breath.'' Given the way that the term ``sample'' is used 
in these portions of the statute, the use in paragraph (d)(5) of 
``sample'' should also be used to refer to liquid body fluid samples 
(i.e., urine and blood). When this paragraph speaks of the ``specimen 
sample be[ing] subdivided,'' then, it is imposing a split sample 
requirement on blood and urine, not on breath.
    Some commenters argued that the language mentioned above from 
paragraph (d)(6), requiring the Department to ``ensure appropriate 
safeguards for testing to detect and quantify alcohol in breath and 
body fluid samples, including urine and blood * * *,'' creates a right 
for employees to have a screening test confirmed by blood testing. This 
language, on its face, does not create such a requirement, since it 
does not specify any particular sort of test for either screening or 
confirmation purposes. There is ambiguous legislative history on the 
point, with the Senate report on the Act saying both that ``an employee 
testing positive for alcohol using a specimen other than blood shall be 
entitled, at that employees [sic] option, to a blood test'' and that 
``the Committee has not specified the type of test to be used in either 
the screening or confirmation test.'' Given that the statute does not 
explicitly require blood testing for confirmation, and that the portion 
of the statute that mandates confirmation testing requires only a 
``scientifically recognized'' confirmation test that can produce 
``quantitative data'' (criteria that breath testing clearly meets), the 
Department does not believe it would be reasonable to view this 
ambiguous legislative history as a mandate for the availability of 
blood confirmation testing in all cases.
    The Department does not believe that regulations of the Food and 
Drug Administration (FDA) would interfere with the implementation of 
breath testing under this rule. FDA does regulate the safety, labeling, 
etc. of medical devices. It is our understanding that FDA may be 
considering initiatives to regulate EBTs used as medical devices in 
medical settings. FDA does not, however, regulate or certify the 
precision or accuracy of EBTs that are currently used for law 
enforcement purposes or that would be used under the DOT alcohol 
testing program. (These would not be viewed as medical devices used in 
medical settings.) We believe that current FDA rules are, and future 
FDA rules would be, consistent with NHTSA certification of EBTs.

Flexibility and Cost

    Many commenters made flexibility in testing methods a high 
priority. The Department agrees that flexibility is desirable. However, 
the Department also believes that any testing system should meet a 
series of criteria, each of which is necessary to execute the statute 
faithfully and to ensure that the safety and accuracy goals of the 
program are met. The Department cannot emphasize too strongly the 
importance of ensuring accuracy and reliability of testing devices and 
methods, at both the screening and confirmation test stages. This is 
needed, among other reasons, to protect employees from even temporarily 
being identified as misusers of alcohol. In the context of drug testing 
litigation, the courts, in upholding the Department's program, relied 
to a substantial extent on the reliability and accuracy safeguards in 
that program.
    Within these constraints, our objective is to provide maximum 
flexibility and minimum cost. The Department's criteria for carrying 
out its objectives in this area are the following:
     As required by the statute, the method used for 
confirmation should be scientifically recognized and able to produce a 
quantitative result. The method should meet NHTSA Conforming Products 
List (CPL) standards at 0.02 and higher alcohol concentrations.
     The confirmation method should be alcohol-specific (i.e., 
does not produce a reading for acetone).
     The confirmation method should generally provide 
documentation of quality control/calibration and be admissible as 
forensic evidence in administrative proceedings.
     The testing method used for confirmation should provide a 
result at the time and place of the test, so that an employee whose 
continued performance of a safety sensitive function may present a 
safety risk can be removed from performing that function.
     The testing method used for the screening test should 
minimize the occurrence of false positives and false negatives and 
should meet stringent standards for precision and accuracy (e.g., +/- 
.005 at 0.02 alcohol concentration).
     The testing method used for screening tests should provide 
a result at the time and place of the test and be specific for 
measuring alcohol concentration.
     The testing methods used for confirmation tests should 
provide a printed, permanent record of the test number and test result, 
in order to avoid uncertainty about whether this employee took this 
test with this result. The testing methods used for screening tests 
should provide either this kind of record or be used in conjunction 
with procedures that provide a record of the test result linked to the 
individual tested through some form of permanent documentation. The 
purpose of this criterion is to prevent collusion and cheating.
     The testing methods used for screening and confirmation 
tests should, as a policy matter, be as non-invasive as possible.
    At the present time, only evidential breath testing methods meet 
all these criteria for screening and confirmation tests. Applying these 
criteria strictly would result in a final rule that, like the NPRM, 
permitted only evidential breath testing for both tests. The points 
made by commenters favoring the NPRM approach further support using 
evidential breath testing for both tests.
    The Department, to achieve a reasonable balance between the legal 
and policy goals on which the criteria are based and commenters' desire 
for greater flexibility, is modifying the approach proposed in the 
NPRM. First, the final rule will permit EBTs that are on the NHTSA CPL, 
but that do not meet the additional requirements for confirmation EBTs 
(e.g., sequential numbering and print-out capability), to be used for 
any screening test. While these EBTs may be used for screening tests at 
this time, because NHTSA has determined them to meet appropriate 
accuracy and precision standards, non-evidential breath screening 
devices (e.g., ``breath tubes'') may not be used at this time.
    Second, in an NPRM published in today's Federal Register, the 
Department will propose to permit blood testing to be used in limited 
circumstances. In the case of a reasonable suspicion test or a post-
accident test, where an EBT meeting the requirements of part 40 is not 
readily available, the employer could use blood testing for the 
confirmation test. Blood alcohol testing would also be available as an 
option in ``shy lung'' situations. This NPRM also proposes blood 
testing procedures to be used in these circumstances. The rationale for 
allowing this limited use of blood testing is discussed in the preamble 
to the NPRM.
    Third, the Department is also publishing in today's Federal 
Register a notice proposing to adopt criteria and procedures that would 
permit additional alcohol screening devices to be used for screening 
tests in the program. This proposal would be intended to result in the 
adoption of model specifications for a conforming products list for 
alcohol screening devices. Under this proposal, manufacturers of 
devices could submit their products to DOT for evaluation and, if their 
devices met the model specifications, the Department would authorize 
their use as screening devices in DOT-mandated alcohol testing. This 
approach will permit greater flexibility in the use of screening 
devices that are not now appropriate for use, including those supported 
by their manufacturers and others in comments to the part 40 docket, if 
they are able to meet DOT model specifications.
    With respect to costs, commenters had three basic concerns, First, 
commenters believed that EBTs meeting all the NPRM's requirements would 
be too expensive. Some commenters believed that adding features such as 
a sequential numbering and printout capability would add considerably 
to the cost of the devices. The Department's information, included in 
our regulatory evaluations, and based on data obtained from 
manufacturers, suggests that the list price per unit of an EBT meeting 
all the NPRM criteria for use in confirmation tests is about $2000. 
(There are some indications that prices may be lower for purchases in 
quantity.) There are other EBTs on the CPL, available under the final 
rule to be used for screening tests, that list for about $1000, again 
with the possibility of lower prices for purchases in quantity.
    Because the Department is proposing to permit blood testing in 
post-accident and reasonable suspicion situations where a breath 
testing unit is not readily available, the numbers of EBTs that any 
employer would have to obtain may be reduced significantly from earlier 
estimates, lowering many commenters' estimated capital costs of the 
program. This is because employers would not have to provide an EBT at 
all its work sites against the contingency of a reasonable suspicion or 
post-accident test happening there, as a number of employers' estimates 
assumed. Commenters identified having to pre-position EBTs at all work 
sites, even the small and remote ones, as a major cost of compliance 
with the NPRM (even though the NPRM would not have imposed this 
requirement). In addition making blood testing available means that the 
time workers would be held out of service pending a test would be 
reduced significantly, resulting in further savings. We refer 
commenters to today's NPRM on blood alcohol testing for further 
information.
    Second, commenters expressed concern about the costs of training 
personnel and maintaining and calibrating the instruments. While 
training can be expensive, we believe that these costs are difficult to 
avoid if the accuracy and integrity of the testing program are to be 
protected. As other devices are approved under the Department's 
forthcoming procedures, employers will have the opportunity to 
determine if use of other methods will reduce their overall costs.
    Third, some commenters (especially from the railroad industry) who 
already use EBTs expressed concern about the costs of the additional 
features that the NPRM would have required (e.g., sequential numbering 
capacity, print-out capability). The final rule responds to these 
concerns by allowing EBTs without these features to be used for 
screening purposes. A railroad could use its existing EBTs (assuming 
they are on the NHTSA CPL) for screening tests, while obtaining only as 
many of the machines with the additional features as it needed for 
confirmation testing. This would reduce the additional costs that these 
employers would have to incur.
    When the Department issues a broad mandate for employee testing, 
the overall effect is likely to be the creation of additional 
opportunities for professionals, manufacturers, and other businesses to 
serve the markets created by the DOT requirements. These opportunities 
can fairly be expected to lead to an influx of participants into the 
market. There is ample evidence that this has been the case in the 
Department's drug testing program, and it is reasonable to expect that 
similar economic opportunities will draw businesses and professionals 
into the alcohol testing market. The Department believes that this 
factor is likely to outweigh, by a substantial margin, any deterrent 
effects on participation in the program related to equipment or 
training costs, the newness of the procedures, liability, or the 
willingness of businesses and professionals to participate.
    Comments that potential participants would be deterred for these 
reasons were, for the most part, speculative. Given the market's 
response to the drug testing rules since 1988, it is fairer to assume 
that the market's response to the even larger-scale alcohol testing 
program will not be timid. With respect to the issue of sufficient EBTs 
being available, the Department has contacted EBT manufacturers, and we 
do not anticipate any serious shortage of devices as the program begins 
operation. If, at any time, the Department learns that there are 
inadequate supplies, the Department could postpone or otherwise modify 
its rules.
    While the image of a large, angry, intoxicated employee confronting 
a 90-pound female BAT over a positive result is a graphic one, the 
speculation and spotty anecdotal evidence provided by commenters to 
back up their concern on this matter is not sufficient to cause the 
Department to retreat from its position that immediate results are 
needed. (This concern goes to any testing method that provides an 
immediate result, not just to breath testing. It might appear even more 
strongly in a situation in which an individual is told, as the result 
of a non-evidential screen, that he is to ``stand down'' and not work 
for three days while a laboratory test result is obtained.)
    The point of getting an immediate result is safety: if an employee, 
of whatever size, has a higher alcohol concentration than the 
Department's rules permit, the individual should not be performing a 
safety-sensitive function. In the interest of safety, we need to stop 
the individual's performance of that function now, not two or three 
days later when a laboratory test result becomes available. We also 
want to prevent the unnecessary cost of holding an employee out of 
service for two or three days pending laboratory results following a 
non-evidential screen. BATs are not given the responsibility of taking 
a driver's keys away. The DOT alcohol testing form includes a 
statement, to be signed by the employee, that persons who test positive 
should not drive or perform other safety-sensitive functions. Employers 
have a responsibility, as part of their alcohol education for 
employees, to emphasize that employees must cease performing safety 
sensitive functions if they test positive.
    The Department does not believe that it is necessary to use two 
separate EBTs in order to have a valid, defensible result. EBTs on the 
NHTSA CPL are designed for accuracy, and the internal and external 
calibration checks built into the Department's procedures are 
sufficient insurance against error. (Where employers choose to use an 
EBT without the additional features for screening tests, of course, the 
employer will necessarily use a different machine for the confirmation 
test.) The Department is convinced that EBTs meeting its requirements 
are sufficiently accurate and reliable, at the alcohol concentrations 
that will be tested for, and that excessive invalidations of tests or 
successful lawsuits or grievances will not occur. Similarly, the 
likelihood of extensive errors by testing personnel should be 
diminished by the BAT training requirements.
    Manufacturers of alternative testing devices, and some other 
commenters as well, advocated various other methods of testing, 
particularly for screening tests. As noted above, the Department 
intends to take action that could result in decisions to authorize use 
of other screening devices and to authorize the use of blood testing in 
some circumstances. The Department has decided not to permit the use of 
these alternative methods until they can meet the criteria we believe 
are necessary for accurate testing meeting the requirements of the 
statute. The following paragraphs summarize the Department's reasons 
for not permitting the use, at this time, of other testing methods:

Blood Testing

     This is the most invasive form of testing.
     Employees may fear needles or fear infection from 
improper medical procedures.
     Additional collection procedures, chain of custody 
procedures, and equipment requirements would be needed, making 
regulatory requirements more complex.
     Laboratory certification standards and testing 
protocols would need to be established. As noted in the accompanying 
NPRM, this poses potentially significant problems even in the 
limited context in which the Department is proposing to permit the 
use of blood testing.
     Results would not be available for at least 24 hours, 
and could take 3-4 days to arrive. Confirmed results would, 
therefore, not be available at the time the employee was affected by 
alcohol, which would reduce the safety benefits of the program.

