SUMMARY: FRA issues a final rule to conform its regulations on control of alcohol and drug use in railroad operations to the requirements of the Omnibus Transportation Employee Testing Act of 1991 (Act). Among other changes, FRA now requires pre-employment and random alcohol testing of safety-sensitive employees, and makes the reasonable suspicion component of for cause testing mandatory for both alcohol and drugs. FRA also amends its procedures to incorporate split sample testing and to incorporate the departmental alcohol testing procedures published elsewhere in today's Federal Register. EFFECTIVE DATES: This final rule is effective January 1, 1995, except that the amendment to Sec. 219.707 and Sec. 219.708 are effective August 15, 1994. ADDRESSES: Any petition for reconsideration should be submitted in triplicate to the Docket Clerk, Docket No. RSOR-6, Office of the Chief Counsel, Federal Railroad Administration, 400 7th Street SW., room 8201, Washington, DC 20590. FOR FURTHER INFORMATION CONTACT: Walter C. Rockey, Executive Assistant to the Associate Administrator for Safety (RRS-3), Office of Safety, FRA, Washington, DC 20590 (Telephone: (202) 366-0897) or Patricia V. Sun, Trial Attorney (RCC-30), Office of Chief Counsel, FRA, Washington, DC 20590 (Telephone: (202) 366-4002). SUPPLEMENTARY INFORMATION: Background On December 15, 1992, FRA published a notice proposing to amend its regulations on alcohol and drug misuse (49 CFR part 219) in response to the testing requirements mandated by the Omnibus Transportation Employee Testing Act of 1991 (57 FR 59588). At hearings in Washington, DC, Chicago, and San Francisco, FRA heard testimony from over 20 parties, including the major industry trade associations (the American Association of Railroads (AAR) and the American Short Line Railroad Association (ASLRA)) and labor organizations (the Brotherhood of Locomotive Engineers (BLE), the Brotherhood of Railroad Signalmen (BRS), and the Railway Labor Executives' Association (RLEA)). FRA also received oral or written comment from the American Public Transit Association (APTA), the Transportation Trades Department (TTD) of the AFL-CIO, the National Transportation Safety Board (NTSB), individual freight and commuter railroads, service providers, and state and local governments, among others. FRA has reconsidered some of its proposals in light of the comments received. First, the effective date for this rule is January 1, 1995, to ensure that EBTs that meet part 40 specifications will be widely available and to allow for implementation of quality control systems. This will also allow railroads time to purchase evidential breath testing devices (EBTs) and phase in other part 40 requirements such as breath alcohol technician training. To ensure a smooth transition, existing provisions will remain in effect and voluntary compliance before the effective date will not be allowed. (However, urine split sample drug testing is effective beginning on August 15, 1994. See part 40 and Secs. 219.707 and 219.708 of this final rule. Second, FRA permits screening tests to be conducted on preliminary breath testing devices (PBTs) found on the Conforming Products List (CPL) of the National Highway Traffic Safety Administration (NHTSA). EBTs that currently qualify for the CPL but do not meet the specifications listed in DOT's amended ``Procedures for Transportation Workplace Drug and Alcohol Testing Programs'' (49 CFR Part 40 (Part 40), published elsewhere in today's Federal Register), may be used as PBTs. As discussed in part 40, NHTSA will develop model specifications for a new CPL to qualify other alcohol testing devices (such as saliva tests and breath tubes) as PBTs. Procedures for these additional PBTs will be addressed in future rulemakings. Third, in a separate notice issued by the Office of the Secretary (OST), the Department proposes to allow blood testing for screening and confirmation tests in the case of reasonable suspicion or after qualifying accidents (for modes other than rail). Combined, the use of PBTs and blood would be an alternative testing methodology that could be used in remote locations or unusual circumstances when EBTs are inaccessible. Allowing a blood test option for for cause testing would substantially reduce the costs of alcohol testing, since railroads would not have to ensure system-wide availability of breath alcohol technicians (BATs) and EBTs for unplanned testing events. Railroads would then be free to plan the deployment or contracting of BATs and EBTs for the remaining types of testing (pre-employment, return to service, follow-up, and random) which are all, of course, scheduled by the railroads. Since FRA has repealed its existing breath and blood testing procedures (Sec. 219.104(e) and former Sec. 219.303), for cause testing will be conducted exclusively under today's amended part 40 procedures, which contain new breath testing safeguards. For now, FRA allows only breath to be used in for cause