Urine Testing

     Present laboratory certification standards and testing 
protocols do not cover urine testing for alcohol. There would have 
to be additional laboratory certification procedures and testing 
protocols developed for urine testing.
     Urine testing for alcohol (as distinct from drugs) 
requires a complex collection process, involving two separate voids 
with an interval between them. Addition of a preservative to prevent 
the creation of alcohol by microbial fermentation is also 
recommended. We would need to add new collection procedures to 
accommodate these requirements, as well as new training requirements 
for collection site personnel. These additional procedures would 
make the collection process more complex and multiply the chances 
for errors.
      Urine testing is regarded as the least accurate method 
currently available for determining the amount of alcohol in the 
body.
      A blood to urine ratio has not been definitively 
established, making it difficult to equate a urine test result for 
alcohol to a particular blood or breath alcohol level.
     There are greater costs of employee ``downtime,'' for 
transporting the employee to a collection site for testing and for 
the longer collection procedure.
     Testing of urine specimens would have to take place in 
a laboratory. Results would not be available for at least 24 hours, 
and could take 3-4 days to arrive. Confirmed results would, 
therefore, not be available at the time the employee was affected by 
alcohol, which would reduce the safety benefits of the program.

Saliva Testing

     Especially at low alcohol levels, saliva devices are 
likely to have a higher rate of false positives and negatives than 
EBTs on the CPL.
     Some saliva devices do not provide quantitative 
results.
     Because saliva screening testing devices are 
disposable, and do not generate a record of the test, ascertaining 
whether a particular employee took a particular test and had a 
particular result, or that the test took place at all, would be 
difficult. (The use of a log book, which helps to address this 
concern where EBTs without sequential numbering or printout 
capabilities are used, would be difficult in the case of disposable 
devices. The log book would accompany the EBT wherever it went, 
which would not be possible with disposable devices.)
     There are different saliva-based technologies, each 
requiring the establishment of criteria for accuracy, reliability, 
etc. Until NHTSA criteria are established for these technologies, it 
is premature to permit their use in the DOT program.
     If laboratory confirmation methods (e.g., blood) are 
used in combination with saliva screens, confirmation results would 
not be available for at least 24 hours, and could take 3-4 days to 
arrive. Confirmed results would, therefore, not be available at the 
time the employee was affected by alcohol, which would reduce the 
safety benefits of the program. If breath testing confirmation is 
used, cost savings claimed for the use of disposable devices over 
the use of breath testing for both screening and confirmation 
testing would be reduced substantially.
     The Department would have to establish additional 
procedures, training requirements, quality control requirements, 
etc. for saliva testing, adding further complexity to the program.

Non-evidential Breath Testing

     Non-evidential breath devices (i.e., disposable devices 
and others not on the CPL) have a higher rate of false positives and 
negatives than evidential EBTs.
     Non-evidential breath screening testing devices do not 
generate a record of the test, so that ascertaining whether a 
particular employee took a particular test and had a particular 
result, or that the test took place at all, would be difficult. (The 
use of a log book, which helps to address this concern where EBTs 
without sequential numbering or printout capabilities are used, 
would be difficult in the case of disposable devices. The log book 
would accompany the EBT wherever it went, which would not be 
possible with disposable devices.)
     If laboratory confirmation methods (e.g., blood) are 
used in combination with non-evidential breath screens, confirmation 
results would not be available for at least 24 hours, and could take 
3-4 days to arrive. Confirmed results would, therefore, not be 
available at the time the employee was affected by alcohol, which 
would reduce the safety benefits of the program. If breath testing 
confirmation is used, cost savings claimed for the use of non-
evidential devices over the use of evidential breath testing for 
both screening and confirmation testing would be reduced 
substantially.
     Non-evidential EBTs on the market appear to vary 
greatly in type of technology used, quality, and accuracy. Until 
NHTSA criteria are established for these devices, it is premature to 
permit their use in the DOT program.
     The Department would have to establish additional 
procedures, training requirements, quality control requirements, 
etc. for non-evidential breath testing, adding further complexity to 
the program.

Performance Testing

     The statute requires testing for alcohol concentration, 
not diminished performance. A test for performance appears not to 
meet this statutory requirement.
     Performance tests are very unspecific, which could 
result in positives caused by a wide variety of things other than 
alcohol use (e.g., illness, prescription or over-the-counter 
medication, fatigue, emotional distress). This would lead to many 
unnecessary confirmation tests and could result in employees being 
taken off the job while awaiting confirmation test results, adding 
extra costs for employers and employees.
     The accuracy of many performance testing devices is 
unproven.
     Many performance testing devices do not generate a 
record of the test. Ascertaining whether a particular employee took 
a particular test and had a particular result, or that the test took 
place at all, could be difficult.
     Most performance testing devices require the 
establishment of individual baseline data for each employee, which 
can be a time-consuming and costly procedure.
     In many systems, performance evaluation must relate to 
critical job skills, measures of which have not been established for 
many occupations.
     Performance testing devices or systems on the market 
appear to vary greatly in quality and accuracy. Until NHTSA criteria 
are established for these devices, it is premature to permit their 
use in the DOT program.
     The Department would have to establish additional 
procedures, training requirements, quality control requirements, 
etc. for performance testing, adding further complexity to the 
program.

    This discussion is in the context of an extensive, multi-modal 
testing program, including pre-employment and random testing as well as 
reasonable suspicion and post-accident testing. Greater protections are 
needed in such a program, particularly in the absence of procedural 
protections present in some existing programs that may use non-
evidential testing in some circumstances. For example, the Coast Guard 
post-accident alcohol testing program can involve administrative 
proceedings in which the employee has the opportunity to challenge test 
results before a license is revoked or an investigative inquiry at 
which further evidence could be introduced.

Breath Alcohol Technicians

    The NPRM proposed that breath alcohol technicians (BATs) be trained 
to proficiency in using EBTs and in DOT alcohol testing procedures, 
using a NHTSA- or state-approved course. The competence of the BAT 
would have to be documented. Additional (i.e., refresher) training 
would be required, as needed, to maintain proficiency. An employee's 
supervisor could not act as the BAT for that employee unless allowed by 
a DOT rule and no other qualified BAT were available.
    Commenters spoke to several provisions of this section. Six 
commenters favored, and 15 opposed, requiring BATs to be tested to 
ensure that they are alcohol free (an issue about which the Department 
had asked a question in the NPRM preamble). A number of the opponents 
said that this issue should be decided by the BATs' employers. The 
Department is not adopting this idea, which we believe to be 
unnecessary to the program.
    Forty-nine comments addressed the training and qualification of 
BATs. All these commenters favored training, though two mentioned that 
training might be very costly or difficult, especially for smaller 
companies. Sixteen comments said that it was not necessary for the 
regulation to specify that BATs be trained in the pharmacology and 
physiology of alcohol, about which the NPRM preamble had asked a 
question. Three commenters took the opposite position. The Department 
agrees that this training is not needed for BATs, whose training should 
be focused on the proper operation of testing devices.
    Seventeen commenters supported the NPRM approach (including the 
concept of ``training to proficiency''), while two thought the NPRM too 
vague. Eleven favored specific numbers of hours of training, ranging 
from 4 to 40, with most of the comments suggesting something between 4 
and 8 hours. Two expressed support of recurrent training, one asking 
for a more specific requirement than the NPRM proposed. The Department 
believes it is most relevant to ensure the BATs' proficiency. Our goal 
is to ensure that BATs are able to use the testing devices that they 
will operate. The Department believes that the best way to make sure 
that BAT training results in proficient operators is to require that 
BAT training include a course that is equivalent to the DOT Model 
Course. Courses followed by state law enforcement agencies and other 
organizations appear to vary substantially from one another, and may be 
focused on breath testing in other contexts (e.g., enforcement of DUI 
laws). NHTSA will review training courses and issue determinations 
concerning whether they are equivalent to the NHTSA Model Course.
    Who should be a BAT? Twenty-two of 23 commenters supported 
permitting a trained law enforcement officer to act as a BAT. The 
Department agrees that it is appropriate to authorize trained law 
enforcement officers to act as BATs (e.g., off-duty officers under 
contract to an employer), as long as they have been certified by a 
state or local law enforcement agency. The officers would have to 
follow DOT testing requirements, including this part, and to be 
certified to operate the EBT used in the DOT-mandated test. The 
officers could perform any type of DOT test. Except for the FHWA rule, 
the OA rules do not permit the substitution of law enforcement tests 
for tests conducted under DOT procedures.
    There was less consensus on the issue of supervisors as BATs. 
Sixteen commenters favored allowing properly trained supervisors to act 
as BATs, pointing out that, particularly in reasonable suspicion or 
post-accident testing, or at remote sites, supervisors may be the most 
readily available, or perhaps the only available, trained BATs. Eleven 
other commenters disagreed, most saying that an employee's supervisor 
should never be the employee's BAT. These commenters appeared concerned 
about the appearance or reality of a conflict of interest between the 
supervisor's managerial role and his objectivity as a BAT. The 
Department believes that, when possible, someone other than an 
employer's supervisor must act as a BAT for the employee's test. 
However, a supervisory BAT is better than no BAT at all. To enable a 
test to go forward when no other BAT is available in a timely manner, 
the Department will permit a BAT-trained supervisor to conduct the 
test. However, if a DOT operating administration regulation prohibits 
the use of a supervisor in this role (e.g., in reasonable suspicion 
testing), the supervisor may not act as the BAT even in this 
circumstance.

EBT Technology

    The NPRM required EBTs used for screening and confirmation testing 
to be on the NHTSA CPL, have the capacity to print out triplicate (or 
three consecutive identical) results, assign a sequential number to 
each test, distinguish alcohol from acetone at the 0.02 alcohol 
concentration level, and have the capability for performing both air 
blanks and external calibration checks. Commenters addressed a number 
of points concerning EBT technology.
    Some commenters pointed to what they viewed as shortcomings of the 
CPL itself, particularly that it did not require EBTs to be accurate at 
the 0.02 level. This was true of the CPL at the time the NPRMs were 
issued; however, NHTSA has since modified the model specifications for 
the CPL to require accuracy and precision at the 0.02 level. Other 
commenters said that since inclusion on the CPL is based on testing of 
a prototype, rather than testing of each device, the CPL was an 
inadequate assurance of accuracy. The final rule does not rely on the 
CPL alone to ensure accuracy, however. The rule requires there to be a 
quality assurance plan (QAP) for the instrument as well as air blanks 
and external calibration checks.
    As noted above, a number of commenters criticized the requirement 
for printing results and sequential numbering capability, saying that 
these features were unnecessarily costly. Any device on the CPL should 
be able to be used, one of these commenters said. The final rule 
responds to these comments by allowing any device on the CPL to be used 
for screening tests, with the additional features required only on 
those machines used for confirmation testing. This should reduce the 
number of the more expensive models employers will have to obtain.
    Some commenters expressed concern about radio frequency 
interference (RFI) affecting the results of some types of EBTs.The 
concern is that, in airports and other locations where communications 
or other electronic equipment is operating, alcohol concentration 
readings could be distorted. DOT asked manufacturers about this issue, 
who said that most models of EBTs are shielded to avoid this problem. 
NHTSA tested three models of EBTs at Washington National Airport and 
detected no RFI effects on their readings. In addition, NHTSA plans, as 
part of its process for reviewing quality assurance plans (see 
discussion below), to have manufacturers establish operational 
guidelines to avoid RFI problems. The Department believes that it is 
not necessary to modify the regulatory text to address the commenters' 
concerns.
    Commenters also expressed concern that some EBTs might not be able 
to distinguish acetone from some alcohols. Commenters also questioned 
the suitability of the CPL for instruments measuring alcohol 
concentrations at the 0.02/0.04 levels, since the CPL, at the time of 
the NPRM, did not address testing at these levels. As noted above, 
NHTSA has revised the model specifications on which CPL listing of 
devices is based. The revised specifications address both issues, and 
EBTs on the CPL will distinguish acetone from alcohol and be accurate 
at the 0.02/0.04 levels.
    A few comments raised other technical issues about the use of EBTs. 
One issue was the effect of altitude on external calibration standards. 
Altitude affects gas aerosol standards; NHTSA will address this problem 
by requiring gas aerosol standards on its CPL for calibration devices 
to be criterion-referenced for various altitudes.
    Another concern was based on the belief that EBTs that display 
results to only two, rather than three, decimal places would round up. 
That is, commenters were concerned that someone whose actual alcohol 
concentration was .036 would be reported as a 0.04, subjecting the 
individual to heavier sanctions. EBTs on the CPL provide three-digit 
displays, so this problem does not arise for these devices.
    Finally, some commenters expressed concern that defining alcohol 
concentration in terms of grams of alcohol per 210 liters of breath was 
not as accurate as desirable (or as accurate as a blood alcohol 
reading), because this ratio could vary among individuals. The 
Department's information is that any variation is very minor and 
unlikely to affect the results of a breath test or its consequences 
under these rules. In addition, EBTs are typically calibrated to 
account for any variation by slightly undercounting alcohol 
concentration.

Quality Assurance Plans

    The NPRM proposed that EBT manufacturers would develop a quality 
assurance plan (QAP) for each EBT model. The plan would cover such 
matters as external calibration methods, tolerances and intervals and 
inspection and maintenance requirements. The manufacturer would have to 
obtain NHTSA approval of the QAP, and employers would have to comply 
with it. This compliance includes making external calibration checks as 
called for in the QAP and taking EBTs out of service if they ``flunk'' 
an external calibration check. In addition, the employer would have to 
ensure that inspection, calibration and maintenance of EBTs is done by 
the manufacturer, a representative certified by the manufacturer, or an 
appropriate state agency.
    On the basic concept of the QAP, five commenters supported the 
NPRM's approach, while another eight said that NHTSA, rather than the 
manufacturer, should establish the standards. Some of the latter 
commenters appeared concerned that manufacturers may have incentives to 
establish requirements for their devices that were not optimal. The 
Department believes that NHTSA approval of the QAPs should be 
sufficient to ensure that the manufacturer's standards are adequate and 
that the manufacturers are better positioned than we are to establish 
model-specific requirements for individual EBTs. For this reason, we 
are retaining the proposed approach. QAPs would be required for all 
EBTs on the NHTSA CPL that would be used in DOT-required alcohol 
testing, whether or not a particular EBT met the additional 
requirements of this part for use in confirmation testing.
    Commenters suggested a wide variety of requirements concerning how 
frequently an external calibration test must be performed. Some of the 
ideas included performing such checks before and/or after every test, 
after every positive test, before, during and after the testing shift, 
every day, after every five tests, every thirty days, or before 
disciplinary action is taken on the basis of a positive test. All these 
comments respond to a basic point: if an EBT ``flunks'' an external 
calibration check, positive tests conducted on that device since the 
last previous successful external calibration check must be regarded as 
invalid. This fact provides a strong incentive to employers and BATs to 
conduct these checks frequently enough to avoid retroactive 
invalidations of positive tests. In conjunction with the manufacturer's 
instructions on the QAP, this incentive should be sufficient to induce 
employers acting in good faith and testers to conduct these checks at 
appropriate intervals. A generally applicable regulatory requirement 
for external checks of calibration at a stated interval, on the other 
hand, would provide less flexibility and might not fit a variety of 
situations well.
    A few commenters suggested specific types of calibration solutions 
or obtaining such solutions from certified laboratories. Others 
suggested that the Department establish particular standards for 
external calibration devices, or allow use of only those external 
calibration devices that are on the NHTSA CPL. Others suggested 
particular tolerance standards (e.g., +/- .005). The Department does 
agree that the employers should use external calibration devices that 
are on the NHTSA CPL, and this requirement has been incorporated into 
the final rule. The Department does not certify laboratories for 
production of external calibration solutions, so we could not 
reasonably require employers to obtain solutions from certified 
laboratories. For the types of solution that work best with a 
particular machine, or for the tolerance standard that is most 
relevant, we believe that reliance on the QAP, based on the 
manufacturer's knowledge of the behavior of its product, makes the most 
sense.
    On the subject of maintenance, most commenters supported the NPRM's 
proposal for maintenance by manufacturers, or their representatives, 
and careful documentation of this activity. These provisions have been 
retained.

Testing Location

    The NPRM called for a testing site that afforded visual and aural 
privacy to the employee, though in unusual circumstances a test could 
be conducted elsewhere. The site would have to be secured. A mobile 
facility (e.g., a van) that met the requirements could be used. At the 
site, the BAT was to supervise only one employee's use of an EBT at a 
time, and the BAT could not leave the site when testing was in 
progress. The Department, with some modifications, is adopting this 
provision in the final rule. In our view, privacy in the context of 
breath alcohol testing is primarily for the purpose limiting other 
persons' access to information about the employee's test result. In 
contrast to urine drug testing, where private elimination functions are 
involved, privacy need not be as strict for breath alcohol testing. We 
have also eliminated references to the site being ``secured,'' as such, 
because this term could lead to confusion. Our concern is that 
unauthorized persons not be in a position to see or overhear test 
results. We are not requiring that testing take place behind locked 
doors, in a totally enclosed space, or in a dedicated facility that is 
not used for other purposes.
    There were few comments on this provision. Two commenters noted 
that privacy could be hard to achieve at a remote site. The NPRM 
already made allowance for this problem, however, by saying that a 
testing location did not have to provide full privacy in unusual 
circumstances such as a post-accident or reasonable suspicion test in a 
remote location. Other comments included a concern that privacy be 
protected adequately, that too much privacy could sharpen the concern 
about confrontations between BATs and employees, and that privacy 
requirements should not exclude a witness (e.g., a union 
representative) from the testing site. The provision establishes a 
general performance standard for privacy of the physical site: It does 
not address the issue of whether a witness may be present (that is a 
matter for labor-management negotiation). It does not require a site 
that is so isolated that a BAT could not find assistance if needed. One 
commenter asked for a DOT-operated national inspection program for test 
sites, analogous to the DHHS laboratory certification program. The 
Department believes that such a system would not be practicable, given 
the very high number of testing sites likely to be involved with the 
program.

Testing Form and Log Book

    The NPRM proposed to require the use of a standard form for DOT-
mandated testing, which employers could not modify. It would be a 
triplicate form, with copies for the BAT, employer, and employee. The 
colors of each copy of the form are intended to be consistent with the 
colors of the Department's drug testing form. The Department has 
decided to adopt this provision with minor modifications.
    Seven commenters supported the NPRM provision as drafted. Thirteen 
commenters favored having space on the form for recording a repeat of a 
test, in order to reduce paperwork. The Department believes that adding 
space for this purpose would result in a longer, more complicated form. 
Moreover, it is likely to be only in a minority of cases that a test 
will have to be repeated, meaning that the extra complexity of the form 
would not serve a useful purpose in most cases. For this reason, the 
Department is not adopting this comment.
    Two commenters suggested that a combined drug/alcohol form be 
developed. The Department responds that, because of the differences 
between drug and alcohol testing, it would be difficult to develop a 
combined form that would not be too cumbersome and would work in both 
situations.
    Two commenters asked that employers be able to modify the form. The 
Department's experience with the drug testing program, where some 
modification of the form has been permitted, is that the resulting 
variety of forms leads to confusion, errors, and difficulty in 
completing the form by collection site personnel. The Department 
believes that an unvarying, standard form will minimize these problems. 
Employers would have to use the form exactly as presented in Appendix A 
to this regulation (though a form directly generated by an EBT could be 
smaller and would not need a space to affix a separate printed result.) 
One commenter suggested that DOT provide the forms to employers free of 
charge. The Department does not believe that this is an appropriate use 
of Federal funds.
    Two commenters asked that the form specify that the test is being 
conducted under the authority of DOT regulations. The Department's 
experience under the drug testing program is that, for lack of such a 
statement, some employees have been confused about whether a particular 
test was being conducted under DOT authority or simply under the 
employer's policy. The form being published with this rule includes 
such a statement. The result of including such a statement is that 
employers are not permitted to use the ``DOT form'' for a test not 
conducted under DOT authority.
    Two commenters questioned the option to have the EBT or printer 
print results directly on the form, preferring to use a separate form. 
The regulation's requirements for EBTs used in confirmation testing 
provides this option, which is appropriate to provide flexibility. An 
employer who is uncomfortable with one approach can use the other.
    This section of the rule includes a new provision requiring the use 
of a log book with EBTs, used for screening tests, that do not have the 
sequential numbering and printing capabilities required for devices 
used for confirmation tests. This section spells out the requirement 
for the log book and what it must contain; the rationale for the log 
book requirement is discussed below.

Preparation for Testing

    The NPRM proposed that the BAT and the employee provide 
identification to one another and that the BAT explain the testing 
procedure to the employee. A commenter suggested that written 
information be provided to the employee, so that the briefing could be 
more detailed and the BAT had less verbal work to perform. The employer 
may provide the information in this fashion, though the regulation will 
not require it. Other comments were few and supportive. The NPRM 
provisions have been retained. Some provisions of this NPRM section, 
concerning filling out of forms and refused or incomplete tests, have 
been moved to the next section.

Initial Breath Test Procedures

    The NPRM proposed to require an air blank before and after the 
screening test, which the machine had to pass in order to stay in 
service. The NPRM also included proposed requirements concerning 
completing the test paperwork.
    Fifteen commenters addressed the issue of air blanks. Seven 
commenters agreed with the NPRM that air blanks should be required 
before and after each screening test. Two said that air blanks are not 
technically relevant with some types of EBTs. Six commenters said that 
an air blank should not be required after a test when the result was 
less than 0.02, as this was a waste of time. Some of these commenters 
favored pre-test air blanks, however. One commenter supported only pre-
test air blanks.
    The Department has decided that it will not require air blanks 
either before or after a screening test. First, most screening test 
results will be below 0.02, making post-test air blanks of limited 
value in those cases. Second, pre-test air blanks, at the screening 
stage, are not crucial in preventing ``false positives'' for employees, 
since no action against an employee may be taken without a confirmation 
test. Third, the Department will require air blanks before confirmation 
tests, which will build this protection into the testing process where 
it matters most. Fourth, the Department is permitting all EBTs on the 
NHTSA CPL to be used in screening tests, and some of these instruments 
would not provide any durable record of an air blank, even if they were 
able to perform air blanks. Finally, the absence of a requirement for 
air blanks on the more frequent screening tests will result in some 
cumulative savings of BAT and employee time and wear on the machines.
    The NPRM called for a 15-20 minute waiting period before the 
confirmation test; no such waiting period was proposed for before the 
screening test. Seven commenters favored a waiting period before the 
screening test, eight opposed it, and two favored employer discretion. 
Because the confirmation testing procedures do provide for a waiting 
period, and since action against an employee can be taken only on the 
basis of a confirmation test, we believe that requiring an additional 
waiting period before the screening test would be superfluous.
    The NPRM provision addressed situations in which the printed and 
displayed results did not match, proposing that such tests would be 
invalid. The final rule modifies this provision, since it is irrelevant 
concerning instruments that do not print out a result. The NPRM 
provision remains in effect for EBTs that do print out.
    The additional flexibility the Department has provided in screening 
testing procedures, by permitting the use of EBTs that do not have 
sequential numbering and result printing capabilities, makes it more 
difficult to determine that a test of a particular employee, with a 
particular result, has taken place, raising the possibility of cheating 
by employers. To mitigate this potential problem, the final rule will 
require a log book to be kept with each EBT used for screening that 
does not have the sequential numbering and printout capabilities. (This 
requirement does not apply to EBTs meeting the requirements for devices 
used for confirmation testing. ) The BAT will fill out a log book entry 
for each test in addition to completing the alcohol testing form. The 
log book entries are intended to serve as a cross-check on the 
performance and result of a test.
    There were several comments both to this section and the next 
section concerning whether the cutoff level for a test to which 
consequences for the employee would attach should be 0.02, 0.04, or, as 
the NPRM proposed, a bifurcated 0.02/0.04 standard, with different 
consequences at each level. The rule takes the latter approach, for 
reasons discussed in the common preamble to the OA rules.
    The employee is told to sign the form after the test has been 
taken. If the employee does not do so, it is not regarded as a refusal 
to take the test. Obviously, it would be silly to regard as a refusal 
to take the test a refusal to sign the form after the test had already 
been successfully conducted. In this situation, the BAT is required to 
not the failure to sign in the remarks section of the form.