testing. As mentioned above, however, the Department is proposing part 40 blood testing procedures (see the DOT notice titled ``Procedures for Transportation Workplace Drug and Alcohol Testing Programs,'' 49 CFR part 40, published elsewhere in today's Federal Register). If the Department adopts this proposal, and allows for cause blood alcohol testing, FRA will allow railroads to conduct for cause blood alcohol testing under part 40 procedures. Also, FRA moves the authority for optional post-accident breath testing formerly contained in Sec. 219.303 from for cause testing to post-accident testing. Although urine and blood samples must still be collected and sent to FRA's post-accident laboratory for full toxicological testing, the EBT breath testing option allows railroads to obtain immediate test results. (Post-accident breath testing must be conducted in accordance with part 40, however, since FRA will not retain Sec. 219.303's procedures.) This relocation means, in effect, that the categorical standards used in subpart C will govern, rather than the reasonable cause criteria of subpart D. Additional discussion of alcohol testing methodology can be found below and in a separate departmental Final Rule, ``Procedures for Transportation Workplace Testing Programs'' (49 CFR part 40), also published elsewhere in today's Federal Register. Finally, for random alcohol testing, FRA will introduce performance-based testing by industry. Performance-based testing was widely supported by commenters. As proposed, FRA will use a graduated submission and implementation schedule similar to the one used to phase-in random drug testing. Railroads will initially be required to conduct random alcohol testing at a 25 percent rate. Performance-based testing, as determined by the Administrator, will begin a year after industry-wide implementation of random alcohol testing has occurred. (In a separate departmental NPRM in today's Federal Register, FRA also proposes to implement performance-based testing for random drug testing. See ``Random Drug Testing Program.'') The section-by-section analysis discusses these and other amendments contained in the final rule. (Editorial changes and several proposed amendments that did not receive comment are adopted without further discussion.) Interested parties should also review the departmental preamble (common preamble) published by the Office of the Secretary of Transportation elsewhere in today's Federal Register, which is incorporated herein by reference. (The Federal Aviation Administration (FAA), Federal Highway Administration (FHWA), Federal Transit Administration (FTA) and Research and Special Projects Administration (RSPA) are also publishing alcohol testing rules in today's Federal Register.) Although the common preamble discusses issues pertaining to all of the modal rules, some of its discussion is not directly applicable to FRA's rule. The basis for any differences can be found in the section-by-section analysis. For convenience and ease of reference, the entire rule text as amended is republished. Multi-modal Coverage For railroads, multi-modal coverage primarily affects those employees (mainly signal maintainers) who both perform covered service and hold Commercial Driver's Licenses (CDLs). These employees are subject to FRA and FHWA regulations. CDL holders who do not perform covered service (such as maintenance of way and shop employees) are subject only to FHWA's regulations. The ASLRA and rail labor wanted all railroad employees to be covered by FRA only. The AAR, on the other hand, recommended that FHWA regulations determine coverage for railroad commercial vehicle drivers, although testing would be conducted in accordance with FRA's regulations. Estimating that approximately 1,000 rail drivers per major railroad are subject to dual coverage, the AAR expressed concern about how to resolve differences between FRA and FHWA regulations and pointed out that this group of employees could be subject to a different random testing rate if performance-based rates are set by industry. While sympathetic to these industry concerns, FRA and DOT believe that modal coverage, like reportability (see Annual Reporting Requirements; Amendments to Alcohol/Drug Regulations, 58 FR 68232), must be determined by employee function to be logically consistent. As guidance, FRA's policy for CDL holders who perform covered service (and other employees subject to dual coverage) is as follows: For pre-employment and random testing, an employee is covered by whichever operating administration (OA) covers more than 50% of that employee's function. For post-accident and reasonable suspicion testing, however, coverage is determined by the function the employee was performing at the time of the accident or incident. Finally, for return to service and follow-up testing, the employee is covered by the same OA to which the initial positive was reported. For example, a signalman who holds a CDL but performs