Confirmation Breath Test Procedures

    The NPRM instructed the BAT to tell the employee to avoid eating, 
drinking, etc. during a 15-20 minute interval between the screening and 
confirmation test, though the test would continue even if the employee 
did not follow the directions. The BAT would also give the employee a 
notice not to drive or perform other safety-sensitive functions if the 
employee's alcohol concentration were 0.04 or greater. After performing 
the same steps as with the screening test, the BAT would note the 
alcohol concentration reading and transmit the results to the employer 
in a confidential manner. The lower of the two readings--screening and 
confirmation--would control the result.
    There were 29 comments concerning the waiting period before the 
confirmation test, fifteen of which supported the 15-minute minimum 
time proposed in the NPRM. Four comments wanted a shorter interval 
(e.g., two or five minutes) and four supported a longer interval (e.g., 
20 or 30 minutes). Two comments opposed any requirement concerning an 
interval. Six comments either wanted no maximum waiting time or 
preferred to rely on the employer's or EBT manufacturer's discretion.
    The waiting period is important. It is intended to give the 
employee the opportunity to ensure that any residual mouth alcohol does 
not influence the result of the confirmation test. According to the 
Department's information, fifteen minutes is the minimum period after 
which one can be confident that any residual mouth alcohol has 
disappeared. A shorter interval is not feasible for this reason. At the 
same time, waiting a long period between tests can be costly in terms 
of lost employee time and could influence the outcome of the 
confirmation test. In order to guard against lengthy delays in the 
performance of confirmation tests, which can allow alcohol 
concentration levels to fall, the final rule retains the 20-minute 
maximum. It should be pointed out that failing to observe the minimum 
15-minute period is a ``fatal flaw'' (see Sec. 40.79 (a)), 
automatically invalidating a test. This is because the Department 
believes it is important to prevent artificially high readings due to 
mouth alcohol residue. However, taking longer than 20 minutes between 
tests is not a ``fatal flaw.'' The Department is aware that 
circumstances may sometimes result in stretching the time between tests 
for a few additional minutes.
    Another issue addressed by commenters in a variety of ways was that 
of whether the screening or confirmation test result prevails when one 
is higher than the other. Eighteen commenters believed that the 
confirmation test should prevail in all cases. Two commenters supported 
using the higher of the two results, while three supported using the 
lower of the two results. The Department believes that it is more 
understandable, and less potentially confusing, for the confirmation 
test result to determine the outcome of the test. The confirmation test 
will always have to be performed using the most reliable methods. Also, 
alcohol concentration can still be rising at the time of the screening 
test. Although it is also possible for alcohol concentration to have 
dropped since the screening test, the Department's requirement for the 
confirmation test to be conducted a short time after the screening test 
should minimize any problem. Finally, this approach is consistent with 
that the Department takes in drug testing. Consequently, in situations 
in which a confirmation test is needed, the final rule will attach 
consequences only to the confirmation test result.
    Nine commenters asked that the final rule, unlike the NPRM, provide 
for medical review officer (MRO) review of the confirmation test 
result, as the Department requires in drug testing. Among their reasons 
were that there could be valid medical or food-related reasons for 
alcohol concentrations, that there could be inadvertent alcohol 
consumption, that someone should review results for procedural errors, 
that an MRO should play the role assigned to the substance abuse 
professional (SAP) by the proposed rules, or that the alcohol rules 
should mirror the drug rules as much as possible.
    In the drug testing context, an MRO determines whether there is a 
legitimate medical explanation for an individual having in his or her 
system a substance which is otherwise illegal. The alcohol rules are 
different in this respect. They prohibit safety sensitive employees 
from having alcohol concentrations above certain levels, regardless of 
the source of the alcohol. An alcohol concentration of 0.04 resulting 
from drinking beverage alcohol has the same consequences under the 
rules as an alcohol concentration of 0.04 resulting from ingesting 
medication. Both uses of alcohol are legal (as long as they do not 
violate OA rules concerning on-duty use, pre-duty abstinence, etc.); 
the resulting alcohol concentration is prohibited by DOT regulations 
equally in both cases. In this context, there is nothing for an MRO to 
decide. Inserting an MRO into the process without this key function 
would add to the complexity and cost of the system without providing 
any benefits. For these reasons, the Department will not require MRO 
review of alcohol testing results.
    The NPRM proposed that employers could use the same EBT for both 
the screening and confirmation tests. Fifteen commenters objected to 
this proposal. Some said that an entirely different methodology should 
be used for the two tests. The legal issues section of the preamble 
discusses this point. Others said that a different EBT should be used 
for each test, some making the argument that using the same machine for 
both tests constituted ``repetition,'' but not ``confirmation.'' This 
semantic argument is not persuasive. The statute does not require 
different machines to be used, as long as the machine used for the 
second test meets statutory requirements. (Of course, where an employer 
chooses to use a preliminary EBT for the screening device, it will 
necessarily use two different machines.) Because of the reliability of 
EBTs meeting the requirements of this rule, we believe it would be 
unnecessarily expensive to require a second device to be used, which 
could have the effect of roughly doubling the capital equipment costs 
of the program.
    Twelve of thirteen commenters opposed requiring a second 
confirmation test after the first confirmation test had been positive, 
a matter about which the NPRM preamble asked a question. The Department 
does not see a basis for requiring a second confirmation test, and we 
are not adding this requirement to the final rule.
    A few commenters suggested getting rid of the requirement for the 
BAT to notify someone testing positive that he or she should not drive. 
The Department has decided to include a notice to this effect on the 
alcohol testing form, making direct participation by the BAT 
unnecessary.
    Two commenters suggested that the rule be clarified to indicate 
that an employer could have more than one representative to whom 
results are transmitted. The Department has done so.
    Two comments supported, and two opposed, the practice of back 
extrapolation to obtain a result. The Department's NPRMs proposed that 
the consequences of test results attach only to employees whose EBT 
readings were in fact at the stated levels. The Department did not 
propose to attach these consequences to inferences from EBT readings 
about what an employee's alcohol concentration might have been at an 
earlier point. For example, if an employee's EBT test result were .03, 
the requirement that the individual not again perform safety-sensitive 
functions until he or she was evaluated by a substance abuse 
professional (SAP) and had passed a return-to-duty test, and the 
requirement that the individual be subject to follow-up testing, would 
not apply because the employer, SAP, or other party believed that the 
individual's alcohol concentration had been 0.04 or greater prior to 
the test. Given the wide individual variations in alcohol metabolism 
among individuals, such inferences involve considerable uncertainty. 
The Department is retaining the NPRM provision on this point. This 
would not prevent an OA from making use of back extrapolation in 
certain situations (e.g., FRA makes some use of back extrapolation in 
its existing toxicological testing program, in a context involving the 
use of samples of two different body fluids; inquiries into accident 
causation or proceedings to revoke DOT-issued certificates or licenses 
held by employees, where expert testimony can be produced with the 
protection of the due process procedures of a hearing). These 
situations are different from the use of back extrapolation by 
employers in interpreting the results of tests conducted under part 40, 
however.
    There will be some cases in which the BAT who conducts the 
screening test and the BAT who conducts the confirmation test are 
different people. For example, BAT #1 conducts a screening test, using 
an EBT not having sequential numbering or printout capabilities, in 
location A. The confirmation test, using a device that has these 
features, happens subsequently in location B, and is conducted by BAT 
#2. In such a case, to minimize the possibility of lost forms or other 
errors, the final rule provides that BAT #1 would complete the form for 
the screening test and give the employee his or her copy of the form. 
BAT #2 would then start a new form. The sections of the rule concerning 
screening and confirmation testing procedures have been modified to 
this effect.

Refused and Incomplete Tests

    The final rule, in Sec. 40.67, picks up paragraphs from the NPRM 
that do not fit conveniently in other sections. The first provides that 
employee refusals to take certain actions (e.g., complete and sign Step 
2 of the form, provide breath) constitute a refusal to be tested. Such 
refusals, under the operating administration rules, have the same 
consequences as a test result of 0.04 or greater. The NPRM provision on 
which this paragraph is based was not the subject of comment. The 
second paragraph provides that if a test cannot be completed, or an 
event occurs that would invalidate the test, the BAT would, if 
practicable, run a retest. All seventeen comments on the subject 
favored this approach, and the Department is including it in the final 
rule.

Inability to Provide Sufficient Breath

    The NPRM proposed that if an employee were unable to provide enough 
breath for an adequate sample, the BAT would ask the employee to try 
again. If the same result occurred, then the employee would be referred 
to a doctor for a medical evaluation. If the doctor determined that the 
inability to provide breath was due, or probably due, to a medical 
condition, the failure to provide the sample would be excused. If not, 
it would be treated as a refusal.
    Four comments supported the NPRM provision. Three others thought 
that this situation was unlikely to arise, since only an employee who 
was seriously disabled, unconscious, or dead would be unable to provide 
the modest quantity of breath required to complete a test. We agree 
that this situation should not occur frequently, but we believe it is 
sensible to have a procedure in place to handle the occasional 
occurrence.
    Nine commenters suggested that, if the employee cannot provide 
sufficient breath, the employee should be required to provide a sample 
of a body fluid (e.g., blood, urine). Two comments urged employer 
discretion in these cases. Ten commenters said that there should be a 
medical evaluation in all cases where an employee cannot produce 
sufficient breath, though these commenters disagreed with each other 
about whether the employee should be held out of safety-sensitive 
functions pending the result of the evaluation.
    Under the final rule, the employer is required to direct the 
employee to be medically evaluated in ``shy lung'' cases. The final 
rule directs the employer to ensure that this evaluation occurs as soon 
as possible. Employers, under their own authority, could choose to 
``stand down'' an employee pending the result of a medical evaluation, 
but the rule does not require this step.
    In addition, the accompanying NPRM proposes that blood testing may 
be used in post-accident and reasonable suspicion testing when an EBT 
is not readily available. Since blood testing, and procedures for it, 
may become part of the rule for these purposes, the Department is 
responding to these comments by proposing blood testing as an option 
(regardless of the type of testing involved) when an employee cannot 
provide a sufficient breath sample. If the NPRM's proposal is made part 
of a final rule, the employer would have discretion concerning which 
alternative (blood alcohol testing or a medical evaluation) to select. 
Persons interested in this issue are asked to comment to the NPRM 
docket.