less than 50% of his time driving is covered by FRA for pre-employment and random testing. If that signalman were to have a reportable accident while driving a commercial motor vehicle, however, the post-accident test and any return to service tests and follow-up tests would be governed by FHWA. On the other hand, if in another instance a supervisor determined that the same signalman was impaired by a controlled substance while performing covered service under FRA's regulations, the reasonable suspicion test result and any post-positive return to service and follow-up test results would be governed by FRA. Under the management information system (MIS), test results should be reported to whichever OA governed the particular type of test. Each OA will calculate the violation rate for its industry, based on reported random test results. It is possible, therefore, that CDL holders covered by FHWA regulations could be tested at a different rate from those under FRA regulations. See the common preamble for further discussion of this issue. Section by Section Analysis Subpart A--General Sec. 219.3 Application. Paragraph (b)(2) FRA continues to exempt small railroads from subparts D, E, F and G. Although they are exempt from mandatory reasonable suspicion testing, small railroads must enforce the prohibitions contained in Secs. 219.101 and 219.102 (presumably through the industry's longstanding Rule G and their own for cause testing programs). Additionally, even though small railroads are exempted from the employee assistance requirements of subpart E, they must provide information on substance abuse services to their covered employees, as required by Sec. 219.23. Paragraph (b)(3) As mentioned above, FRA recently published a final rule implementing the drug testing portion of the management information system, a new reporting system for alcohol and drug program information that replaces the data that railroads currently submit in their annual reports under Sec. 217.13(d). For reasons discussed in the drug MIS rule, FRA exempts railroads with fewer than 400,000 manhours from all MIS reporting requirements (including the alcohol program data elements added in today's final rule). Foreign Application Foreign railroads have been subject to portions of FRA's regulations on the control of alcohol and drug use (49 CFR part 219) since February 10, 1986. In a Notice of Termination of Rulemaking Proceedings published elsewhere in today's Federal Register, FRA withdraws its advance notice of proposed rulemaking (ANPRM) on application of the Act's new requirements to foreign railroads operating within the United States (57 FR 59605). In lieu of a separate rulemaking on this issue, FRA revises Sec. 219.3(c) to continue and make permanent its current level of application of alcohol and drug testing to foreign railroads operating within the United States. As applied, FRA's current approach affects only Canadian employees, since Mexican employees do not operate in U.S. territory. FRA does not seek extra-territorial application of its regulations. A covered service employee whose primary place of service or point of departure (``home terminal'') for rail transportation services is located outside the U.S. continues to be subject to limited exceptions in coverage, since the primary terms and conditions of his or her employment were established under foreign law. Thus, the employee is subject to FRA's prohibitions and return to service conditions, as well as post-accident and for cause testing, only when operating in U.S. territory. (Pursuant to the Act, the employee is subject to mandatory reasonable suspicion alcohol and drug testing while on U.S. soil, although the accident/ incident and rules violation components of FRA's for cause testing program remain discretionary.) Employee assistance policies (subpart E), pre-employment testing (subpart F) and random testing (subpart G) remain subject to the law of the country where the employee is based (as determined by the employee's home terminal or reporting point). A U.S.-based covered service employee of a foreign railroad continues to be subject to all components of FRA's program, as amended. For both U.S.-based and foreign based covered service employees, all testing pursuant to part 219 must be conducted under the procedures set forth in part 219 and part 40, as amended in today's Federal Register. Sec. 219.5 Definitions. FRA revises its definition of alcohol to conform to the definition published by NHTSA in its September 17, 1993 notice on ``Highway Safety Programs; Model Specifications for Devices to Measure Breath Alcohol'' (58 FR 48705), which proposed to revise model specifications for EBTs. As discussed in both part 40 and the common preamble, FRA now requires employers to look to NHTSA's CPL for both PBTs and EBTs. Commenters were split on FRA's proposal to expand its definition of covered employees to include first line supervisors in random alcohol and drug testing programs. Most industry commenters found this proposal unjustified, since supervisors as a group have no record of substance abuse directly affecting railroad safety and are already subject to post-accident and for cause testing if they perform covered service. The ASLRA did not want supervisors to count as covered employees for purposes of the small railroad exemption (under FRA's proposal, supervisors would not have been considered Hours of Service Act employees towards the 15 employee threshold). On the other hand, the Long Island Railroad wanted testing expanded to include every supervisor in the chain of command, and Amtrak already uses its own authority to test its first two levels of supervisors (although Amtrak was concerned about preemption by FRA). Rail labor has always supported supervisory testing, for both equity and safety reasons. On balance, FRA has decided to drop this proposal since there is no statutory mandate to include supervisors in testing. Moreover, it would often be difficult to determine where first-line supervisory lines should be drawn since supervisors frequently perform multiple functions. Further, other modal administrations have not included supervisory personnel in the covered class. FRA amends its proposed definition of substance abuse professional (SAP), to include addiction counselors who have been certified by the National Association of Alcoholism and Drug Abuse Counselors Commission. This DOT-wide amendment is discussed in more detail in the common preamble. FRA also substitutes substance abuse professional for EAP counselor wherever that term appeared in the text of the Locomotive Engineer Certification regulation. Commenters did not offer any other changes to conform the language of 49 CFR part 240 to the language of this part. FRA adds a definition, ``violation rate,'' which measures the rail industry's overall random alcohol ``positive'' rate. Each year, the Administrator will examine the violation rate to determine whether the industry testing rate should be adjusted, as explained in Sec. 219.608 on performance-based testing. Sec. 219.9 Responsibility for compliance. Paragraphs (a) and (c) The recently enacted Rail Safety Enforcement and Review Act (RSREA), Public Law No. 102-365, amended the Federal Railroad Safety Act of 1970 (FRSA) (See 45 U.S.C. 438(a)) to clarify that FRA's safety jurisdiction extends to all entities that may violate the railroad safety laws. FRA amends this section to make clear that this part, like all regulations issued under authority of the FRSA, applies not only to railroads but also to any other entity that may violate this part, including independent contractors who provide goods and services to railroads and the employees of such contractors. Metro-North and the Metropolitan Transit Authority commented that this amendment appeared to make a railroad responsible for administering substance abuse programs for its independent contractor employees. That is not FRA's intent. FRA is concerned at this time only with ensuring that all persons who perform covered service for a railroad are subject to the same testing, regardless of whether the person has a direct employment relationship with the railroad. For FRA purposes, a railroad is in compliance if it contracts with its contractors who perform covered service to assure their compliance with part 219. The railroad may then, for example, attach copies of these contracts as part of its random alcohol testing plan submission. In practice, FRA anticipates that many contractors will develop a single substance abuse program for their employees, which can be recognized by all the railroads that the contractor serves. FRA has added regulatory text to clarify what is intended here. FRA will publish a separate notice to make conforming revisions to the penalty schedule set out in Appendix A to part 219.) Sec. 219.11 General conditions for chemical tests. Paragraph (g) As proposed, FRA continues to require supervisors to receive a minimum of three hours of combined training on alcohol and drugs. FRA received little comment on this issue. While two commenters felt that 60 minutes should be sufficient, particularly if offered as an annual refresher course, a commuter railroad that already provides 8 hours of supervisory training felt that 3 hours was a bare minimum. FRA believes that a minimum of 3 hours is sufficient, but now requires training on post-accident testing criteria and collection procedures. Railroads may integrate these required topics into their current training programs, which must be made available for inspection upon request. Sec. 219.13 Preemptive effect. Section 219.13, which states the preemptive effect of safety regulations issued under the FRSA (See 45 U.S.C. 434), remains unchanged. The Act expressly did not provide a new preemption standard; instead, it amended the FRSA to require the Secretary to issue rules, regulations, standards, and orders relating to alcohol and drug use in railroad