Invalid Tests

    The original NPRM listed nine ``fatal flaws'' that would invalidate 
breath tests. An invalid test is neither positive nor negative, and it 
has no consequences for an employee. The NPRM being published today 
proposes a similar list of fatal flaws for blood tests.
    The NPRM proposed that failure to observe the 15-minute minimum 
waiting period before the confirmation test would be a fatal flaw; 
going over the 20-minute maximum would not. Comments generally agreed 
with this approach, some noting that if exceeding a maximum waiting 
time were to be a fatal flaw, the outer limit should be 30 or 60 
minutes rather than 20. One commenter opposed making observance of the 
minimum a fatal flaw. The Department is retaining the NPRM provision on 
this point.
    The Department is changing the provision concerning air blanks to 
reflect the final rule's requirement of an air blank before only the 
confirmation test. Likewise, the NPRM provision making the device's 
failure to print out a result a fatal flaw has been changed to apply 
only to confirmation tests. The provision on disagreement between the 
printout and the machine display concerning sequential test numbers or 
alcohol concentration has been modified for the same reason. If the 
employee fails to sign Step 4 of the form, that is not a fatal flaw; 
the BAT's failure to note the employee's failure to sign that portion 
of the form would be a fatal flaw, however.
    The NPRM proposed that if an EBT fails an external calibration 
check, every test performed on the device since the last valid external 
calibration test would be invalidated. Ten commenters opposed this 
provision, pointing out that it would cause numerous problems for 
employers if they had to invalidate tests after the fact, and perhaps 
had to reverse personnel actions as well. Four commenters supported the 
proposed requirement. The Department is well aware that after-the-fact 
invalidations of tests can create serious problems for employers. The 
Department does not see a workable alternative, however. If a valid 
external calibration check was performed after test A, and an invalid 
external calibration test was performed after test K, all we know for 
certain is that the machine went out of kilter somewhere between tests 
B and K. We cannot say for certain that test B or C was valid, or 
assume that the error occurred only on test K. Since we cannot 
determine that these tests were valid, we must, in fairness to the 
employees involved, treat them as invalid. Tests with results of 0.02 
and above would be deemed invalid in this situation. This is surely 
incentive for employers to conduct frequent external calibration 
checks, particularly after positive tests.
    One commenter suggested additional fatal flaws, such as failure to 
use a clean mouthpiece, inadequate grounds for reasonable suspicion, 
etc. One commenter suggested that all flaws should be regarded as 
fatal. The Department believes that only certain serious problems in 
the process, that directly affect the integrity of the test or accuracy 
of the result, should automatically invalidate the test. Other errors, 
particularly in combination with one another, could form the basis for 
a determination that a test is invalid (i.e., the listed fatal flaws 
are not intended to be the only possible grounds for invalidation). The 
Office of Drug Enforcement and Program Compliance is charged with 
providing, on behalf of the Department, definitive guidance on issues 
concerning the invalidation of tests.

Availability of Testing Information

    The NPRM proposed provisions on alcohol test information 
availability parallel to the existing provisions on the availability of 
drug testing information, as the Department has interpreted them. 
Employers could release information to a third party only with the 
specific written consent of the employee, must keep confidential 
information secure, but may make the information available in certain 
litigation situations. Employers must make information available to DOT 
or, under some circumstances, to the National Transportation Safety 
Board (NTSB). Employers must also make information about an employee's 
test available to that employee.
    Seven commenters, most of whom were from the motor carrier 
industry, asked that employers be authorized or required to make 
testing information available to third parties without the employee's 
consent. In this industry, the commenters said, there was a high 
turnover rate. Employees move rapidly from employer to employer. In the 
absence of authorization or requirement for a former employer to 
provide testing information to a potential new employer, either the 
hiring process would be slowed or important information about positive 
tests in the employee's past would be unavailable to the new employer.
    In response, the Department points out that an employer may, 
without authorization from DOT, require an applicant, as a condition of 
employment, to give written consent to the disclosure of this 
information by a former employer. The Department is adding a sentence 
to this provision of the rule telling employers that they must provide 
the information when the employee consents to its transmission to a 
third party. However, in order to maintain the confidentiality of 
sensitive information, in which employees have a significant privacy 
interest, the Department will not authorize the transmission of this 
information among employers or potential employers without written 
employee consent.
    The Department emphasizes that the consent involved must be a 
specific written consent for information to be sent from one named 
party to another named party. Blanket consents (i.e., a consent for 
testing information to be sent to all present or future employers or 
members of a consortium) are not permitted. Each consent must pertain 
to one specific employer providing the information about a particular 
employee to another specific employer.
    Two commenters suggested that an employee should not have to pay 
for obtaining information in his or her own file concerning alcohol 
tests. The Department believes that this is a matter better left to 
employer-employee agreements. As the Department interprets this 
provision, employers may impose reasonable charges to cover the cost of 
retrieval, copying, and transmission of the records requested. The 
employer is also expected only to provide copies within its possession 
or control (including documents that may be maintained by a consortium 
or third-party provider that conducted testing for the employer).

Records Concerning BATs and EBTs

    The NPRM proposed that the employer maintain various records 
concerning EBTs and BATs for five years. One commenter suggested that 
consortia and third-party providers be authorized to keep the records 
instead of the employer. The Department agrees that this is reasonable, 
and the final rule requires the employer or its agent to maintain the 
records. The employer retains ultimate responsibility for producing the 
records, however. Two commenters suggested we reduce the record 
retention period to two years, while one commenter said that the 
recordkeeping requirements in the NPRM were not burdensome. Consistent 
with the OA rules, the final part 40 rule establishes a 5-year 
retention period for calibration records and a two-year retention 
period for other records.

Other Issues

    A number of commenters asked that we modify the definition of 
alcohol to include alcohols other than ethanol (e.g., methanol, 
isopropanol), in order to avoid loopholes in the program that would 
allow an employee to claim that his or her alcohol concentration 
reading was the result of ingesting a non-ethanol substance. The 
Department agrees that the definition should be broadened to avoid any 
potential problems with the use of non-ethanol alcohols, and the final 
rule includes a modified definition to this effect. This revised 
definition is consistent with that used by NHTSA in its model 
specifications for evidential EBTs. We have also added a companion 
definition of alcohol use, which emphasizes that any consumption of a 
preparation including alcohol (e.g., beverages, medicines) counts as 
alcohol use.
    A few commenters asked that, for convenience, we centralize all the 
definitions in part 40 in one section. We have done so, and all the 
definitions are now in Sec. 40.3.
    The NPRM preamble asked for suggestions on how to deal with 
situations in which an arbitrator overturns an employer's personnel 
action based on an alcohol test result. Employers had expressed concern 
about perceived conflicts between the arbitrator's decisions and DOT 
regulations, and several commenters echoed these concerns. The 
Department is not convinced, however, that this problem is either 
frequent enough or serious enough to warrant a mandate in the 
regulatory text. Such a mandate, because it could not anticipate all 
the nuances of the factual situations involved, might interfere with 
reasonable resolutions of particular disputes.
    However, it is clear that employers are obligated to comply with 
DOT safety regulations, which have the force and effect of law. As a 
matter of law, no decision by an employer, employee organization, or 
individual or group appointed by those or other parties, can have the 
effect of excusing noncompliance by an employer with a provision of a 
DOT safety regulation. If a violation of DOT rules has occurred, then 
the consequences prescribed by DOT rules must follow (e.g., the 
employee must be removed from performing a safety-sensitive function).
    In the NPRM preamble, the Department included a discussion of 
handling of perceived conflicts between part 40 and operating 
administration regulations, exemptions, and the obligations of 
consortia and third-party providers (57 FR 59410; December 15, 1992). 
This discussion applies to the implementation of the final part 40 as 
well. The relevant language is reprinted below:

    Although implementation of part 40 generally would be done 
through an operating administration, part 40 is an Office of the 
Secretary of Transportation (OST) regulation. As such, requests for 
exemption would be processed under 49 CFR part 5, an existing 
regulation covering requests for exemption from or amendment to all 
OST rules, rather than through separate operating administration 
exemption procedures. This would add an additional element of 
consistency. This approach is consistent with the existing part 40 
drug testing procedures, from which exemptions would also be granted 
under part 40. (See 54 FR 49863; December 1, 1989).
    The grant of an exemption under part 40 must be based on special 
or exceptional circumstances. It is not appropriate to carve out a 
generally applicable exception to a rule. Also, an exemption must be 
based on circumstances not contemplated as part of the rulemaking. 
The exemption process is not designed to revisit issues settled in 
the rulemaking process.
    Section 40.1 would also emphasize that other parties involved in 
the testing process--such as consortia, contractors, and agents--
``stand in the shoes'' of the employer. They are, therefore, subject 
to the same obligations and requirements as the employer. If an 
employer is required to do something, so is the consortium that is 
conducting testing for the employer. If the consortium fails to do 
something correctly, the employer is in noncompliance.

    Since, as noted above, part 40 is a regulation of the Office of the 
Secretary of Transportation, the source of definitive interpretations 
of the rule is the Office of the Secretary. Interpretations have been 
and will continue to be made in close coordination among the OAs, the 
Office of Drug Enforcement and Program Compliance (DEPC), and the 
Office of General Counsel.

Regulatory Analyses and Notices

    Because of substantial public interest and substantial impacts on a 
wide range of private and public sector organizations, the Department 
has determined that this rule--in conjunction with the operating 
administration alcohol and drug testing rules--is significant under 
Executive Order 12866. The rule has been reviewed under this Order. It 
is also significant under the Department's regulatory policies and 
procedures. The Department has prepared a regulatory evaluation for 
part 40, which we have included in the docket. The costs of the 
application of part 40 procedures to the programs of the various OAs 
are estimated in each of the OAs' regulatory evaluations for their drug 
and alcohol rules being published today.
    This rule, in conjunction with the operating administration drug 
and alcohol testing rules, is likely to have a significant economic 
impact on a substantial number of small entities. These impacts are 
assessed in the OAs' regulatory evaluations. The Federalism impacts of 
this rule are either minimal or required by statute; for these reasons, 
we have not prepared a Federalism assessment.
    This rule also contains collection of information requirements. The 
Department has submitted these requirements to the Office of Management 
and Budget for review and approval under the Paperwork Reduction Act 
(44 U.S.C. 350, et. seq.). Please see the Common Preamble on the status 
of Paperwork Reduction Act approvals.

List of Subjects in 49 CFR Part 40

    Drug testing, Reporting and recordkeeping requirements, Safety, 
Transportation.

    Issued This 25th day of January, 1994, at Washington, D.C.
Federico Pena,
Secretary of Transportation.
David R. Hinson,
Administrator, Federal Aviation Administration.
Rodney E. Slater,
Administrator, Federal Highway Administration.
Jolene M. Molitoris,
Administrator, Federal Railroad Administration.
Gordon J. Linton,
Administrator, Federal Transit Administration.
Ana Sol Gutierrez,
Acting Administrator, Research and Special Programs Administration.
Adm. J. William Kime,
Commandant, United States Coast Guard.

    For the reasons set forth in the preamble, the Department of 
Transportation amends Title 49, Code of Federal Regulations, part 40, 
as follows:

PART 40--PROCEDURES FOR TRANSPORTATION WORKPLACE DRUG AND ALCOHOL 
TESTING PROGRAMS

    1. The authority citation for Part 40 is revised to read as 
follows:

    Authority: 49 U.S.C. 102,301,322; 49 U.S.C. app. 1301nt., app. 
1434nt., app. 2717, app. 1618a.

    2. Secs. 40.1 through 40.19 are designated as subpart A and revised 
to read as follows:

Subpart A--General

40.1  Applicability.
40.3  Definitions.
40.5-40.19  [Reserved]

Subpart A--GENERAL


Sec. 40.1  Applicability.

    This part applies, through regulations that reference it issued by 
agencies of the Department of Transportation, to transportation 
employers, including self-employed individuals, required to conduct 
drug and/or alcohol testing programs by DOT agency regulations and to 
such transportation employers' officers, employees, agents and 
contractors (including, but not limited to, consortia). Employers are 
responsible for the compliance of their officers, employees, agents, 
consortia and/or contractors with the requirements of this part.


Sec. 40.3  Definitions.