operations. Sec. 219.15 Alcohol concentrations in blood and breath. This section is deleted since a revised alcohol concentration definition is included in Sec. 219.5. Sec. 219.21 Information collection. As mentioned above, FRA has replaced Sec. 217.13(d)'s reporting requirements with a management information system to collect information on railroad drug misuse programs. (MIS comments were addressed in that rule.) After railroads have implemented alcohol testing, FRA will require reporting of alcohol misuse programs data as well. In Sec. 219.801, FRA adds data elements on alcohol testing to complete the MIS' information collection on railroad substance abuse programs. FRA will use this data to monitor compliance and enforcement. Sec. 219.23 Railroad policies. Paragraph (f) FRA received no comments on this paragraph. In addition to the proposed requirements, FRA requires railroads to publish information to each covered employee on the signs, symptoms and effects of alcohol misuse, as well as available methods of intervention once an alcohol problem is suspected. FRA hopes that publishing such information will facilitate self- and peer-referrals, particularly when supplemented by the list of EAP resources that must be made available to covered employees under paragraph (d) of this section. Railroads must therefore provide the required information to all of their covered employees (not just those who have been identified as having a substance-abuse problem). Railroads may of course provide additional information beyond these requirements. Subpart B--Prohibitions As discussed above, existing provisions will remain in effect to allow for a smooth transition. The effective date for the amendments to this subpart is January 1, 1995. Sec. 219.101 Alcohol and drug use prohibited. Paragraph (a)(3), Pre-Duty Abstinence FRA proposed to prohibit a safety-sensitive employee from using alcohol either for four hours before reporting to perform covered service, or for the period of time running from when the employee receives a ``call,'' or notice to report for covered service, to when the employee actually reports for covered service, whichever is the lesser of the two periods. FRA also proposed to prohibit a railroad, through an employer's duty to prevent violations of Sec. 219.101 and Sec. 219.102, from allowing an employee to report for covered service, or go or remain on duty in covered service, if the railroad has actual knowledge that the employee is in violation of Sec. 219.101(a)(3). Industry commenters were concerned that specifying a pre-duty abstinence period would erode Rule G's prohibition on working with any alcohol in one's system, by implying that employees may report to work with alcohol in their system so long as consumption occurred before the required abstinence period. The Chicago and North Western (CNW) commented that pre-duty abstinence would be difficult to enforce, and should more properly be identified as the employee's responsibility. Several commenters felt that FRA's proposed period was too short. APTA, the Metropolitan Transportation Authority and SEPTA suggested mandating eight hours of pre-duty abstinence. The NTSB preferred a longer period, but would also support an eight hours of pre-duty abstinence if uniformly applied. New Jersey Transit (NJT), however, while recommending that pre-duty drinking should be absolutely prohibited, felt that the proposed abstinence period would be useful both as an enforcement and a public relations tool. NJT also felt that the employee should have the option to take a pre-duty test if either the employee or the employer suspects that the employee may be in violation. FRA had asked for comment on whether this pre-duty prohibition is workable for ``short-call'' employees, such as those who operate trains in pool crew service and off extra boards and signal maintainers who are subject to call without notice round the clock to handle ``trouble calls.'' In response to this query, the BLE and BRS recommended that FRA mandate by-pass provisions (mark-offs), particularly for signal employees subject to short call. The RLEA agreed that employees should be allowed to mark off and stay off-duty when impaired, and suggested that any potential for abuse could be curbed by setting limits; for example, requiring an EAP referral for any employee who marks off three or more times in a year. The TTD commented that pre-duty abstinence would be impracticable for those employees who are always subject to call, and that FRA should allow employees to mark off for any type of impairment. As discussed in the NPRM, FRA encourages railroads to adopt mark off procedures in the concept of voluntary programs such as Operation Red Block. However, because the successful implementation of such programs requires faithful adherence to mutual undertakings, FRA believes that implementation of such programs should be a bargaining issue rather than a federal mandate. Implementation of mark-off procedures therefore remains voluntary. In contrast, the eight hour pre-duty abstinence period suggested by the NTSB and some carriers would in effect impose total prohibition on short-call employees. After considering the comments, FRA believes that its proposal is the most workable compromise and thus the approach most likely to elicit voluntary compliance. Paragraph (a)(4) Rail management commenters uniformly supported a .02 standard (the equivalent of zero tolerance, because of the technological limitations of EBTs), because of concerns that the proposed system would preempt Rule G's longstanding prohibitions against alcohol use. Railroads opposed a .04 standard for the same reason. Because of the technological limitations of EBTs, rail labor considered a .04 standard to be the most defensible, and suggested that any alcohol level below be considered a negative. The RLEA saw no conflict between a .04 standard and Rule G, since railroads would remain free to maintain a zero tolerance standard under their own authority. The NTSB supported a .00 standard, which would not permit an employee with any positive BAC to perform ``safety-sensitive functions,'' citing evidence in aviation and highway research of a hangover effect on performance many hours after a person's BAC had returned to zero. After considering the comments, FRA believes that none of these recommended standards is appropriate. A true .00 standard could not be enforced, as a practical matter. A .02 standard would probably cause SAPs to waste time and effort on employees who do not in fact have alcohol or substance abuse problems, while a .04 standard would actually invite some employees to risk using alcohol, in the hope that their BAC would be below .04 by the time that they could be tested (e.g., employees at remote work sites). This approach would also deprive railroads of information that could be used to enforce Rule G and other, more stringent company policies. FRA therefore believes that the proposed ``bifurcated'' or ``two- tier'' system is necessary to avoid having the full consequences of a violation apply to those situations where an employee's test result indicates an alcohol concentration between .02 percent and .039 percent. The employee will not have to be evaluated by a SAP, or administered a return to duty test. However, since use of alcohol indicates that the employee may present a safety risk, the employee will not be allowed to perform covered service for a minimum of eight hours after administration of the test. With the sole exception of pre-employment tests for final applicants for employment (discussed below), FRA does not adopt the proposed retest option. Railroads were unsure as to what to do with an employee while waiting for his or her BAC to drop to below .02., since Rule G prohibits a covered employee who has any alcohol in his or her system from being on company property. Since this option is incompatible with traditional railroad policy, FRA believes that employees who are identified to have low levels of alcohol can more appropriately be handled through Operation Redblock and other existing programs specifically developed to fit industry needs. Like the very similar rule currently in effect, the bifurcated system does not preempt a railroad's independent authority to test and discipline under Rule G. As stated in Sec. 219.1, railroads retain the latitude to adopt more stringent standards under their own authority. For instance, railroads retain their authority to discipline an employee under company policy for a .02-.039 test result conducted under FRA authority or to discipline an employee found to have violated Rule G based solely on supervisory observations. The latter option should prove useful for very small railroads (who will not be required to conduct mandatory reasonable suspicion testing) when faced with an obviously drunk employee reporting for duty, particularly if an evidential breath testing device (EBT) is not readily available. Sec. 219.104 Responsive action. Existing provisions will remain in effect to allow for a smooth transition. As discussed above, however, the effective date for implementation of post-positive return to service and follow-up alcohol testing is January 1, 1995. Paragraph (d), Return to Covered Service FRA proposed to require a return to service test for the substance of the original positive only. Because of concerns about polydrug and alcohol abuse, discussed in greater detail in the section below, FRA also authorizes a railroad to require an employee to be tested for both alcohol and drugs, if such return to service testing has been recommended by the SAP for that employee. Paragraph (e), Follow-up Testing Earlier, FRA discussed the deletion of its current blood testing procedures. As proposed, FRA deletes the breath and blood testing procedures formerly contained in this paragraph. Testing will instead be conducted under part 40's more rigorous evidential breath testing