    The following definitions apply to this part:
    Air blank. A reading by an EBT of ambient air containing no 
alcohol. (In EBTs using gas chromatography technology, a reading of the 
device's internal standard.)
    Alcohol. The intoxicating agent in beverage alcohol, ethyl alcohol 
or other low molecular weight alcohols including methyl or isopropyl 
alcohol.
    Alcohol concentration. The alcohol in a volume of breath expressed 
in terms of grams of alcohol per 210 liters of breath as indicated by a 
breath test under this part.
    Alcohol use. The consumption of any beverage, mixture or 
preparation, including any medication, containing alcohol.
    Aliquot. A portion of a specimen used for testing.
    Blind sample or blind performance test specimen. A urine specimen 
submitted to a laboratory for quality control testing purposes, with a 
fictitious identifier, so that the laboratory cannot distinguish it 
from employee specimens, and which is spiked with known quantities of 
specific drugs or which is blank, containing no drugs.
    Breath Alcohol Technician (BAT). An individual who instructs and 
assists individuals in the alcohol testing process and operates an EBT.
    Canceled or invalid test. In drug testing, a drug test that has 
been declared invalid by a Medical Review Officer. A canceled test is 
neither a positive nor a negative test. For purposes of this part, a 
sample that has been rejected for testing by a laboratory is treated 
the same as a canceled test. In alcohol testing, a test that is deemed 
to be invalid under Sec. 40.79. It is neither a positive nor a negative 
test.
    Chain of custody. Procedures to account for the integrity of each 
urine or blood specimen by tracking its handling and storage from point 
of specimen collection to final disposition of the specimen. With 
respect to drug testing, these procedures shall require that an 
appropriate drug testing custody form (see Sec. 40.23(a)) be used from 
time of collection to receipt by the laboratory and that upon receipt 
by the laboratory an appropriate laboratory chain of custody form(s) 
account(s) for the sample or sample aliquots within the laboratory.
    Collection container. A container into which the employee urinates 
to provide the urine sample used for a drug test.
    Collection site. A place designated by the employer where 
individuals present themselves for the purpose of providing a specimen 
of their urine to be analyzed for the presence of drugs.
    Collection site person. A person who instructs and assists 
individuals at a collection site and who receives and makes a screening 
examination of the urine specimen provided by those individuals.
    Confirmation (or confirmatory) test. In drug testing, a second 
analytical procedure to identify the presence of a specific drug or 
metabolite that is independent of the screening test and that uses a 
different technique and chemical principle from that of the screening 
test in order to ensure reliability and accuracy. (Gas chromatography/
mass spectrometry (GC/MS) is the only authorized confirmation method 
for cocaine, marijuana, opiates, amphetamines, and phencyclidine.) In 
alcohol testing, a second test, following a screening test with a 
result of 0.02 or greater, that provides quantitative data of alcohol 
concentration.
    DHHS. The Department of Health and Human Services or any designee 
of the Secretary, Department of Health and Human Services.
    DOT agency. An agency of the United States Department of 
Transportation administering regulations related to drug or alcohol 
testing, including the United States Coast Guard (for drug testing 
purposes only), the Federal Aviation Administration, the Federal 
Railroad Administration, the Federal Highway Administration, the 
Federal Transit Administration, the Research and Special Programs 
Administration, and the Office of the Secretary.
    Employee. An individual designated in a DOT agency regulation as 
subject to drug testing and/or alcohol testing. As used in this part 
``employee'' includes an applicant for employment. ``Employee'' and 
``individual'' or ``individual to be tested'' have the same meaning for 
purposes of this part.
    Employer. An entity employing one or more employees that is subject 
to DOT agency regulations requiring compliance with this part. As used 
in this part, employer includes an industry consortium or joint 
enterprise comprised of two or more employing entities.
    EBT (or evidential breath testing device). An EBT approved by the 
National Highway Traffic Safety Administration (NHTSA) for the 
evidential testing of breath and placed on NHTSA's ``Conforming 
Products List of Evidential Breath Measurement Devices'' (CPL).
    Medical Review Officer (MRO). A licensed physician (medical doctor 
or doctor of osteopathy) responsible for receiving laboratory results 
generated by an employer's drug testing program who has knowledge of 
substance abuse disorders and has appropriate medical training to 
interpret and evaluate an individual's confirmed positive test result 
together with his or her medical history and any other relevant 
biomedical information.
    Screening test (or initial test). In drug testing, an immunoassay 
screen to eliminate ``negative'' urine specimens from further analysis. 
In alcohol testing, an analytic procedure to determine whether an 
employee may have a prohibited concentration of alcohol in a breath 
specimen.
    Secretary. The Secretary of Transportation or the Secretary's 
designee.
    Shipping container. A container capable of being secured with a 
tamper-evident seal that is used for transfer of one or more urine 
specimen bottle(s) and associated documentation from the collection 
site to the laboratory.
    Specimen bottle. The bottle that, after being labeled and sealed 
according to the procedures in this part, is used to transmit a urine 
sample to the laboratory.


Secs. 40.5--40.19  [Reserved]

    2. Secs. 40.21 through 40.39 are designated subpart B.

Subpart B--Drug Testing

40.21  The drugs.
40.23  Preparation for testing.
40.25  Specimen collection procedures.
40.27  Laboratory personnel.
40.29  Laboratory analysis procedures.
40.31  Quality assurance and quality control.
40.33  Reporting and review of results.
40.35  Protection of employee records.
40.37  Individual access to test and laboratory certification 
results.
40.39  Use of DHHS--certified laboratories.

    Authority: 49 U.S.C. 102, 301, 322; 49 U.S.C. app. 1301nt., app. 
1434nt., app. 2717, app. 1618a.

    3. In Sec. 40.25, paragraph (f)(10) is revised to read as follows:


Sec. 40.25   Specimen collection procedures.

* * * * *
    (f) * * *
    (10) The collection site person shall instruct the employee to 
provide at least 45 ml of urine under the split sample method of 
collection or 30 ml of urine under the single sample method of 
collection.
    (i)(A) Employers with employees subject to drug testing only under 
the drug testing rules of the Research and Special Programs 
Administration and/or Coast Guard may use the ``split sample'' method 
of collection or may collect a single sample for those employees.
    (B) Employers with employees subject to drug testing under the drug 
testing rules of the Federal Highway Administration, Federal Railroad 
Administration, Federal Transit Administration, or Federal Aviation 
Administration shall use the ``split sample'' method of collection for 
those employees.
    (ii) Employers using the split sample method of collection shall 
follow the procedures in this paragraph (f)(10)(ii):
    (A) The donor shall urinate into a collection container or a 
specimen bottle capable of holding at least 60 ml.
    (B) If a collection container is used, the collection site person, 
in the presence of the donor, pours the urine into two specimen 
bottles. Thirty (30) ml shall be poured into one bottle, to be used as 
the primary specimen. At least 15 ml shall be poured into the other 
bottle, to be used as the split specimen.
    (C) If a single specimen bottle is used as a collection container, 
the collection site person shall pour 30 ml of urine from the specimen 
bottle into a second specimen bottle (to be used as the primary 
specimen) and retain the remainder (at least 15 ml) in the collection 
bottle (to be used as the split specimen).
    (D) Both bottles shall be shipped in a single shipping container, 
together with copies 1,2, and the split specimen copy of the chain of 
custody form, to the laboratory.
    (E) If the test result of the primary specimen is positive, the 
employee may request that the MRO direct that the split specimen be 
tested in a different DHHS-certified laboratory for presence of the 
drug(s) for which a positive result was obtained in the test of the 
primary specimen. The MRO shall honor such a request if it is made 
within 72 hours of the employee having been notified of a verified 
positive test result.
    (F) When the MRO informs the laboratory in writing that the 
employee has requested a test of the split specimen, the laboratory 
shall forward, to a different DHHS-approved laboratory, the split 
specimen bottle, with seal intact, a copy of the MRO request, and the 
split specimen copy of the chain of custody form with appropriate chain 
of custody entries.
    (G) The result of the test of the split specimen is transmitted by 
the second laboratory to the MRO.
    (H) Action required by DOT agency regulations as the result of a 
positive drug test (e.g., removal from performing a safety-sensitive 
function) is not stayed pending the result of the test of the split 
specimen.
    (I) If the result of the test of the split specimen fails to 
reconfirm the presence of the drug(s) or drug metabolite(s) found in 
the primary specimen, the MRO shall cancel the test, and report the 
cancellation and the reasons for it to the DOT, the employer, and the 
employee.
    (iii) Employers using the single sample collection method shall 
follow the procedures in paragraph:
    (A) The collector may choose to direct the employee to urinate 
either directly into a specimen bottle or into a separate collection 
container.
    (B) If a separate collection container is used, the collection site 
person shall pour at least 30 ml of the urine from the collection 
container into the specimen bottle in the presence of the employee.
    (iv) In either collection methodology, upon receiving the specimen 
from the individual, the collection site person shall determine if it 
has at least 30 milliliters of urine for the primary or single specimen 
bottle and, where the split specimen collection method is used, an 
additional 15 ml of urine for the split specimen bottle. If the 
individual is unable to provide such a quantity of urine, the 
collection site person shall instruct the individual to drink not more 
than 24 ounces of fluids and, after a period of up to two hours, again 
attempt to provide a complete sample using a fresh collection 
container. The original insufficient specimen shall be discarded. If 
the employee is still unable to provide an adequate specimen, the 
insufficient specimen shall be discarded, testing discontinued, and the 
employer so notified. The MRO shall refer the individual for a medical 
evaluation to develop pertinent information concerning whether the 
individual's inability to provide a specimen is genuine or constitutes 
a refusal to test. (In preemployment testing, if the employer does not 
wish to hire the individual, the MRO is not required to make such a 
referral.) Upon completion of the examination, the MRO shall report his 
or her conclusions to the employer in writing.
* * * * *
    4. In Sec. 40.29, paragraph (b)(2) is revised and paragraph (b)(3) 
is added, as follows:


Sec. 40.29   Laboratory analysis procedures.

* * * * *
    (b) * * * *
    (2) In situations where the employer uses the split sample 
collection method, the laboratory shall log in the split specimen, with 
the split specimen bottle seal remaining intact. The laboratory shall 
store this sample securely (see paragraph (c) of this section). If the 
result of the test of the primary specimen is negative, the laboratory 
may discard the split specimen. If the result of the test of the 
primary specimen is positive, the laboratory shall retain the split 
specimen in frozen storage for 60 days from the date on which the 
laboratory acquires it (see paragraph (h) of this section). Following 
the end of the 60-day period, if not informed by the MRO that the 
employee has requested a test of the split specimen, the laboratory may 
discard the split specimen.
    (3) When directed in writing by the MRO to forward the split 
specimen to another DHHS-certified laboratory for analysis, the second 
laboratory shall analyze the split specimen by GC/MS to reconfirm the 
presence of the drug(s) or drug metabolite(s) found in the primary 
specimen. Such GC/MS confirmation shall be conducted without regard to 
the cutoff levels of Sec. 40.29(f). The split specimen shall be 
retained in long-term storage for one year by the laboratory conducting 
the analysis of the split specimen (or longer if litigation concerning 
the test is pending).
* * * * *
    6. In Sec. 40.33 paragraphs (e), (f) and (g) are revised; paragraph 
(h) is redesignated as paragraphs (i), and a new paragraph (h) is 
added, as follows:


Sec. 40.33   Reporting and review of results.