procedures, which should provide employees ample reassurance that breath testing will be fair and technically sound. FRA had proposed that an employee's first year of follow-up testing after a post-positive return to covered service include a minimum of six unannounced tests that would track the basis for the employee's removal; that is, an employee who was removed for misuse of controlled substances or for refusing to provide body fluid samples under a mandatory provision of this part would be follow-up tested only for drugs, while an employee who was removed for misuse of alcohol or for refusing to provide breath under a mandatory provision of this part would be follow-up tested only for alcohol. After this first year, additional follow-up testing would be discretionary upon recommendation of the SAP, who is authorized to recommend follow-up testing for up to 48 months. FRA asked for comment on the issues of recidivism and polydrug abuse. In its compliance reviews, FRA uncovered instances of railroads failing to conduct follow-up tests on employees who had returned to covered service following positive drug tests. In several cases, routine random testing had detected drug use by post-positive employees. Moreover, FRA's study of clinical literature shows some incidence (10-50 percent) of individuals abusing both alcohol and one or more illicit or controlled substances. There is more evidence for individuals who cease drug misuse crossing over from drugs to alcohol; the cross-over from alcohol to drugs (and particularly those drugs contained in the DHHS test panel) is not as well documented. Comment was divided. The AAR and the ASLRA, among others, felt that follow-up testing should be left completely to the discretion of the SAP. Rail labor agreed. SEPTA and APTA however, wanted mandatory testing for 60 months, with APTA recommending a minimum of 12 tests the first year and 6 tests each year thereafter. APTA and all rail commenters wanted the authority to follow-up test for both alcohol and drugs, regardless of the substance of the positive. SEPTA offered data from its program, which includes 30 months of follow-up testing. Other than reasonable suspicion testing, SEPTA's highest positive rate (10.17 percent in the last fiscal year) was for post-positive return to service tests. To address concerns about employee relapse, FRA retains its proposed testing minimums while allowing for greater cross-substance testing. Broader monitoring may help detect instances where an employee switches to alternative drugs after being deprived of his or her primary drug of abuse. FRA therefore requires a return to service test and six follow-up tests in the first 12 months for the substance (or class) of the original positive. In addition, a railroad may choose to test for both alcohol and drugs at the return to service test, and at follow-up tests for a total of up to 60 months, if such testing is recommended by the SAP. The SAP may make such additional testing part of a conditional recommendation for the employee's return to service. This approach allows for flexible case management, recognizing that an employee who misuses either alcohol or a controlled substance could have a more general propensity to abuse psychoactive drugs (i.e., active polydrug abuse or incipient cross-addiction). Its intent is to eliminate any question as to the employer's right and obligation to provide for aftercare monitoring of employees who have violated alcohol or drug rules. Paragraph (f) (As discussed above, the .02-.039 rule text is now in Sec. 219.101(a)(4).) This paragraph seeks to ensure that the SAP places the interest of safety above other considerations, when recommending treatment or advising the railroad about whether to return an employee to covered service. Subpart C--Post-Accident Toxicological Testing The common preamble discussion of post-accident testing does not apply to FRA's program, which unlike those of other modes, requires full toxicological testing following designated accidents and incidents. As discussed above, existing provisions will remain in effect to allow for a smooth transition. The effective date for the amendments to this subpart is January 1, 1995. Sec. 219.201 Events for which testing is required. Paragraphs (a)(1)(iii) and (a)(2)(ii), Damage thresholds for Major Train Accidents and Impact Accidents FRA received little comment on its proposal to raise the amount of railroad property damage required for major train accidents and for impact accidents, which was prompted by recent low positive rates. Accordingly, FRA redefines Sec. 219.201(a)(1)(iii) (major train accidents) to require testing for accidents that cause $1,000,000 or more in damage. FRA also redefines Sec. 219.201(a)(2)(ii) (impact accidents) to require testing for non-injury accidents that cause $150,000 or more in damage. The NTSB uses this reporting threshold (although the NTSB, unlike FRA, includes damage to lading). FRA expects that much of