* * * * *
    (e) In a situation in which the employer has used the single sample 
method of collection, the MRO shall notify each employee who has a 
confirmed positive test that the employee has 72 hours in which to 
request a reanalysis of the original specimen, if the test is verified 
positive. If requested to do so by the employee within 72 hours of the 
employee's having been informed of a verified positive test, the 
Medical Review Officer shall direct, in writing, a reanalysis of the 
original sample. The MRO may also direct, in writing, such a reanalysis 
if the MRO questions the accuracy or validity of any test result. Only 
the MRO may authorize such a reanalysis, and such a reanalysis may take 
place only at laboratories certified by DHHS. If the reanalysis fails 
to reconfirm the presence of the drug or drug metabolite, the MRO shall 
cancel the test and report the cancellation and the reasons for it to 
the DOT, the employer and the employee.
    (f) In situations in which the employer uses the split sample 
method of collection, the MRO shall notify each employee who has a 
confirmed positive test that the employee has 72 hours in which to 
request a test of the split specimen, if the test is verified as 
positive. If the employee requests an analysis of the split specimen 
within 72 hours of having been informed of a verified positive test, 
the MRO shall direct, in writing, the laboratory to provide the split 
specimen to another DHHS-certified laboratory for analysis. If the 
analysis of the split specimen fails to reconfirm the presence of the 
drug(s) or drug metabolite(s) found in the primary specimen, or if the 
split specimen is unavailable, inadequate for testing or untestable, 
the MRO shall cancel the test and report cancellation and the reasons 
for it to the DOT, the employer, and the employee.
    (g) If an employee has not contacted the MRO within 72 hours, as 
provided in paragraphs (e) and (f) of this section, the employee may 
present to the MRO information documenting that serious illness, 
injury, inability to contact the MRO, lack of actual notice of the 
verified positive test, or other circumstances unavoidably prevented 
the employee from timely contacting the MRO. If the MRO concludes that 
there is a legitimate explanation for the employee's failure to contact 
the MRO within 72 hours, the MRO shall direct that the reanalysis of 
the primary specimen or analysis of the split specimen, as applicable, 
be performed.
    (h) When the employer uses the split sample method of collection, 
the employee is not authorized to request a reanalysis of the primary 
specimen as provided in paragraph (e) of this section.
* * * * *
    7. A new subpart C is added to part 40, to read as follows:

Subpart C--Alcohol Testing

40.51  The breath alcohol technician.
40.53  Devices to be used for breath alcohol tests.
40.55  Quality assurance plans for EBTs.
40.57  Locations for breath alcohol testing.
40.59  The breath alcohol testing form and log book.
40.61  Preparation for breath alcohol testing.
40.63  Procedures for screening tests.
40.65  Procedures for confirmation tests.
40.67  Refusals to test and uncompleted tests.
40.69  Inability to provide an adequate amount of breath.
40.71  [Reserved]
40.73  [Reserved]
40.75  [Reserved]
40.77  [Reserved]
40.79  Invalid Tests.
40.81  Availability and disclosure of alcohol testing information 
about individual employees.
40.83  Maintenance and disclosure of records concerning EBTs and 
BATs.

Appendix A--The Breath Alcohol Testing Form

    Authority: 49 U.S.C. 102, 301, 322; 49 U.S.C. app. 1301nt., app. 
1434nt., app. 2717, app. 1618a.


Sec. 40.51  The breath alcohol technician.

    (a) The breath alcohol technician (BAT) shall be trained to 
proficiency in the operation of the EBT he or she is using and in the 
alcohol testing procedures of this part.
    (1) Proficiency shall be demonstrated by successful completion of a 
course of instruction which, at a minimum, provides training in the 
principles of EBT methodology, operation, and calibration checks; the 
fundamentals of breath analysis for alcohol content; and the procedures 
required in this part for obtaining a breath sample, and interpreting 
and recording EBT results.
    (2) Only courses of instruction for operation of EBTs that are 
equivalent to the Department of Transportation model course, as 
determined by the National Highway Traffic Safety Administration 
(NHTSA), may be used to train BATs to proficiency. On request, NHTSA 
will review a BAT instruction course for equivalency.
    (3) The course of instruction shall provide documentation that the 
BAT has demonstrated competence in the operation of the specific EBT(s) 
he/she will use.
    (4) Any BAT who will perform an external calibration check of an 
EBT shall be trained to proficiency in conducting the check on the 
particular model of EBT, to include practical experience and 
demonstrated competence in preparing the breath alcohol simulator or 
alcohol standard, and in maintenance and calibration of the EBT.
    (5) The BAT shall receive additional training, as needed, to ensure 
proficiency, concerning new or additional devices or changes in 
technology that he or she will use.
    (6) The employer or its agent shall establish documentation of the 
training and proficiency test of each BAT it uses to test employees, 
and maintain the documentation as provided in Sec. 40.83.
    (b) A BAT-qualified supervisor of an employee may conduct the 
alcohol test for that employee only if another BAT is unavailable to 
perform the test in a timely manner. A supervisor shall not serve as a 
BAT for the employee in any circumstance prohibited by a DOT operating 
administration regulation.
    (c) Law enforcement officers who have been certified by state or 
local governments to conduct breath alcohol testing are deemed to be 
qualified as BATs. In order for a test conducted by such an officer to 
be accepted under Department of Transportation alcohol testing 
requirements, the officer must have been certified by a state or local 
government to use the EBT that was used for the test.


Sec. 40.53  Devices to be used for breath alcohol tests.

    (a) For screening tests, employers shall use only EBTs. When the 
employer uses for a screening test an EBT that does not meet the 
requirements of paragraphs (b) (1) through (3) of this section, the 
employer shall use a log book in conjunction with the EBT (see 
Sec. 40.59(c)).
    (b) For confirmation tests, employers shall use EBTs that meet the 
following requirements:
    (1) EBTs shall have the capability of providing, independently or 
by direct link to a separate printer, a printed result in triplicate 
(or three consecutive identical copies) of each breath test and of the 
operations specified in paragraphs (b) (2) and (3) of this section.
    (2) EBTs shall be capable of assigning a unique and sequential 
number to each completed test, with the number capable of being read by 
the BAT and the employee before each test and being printed out on each 
copy of the result.
    (3) EBTs shall be capable of printing out, on each copy of the 
result, the manufacturer's name for the device, the device's serial 
number, and the time of the test.
    (4) EBTs shall be able to distinguish alcohol from acetone at the 
0.02 alcohol concentration level.
    (5) EBTs shall be capable of the following operations:
    (i) Testing an air blank prior to each collection of breath; and
    (ii) Performing an external calibration check.


Sec. 40.55  Quality assurance plans for EBTs.

    (a) In order to be used in either screening or confirmation alcohol 
testing subject to this part, an EBT shall have a quality assurance 
plan (QAP) developed by the manufacturer.
    (1) The plan shall designate the method or methods to be used to 
perform external calibration checks of the device, using only 
calibration devices on the NHTSA ``Conforming Products List of 
Calibrating Units for Breath Alcohol Tests.''
    (2) The plan shall specify the minimum intervals for performing 
external calibration checks of the device. Intervals shall be specified 
for different frequencies of use, environmental conditions (e.g., 
temperature, altitude, humidity), and contexts of operation (e.g., 
stationary or mobile use).
    (3) The plan shall specify the tolerances on an external 
calibration check within which the EBT is regarded to be in proper 
calibration.
    (4) The plan shall specify inspection, maintenance, and calibration 
requirements and intervals for the device.
    (5) For a plan to be regarded as valid, the manufacturer shall have 
submitted the plan to NHTSA for review and have received NHTSA approval 
of the plan.
    (b) The employer shall comply with the NHTSA-approved quality 
assurance plan for each EBT it uses for alcohol screening or 
confirmation testing subject to this part.
    (1) The employer shall ensure that external calibration checks of 
each EBT are performed as provided in the QAP.
    (2) The employer shall take an EBT out of service if any external 
calibration check results in a reading outside the tolerances for the 
EBT set forth in the QAP. The EBT shall not again be used for alcohol 
testing under this part until it has been serviced and has had an 
external calibration check resulting in a reading within the tolerances 
for the EBT.
    (3) The employer shall ensure that inspection, maintenance, and 
calibration of each EBT are performed by the manufacturer or a 
maintenance representative certified by the device's manufacturer or a 
state health agency or other appropriate state agency. The employer 
shall also ensure that each BAT or other individual who performs an 
external calibration check of an EBT used for alcohol testing subject 
to this part has demonstrated proficiency in conducting such a check of 
the model of EBT in question.
    (4) The employer shall maintain records of the external calibration 
checks of EBTs as provided in Sec. 40.83.
    (c) When the employer is not using the EBT at an alcohol testing 
site, the employer shall store the EBT in a secure space.


Sec. 40.57  Locations for breath alcohol testing.

    (a) Each employer shall conduct alcohol testing in a location that 
affords visual and aural privacy to the individual being tested, 
sufficient to prevent unauthorized persons from seeing or hearing test 
results. All necessary equipment, personnel, and materials for breath 
testing shall be provided at the location where testing is conducted.
    (b) An employer may use a mobile collection facility (e.g., a van 
equipped for alcohol testing) that meets the requirements of paragraph 
(a) of this section.
    (c) No unauthorized persons shall be permitted access to the 
testing location when the EBT remains unsecured or, in order to prevent 
such persons from seeing or hearing a testing result, at any time when 
testing is being conducted.
    (d) In unusual circumstances (e.g., when it is essential to conduct 
a test outdoors at the scene of an accident), a test may be conducted 
at a location that does not fully meet the requirements of paragraph 
(a) of this section. In such a case, the employer or BAT shall provide 
visual and aural privacy to the employee to the greatest extent 
practicable.
    (e) The BAT shall supervise only one employee's use of the EBT at a 
time. The BAT shall not leave the alcohol testing location while the 
testing procedure for a given employee (see Secs. 40.61 through 40.65) 
is in progress.


Sec. 40.59  The breath alcohol testing form and log book.

    (a) Each employer shall use the breath alcohol testing form 
prescribed under this part. The form is found in appendix A to this 
subpart. Employers may not modify or revise this form, except that a 
form directly generated by an EBT may omit the space for affixing a 
separate printed result to the form.
    (b) The form shall provide triplicate (or three consecutive 
identical) copies. Copy 1 (white) shall be retained by the BAT. Copy 2 
(green) shall be provided to the employee. Copy 3 (blue) shall be 
transmitted to the employer. Except for a form generated by an EBT, the 
form shall be 8\1/2\ by 11 inches in size.
    (c) A log book shall be used in conjunction with any EBT used for 
screening tests that does not meet the requirements of Sec. 40.53(b) 
(1) through (3). There shall be a log book for each such device, that 
is not used in conjunction with any other device and that is used to 
record every test conducted on the device. The log book shall include 
columns for the test number, date of the test, name of the BAT, 
location of the test, quantified test result, and initials of the 
employee taking each test.


Sec. 40.61  Preparation for breath alcohol testing.

    (a) When the employee enters the alcohol testing location, the BAT 
will require him or her to provide positive identification (e.g., 
through use of a photo I.D. card or identification by an employer 
representative). On request by the employee, the BAT shall provide 
positive identification to the employee.
    (b) The BAT shall explain the testing procedure to the employee.


Sec. 40.63  Procedures for screening tests.

    (a) The BAT shall complete Step 1 on the Breath Alcohol Testing 
Form. The employee shall then complete Step 2 on the form, signing the 
certification. Refusal by the employee to sign this certification shall 
be regarded as a refusal to take the test.
    (b) An individually-sealed mouthpiece shall be opened in view of 
the employee and BAT and attached to the EBT in accordance with the 
manufacturer's instructions.
    (c) The BAT shall instruct the employee to blow forcefully into the 
mouthpiece for at least 6 seconds or until the EBT indicates that an 
adequate amount of breath has been obtained.
    (d)(1) If the EBT does not meet the requirements of 
Sec. 40.53(b)(1) through (3), the BAT and the employee shall take the 
following steps:
    (i) Show the employee the result displayed on the EBT. The BAT 
shall record the displayed result, test number, testing device, serial 
number of the testing device, time and quantified result in Step 3 of 
the form.
    (ii) Record the test number, date of the test, name of the BAT, 
location, and quantified test result in the log book. The employee 
shall initial the log book entry.
    (2) If the EBT provides a printed result, but does not print the 
results directly onto the form, the BAT shall show the employee the 
result displayed on the EBT. The BAT shall then affix the test result 
printout to the breath alcohol test form in the designated space, using 
a method that will provide clear evidence of removal (e.g., tamper-
evident tape).
    (3) If the EBT prints the test results directly onto the form, the 
BAT shall show the employee the result displayed on the EBT.
    (e)(1) In any case in which the result of the screening test is a 
breath alcohol concentration of less than 0.02, the BAT shall date the 
form and sign the certification in Step 3 of the form. The employee 
shall sign the certification and fill in the date in Step 4 of the 
form.
    (2) If the employee does not sign the certification in Step 4 of 
the form or does not initial the log book entry for a test, it shall 
not be considered a refusal to be tested. In this event, the BAT shall 
note the employee's failure to sign or initial in the ``Remarks'' 
section of the form.
    (3) If a test result printed by the EBT (see paragraph (d)(2) or 
(d)(3) of this section) does not match the displayed result, the BAT 
shall note the disparity in the remarks section. Both the employee and 
the BAT shall initial or sign the notation. In accordance with 
Sec. 40.79, the test is invalid and the employer and employee shall be 
so advised.
    (4) No further testing is authorized. The BAT shall transmit the 
result of less than 0.02 to the employer in a confidential manner, and 
the employer shall receive and store the information so as to ensure 
that confidentiality is maintained as required by Sec. 40.81.
    (f) If the result of the screening test is an alcohol concentration 
of 0.02 or greater, a confirmation test shall be performed as provided 
in Sec. 40.65.
    (g) If the confirmation test will be conducted by a different BAT, 
the BAT who conducts the screening test shall complete and sign the 
form and log book entry. The BAT will provide the employee with Copy 2 
of the form.


Sec. 40.65  Procedures for confirmation tests.

    (a) If a BAT other than the one who conducted the screening test is 
conducting the confirmation test, the new BAT shall follow the 
procedures of Sec. 40.61.
    (b) The BAT shall instruct the employee not to eat, drink, put any 
object or substance in his or her mouth, and, to the extent possible, 
not belch during a waiting period before the confirmation test. This 
time period begins with the completion of the screening test, and shall 
not be less than 15 minutes. The confirmation test shall be conducted 
within 20 minutes of the completion of the screening test. The BAT 
shall explain to the employee the reason for this requirement (i.e., to 
prevent any accumulation of mouth alcohol leading to an artificially 
high reading) and the fact that it is for the employee's benefit. The 
BAT shall also explain that the test will be conducted at the end of 
the waiting period, even if the employee has disregarded the 
instruction. If the BAT becomes aware that the employee has not 
complied with this instruction, the BAT shall so note in the 
``Remarks'' section of the form.
    (c) (1) If a BAT other than the one who conducted the screening 
test is conducting the confirmation test, the new BAT shall initiate a 
new Breath Alcohol Testing form. The BAT shall complete Step 1 on the 
form. The employee shall then complete Step 2 on the form, signing the 
certification. Refusal by the employee to sign this certification shall 
be regarded as a refusal to take the test. The BAT shall note in the 
``Remarks'' section of the form that a different BAT conducted the 
screening test.
    (2) In all cases, the procedures of Sec. 40.63 (a), (b), and (c) 
shall be followed. A new mouthpiece shall be used for the confirmation 
test.
    (d) Before the confirmation test is administered for each employee, 
the BAT shall ensure that the EBT registers 0.00 on an air blank. If 
the reading is greater than 0.00, the BAT shall conduct one more air 
blank. If the reading is greater than 0.00, testing shall not proceed 
using that instrument. However, testing may proceed on another 
instrument.
    (e) Any EBT taken out of service because of failure to perform an 
air blank accurately shall not be used for testing until a check of 
external calibration is conducted and the EBT is found to be within 
tolerance limits.
    (f) In the event that the screening and confirmation test results 
are not identical, the confirmation test result is deemed to be the 
final result upon which any action under operating administration rules 
shall be based.
    (g) (1) If the EBT provides a printed result, but does not print 
the results directly onto the form, the BAT shall show the employee the 
result displayed on the EBT. The BAT shall then affix the test result 
printout to the breath alcohol test form in the designated space, using 
a method that will provide clear evidence of removal (e.g., tamper-
evident tape).
    (2) If the EBT prints the test results directly onto the form, the 
BAT shall show the employee the result displayed on the EBT.
    (h) (1) Following the completion of the test, the BAT shall date 
the form and sign the certification in Step 3 of the form. The employee 
shall sign the certification and fill in the date in Step 4 of the 
form.
    (2) If the employee does not sign the certification in Step 4 of 
the form or does not initial the log book entry for a test, it shall 
not be considered a refusal to be tested. In this event, the BAT shall 
note the employee's failure to sign or initial in the ``Remarks'' 
section of the form.
    (3) If a test result printed by the EBT (see paragraph (g)(1) or 
(g)(2) of this section) does not match the displayed result, the BAT 
shall note the disparity in the remarks section. Both the employee and 
the BAT shall initial or sign the notation. In accordance with 
Sec. 40.79, the test is invalid and the employer and employee shall be 
so advised.
    (4) The BAT shall conduct an air blank. If the reading is greater 
than 0.00, the test is invalid.
    (i) The BAT shall transmit all results to the employer in a 
confidential manner.
    (1) Each employer shall designate one or more employer 
representatives for the purpose of receiving and handling alcohol 
testing results in a confidential manner. All communications by BATs to 
the employer concerning the alcohol testing results of employees shall 
be to a designated employer representative.
    (2) Such transmission may be in writing, in person or by telephone 
or electronic means, but the BAT shall ensure immediate transmission to 
the employer of results that require the employer to prevent the 
employee from performing a safety-sensitive function.
    (3) If the initial transmission is not in writing (e.g., by 
telephone), the employer shall establish a mechanism to verify the 
identity of the BAT providing the information.
    (4) If the initial transmission is not in writing, the BAT shall 
follow the initial transmission by providing to the employer the 
employer's copy of the breath alcohol testing form. The employer shall 
store the information so as to ensure that confidentiality is 
maintained as required by Sec. 40.81.


Sec. 40.67  Refusals to test and uncompleted tests.

    (a) Refusal by an employee to complete and sign the breath alcohol 
testing form (Step 2), to provide breath, to provide an adequate amount 
of breath, or otherwise to cooperate with the testing process in a way 
that prevents the completion of the test, shall be noted by the BAT in 
the remarks section of the form. The testing process shall be 
terminated and the BAT shall immediately notify the employer.
    (b) If a screening or confirmation test cannot be completed, or if 
an event occurs that would invalidate the test, the BAT shall, if 
practicable, begin a new screening or confirmation test, as applicable, 
using a new breath alcohol testing form with a new sequential test 
number (in the case of a screening test conducted on an EBT that meets 
the requirements of Sec. 40.53(b) or in the case of a confirmation 
test).


Sec. 40.69  Inability to provide an adequate amount of breath.

    (a) This section sets forth procedures to be followed in any case 
in which an employee is unable, or alleges that he or she is unable, to 
provide an amount of breath sufficient to permit a valid breath test 
because of a medical condition.
    (b) The BAT shall again instruct the employee to attempt to provide 
an adequate amount of breath. If the employee refuses to make the 
attempt, the BAT shall immediately inform the employer.
    (c) If the employee attempts and fails to provide an adequate 
amount of breath, the BAT shall so note in the ``Remarks'' section of 
the breath alcohol testing form and immediately inform the employer.
    (d) If the employee attempts and fails to provide an adequate 
amount of breath, the employer shall proceed as follows:
    (1) [Reserved]
    (2) The employer shall direct the employee to obtain, as soon as 
practical after the attempted provision of breath, an evaluation from a 
licensed physician who is acceptable to the employer concerning the 
employee's medical ability to provide an adequate amount of breath.
    (i) If the physician determines, in his or her reasonable medical 
judgment, that a medical condition has, or with a high degree of 
probability, could have, precluded the employee from providing an 
adequate amount of breath, the employee's failure to provide an 
adequate amount of breath shall not be deemed a refusal to take a test. 
The physician shall provide to the employer a written statement of the 
basis for his or her conclusion.
    (ii) If the licensed physician, in his or her reasonable medical 
judgment, is unable to make the determination set forth in paragraph 
(d)(2)(i) of this section the employee's failure to provide an adequate 
amount of breath shall be regarded as a refusal to take a test. The 
licensed physician shall provide a written statement of the basis for 
his or her conclusion to the employer.


Secs. 40.71-40.77  [Reserved]


Sec. 40.79  Invalid tests.

    (a) A breath alcohol test shall be invalid under the following 
circumstances:
    (1) The next external calibration check of an EBT produces a result 
that differs by more than the tolerance stated in the QAP from the 
known value of the test standard. In this event, every test result of 
0.02 or above obtained on the device since the last valid external 
calibration check shall be invalid;
    (2) The BAT does not observe the minimum 15-minute waiting period 
prior to the confirmation test, as provided in Sec. 40.65(b);
    (3) The BAT does not perform an air blank of the EBT before a 
confirmation test, or an air blank does not result in a reading of 0.00 
prior to or after the administration of the test, as provided in 
Sec. 40.65;
    (4) The BAT does not sign the form as required by Secs. 40.63 and 
40.65;
    (5) The BAT has failed to note on the remarks section of the form 
that the employee has failed or refused to sign the form following the 
recording or printing on or attachment to the form of the test result;
    (6) An EBT fails to print a confirmation test result; or
    (7) On a confirmation test and, where applicable, on a screening 
test, the sequential test number or alcohol concentration displayed on 
the EBT is not the same as the sequential test number or alcohol 
concentration on the printed result.
    (b) [Reserved]


Sec. 40.81  Availability and disclosure of alcohol testing information 
about individual employees.

    (a) Employers shall maintain records in a secure manner, so that 
disclosure of information to unauthorized persons does not occur.
    (b) Except as required by law or expressly authorized or required 
in this section, no employer shall release covered employee information 
that is contained in the records required to be maintained by this part 
or by DOT agency alcohol misuse rules.
    (c) An employee subject to testing is entitled, upon written 
request, to obtain copies of any records pertaining to the employee's 
use of alcohol, including any records pertaining to his or her alcohol 
tests. The employer shall promptly provide the records requested by the 
employee. Access to an employee's records shall not be contingent upon 
payment for records other than those specifically requested.
    (d) Each employer shall permit access to all facilities utilized in 
complying with the requirements of this part and DOT agency alcohol 
misuse rules to the Secretary of Transportation, any DOT agency with 
regulatory authority over the employer, or a state agency with 
regulatory authority over the employer (as authorized by DOT agency 
regulations).
    (e) When requested by the Secretary of Transportation, any DOT 
agency with regulatory authority over the employer, or a state agency 
with regulatory authority over the employer (as authorized by DOT 
agency regulations), each employer shall make available copies of all 
results for employer alcohol testing conducted under the requirements 
of this part and any other information pertaining to the employer's 
alcohol misuse prevention program. The information shall include name-
specific alcohol test results, records and reports.
    (f) When requested by the National Transportation Safety Board as 
part of an accident investigation, an employer shall disclose 
information related to the employer's administration of any post-
accident alcohol tests administered following the accident under 
investigation.
    (g) An employer shall make records available to a subsequent 
employer upon receipt of a written request from a covered employee. 
Disclosure by the subsequent employer is permitted only as expressly 
authorized by the terms of the employee's written request.
    (h) An employer may disclose information required to be maintained 
under this part pertaining to a covered employee to that employee or to 
the decisionmaker in a lawsuit, grievance, or other proceeding 
initiated by or on behalf of the individual, and arising from the 
results of an alcohol test administered under the requirements of this 
part, or from the employer's determination that the employee engaged in 
conduct prohibited by a DOT agency alcohol misuse regulation 
(including, but not limited to, a worker's compensation, unemployment 
compensation, or other proceeding relating to a benefit sought by the 
employee).
    (i) An employer shall release information regarding a covered 
employee's records as directed by the specific, written consent of the 
employee authorizing release of the information to an identified 
person. Release of such information is permitted only in accordance 
with the terms of the employee's consent.


Sec. 40.83  Maintenance and disclosure of records concerning EBTs and 
BATs.

    (a) Each employer or its agent shall maintain the following records 
for two years:
    (1) Records of the inspection and maintenance of each EBT used in 
employee testing;
    (2) Documentation of the employer's compliance with the QAP for 
each EBT it uses for alcohol testing under this part;
    (3) Records of the training and proficiency testing of each BAT 
used in employee testing;
    (4) The log books required by Sec. 40.59(c).
    (b) Each employer or its agent shall maintain for five years 
records pertaining to the calibration of each EBT used in alcohol 
testing under this part, including records of the results of external 
calibration checks.
    (c) Records required to be maintained by this section shall be 
disclosed on the same basis as provided in Sec. 40.81.

Appendix A to Subpart C of Part 40--The Breath Alcohol Testing Form

BILLING CODE 4910-62-U

TR15FE94.014


TR15FE94.015


TR15FE94.016


TR15FE94.017


TR15FE94.018


TR15FE94.019

[FR Doc. 94-2030 Filed 2-3-94; 1:00 pm]
BILLING CODE 4910-62-C