this data will be replaced through railroad for cause testing. (Commenters were silent as to whether Subpart D testing should be required for events between the existing $50,000 threshold and the proposed $150,000 threshold where there is a reasonable belief that the particular employee contributed to the accident.) Paragraph (b) [Exceptions] All commenters supported FRA's proposal to exclude from post- accident testing accidents that otherwise qualify but are clearly attributable to vandalism. (The NTSB did not comment.) FRA therefore adopts this proposal. Consistent with the other exceptions, the vandalism exception holds railroad supervisors to Sec. 219.201(c)'s reasonable inquiry/good faith judgment standard when making determinations. Sec. 219.203 Responsibilities of railroads and employees. Paragraph (a), Employees Tested As discussed above, once a railroad has made a good faith determination that a qualifying event has occurred, the railroad may conduct breath alcohol testing if an EBT is available and breath testing does not interfere with or delay collection of samples for mandatory blood and urine testing. While railroads must still collect blood and urine samples for full toxicological testing, this option allows quick enforcement action since there is no need to wait for laboratory analysis. Although FRA has moved the authority for optional breath testing from for cause testing to post-accident testing, the procedures specified in Sec. 219.303 have not been retained. Instead, railroads who choose to exercise this option must conduct breath testing in accordance with the new Part 40 procedures published today. FRA will revise Forms 6180.73 and 6180.74 to allow for railroad reporting of EBT test results. Form 6180.73 will add a check off box on the railroad representative's copy asking whether one or more employees were breath tested. Form 6180.74 will add a similar check off box for each tested employee. If breath alcohol tests are conducted, the railroad may either attach a copy of each Part 40 breath alcohol testing form (the railroad will have to make an additional copy for FRA purposes, since the Part 40 form allows only for BAT, employee, and employer copies) to the FRA forms to be shipped with the toxicology kit, or send a copy of each Part 40 form directly to FRA within 10 days of the tests. (Appendix C will contain revised instructions.) FRA expects the results from breath and blood alcohol testing to almost always vary, since an EBT test will usually be conducted hours before the railroad is able to collect blood and urine samples, allowing the employee's BAC to change in the interim. The breath test result will always stand on its own, however, since by definition an EBT test always meets evidential standards. FRA will use the results from all three types of testing (breath, blood, and urine) for post-accident toxicological analysis, since the purpose of mandatory toxicological testing remains unchanged. Blood and urine samples will continue to be collected as independent aids to accident investigation, not as backups for EBT test results. Sec. 219.205 Sample collection and handling. FRA will publish an amended Appendix C to part 219 to modify post- accident procedures, toxicology kits, forms, and instructions to accommodate optional breath alcohol testing and mandatory urine split sample procedures. (FRA already splits blood samples into two sealed 10 milliliter tubes.) FRA will announce when the new kits and forms become available. Upon employee request, FRA's post-accident toxicological laboratory will ship the employee's sealed split samples to another DHHS-certified laboratory for testing. Sec. 219.209 Reports of tests and refusals. Paragraph (b) With the addition of the breath test option described above, a railroad must report to FRA any refusals to provide breath for testing. Paragraph (c) In an earlier Safety Recommendation, the NTSB recommended that FRA require railroads to submit notification, including reasons for the delay, whenever a post-accident test could not be conducted within four hours of the qualifying accident or incident. The NTSB's current recommendation reduces this testing window from four hours to two. Such a requirement would in effect compel railroads to submit a report for each qualifying event, since the average FRA post-accident collection time is 5.5 hours. FRA requires reporting but will minimize the burden by adopting the NTSB's original four hour recommendation, which is a more workable window for post-accident testing. In addition, while reports must be prepared and maintained for inspection upon request, they do not have to be submitted to FRA. Sec. 219.211 Analysis and follow-up. Paragraph (a) As part of its amended Appendix C, FRA will publish a summary of its post-accident testing protocols, which have been submitted to the DHHS. Sec. 219.213 Unlawful refusals; consequences. A refusal to provide breath for optional testing will have the same consequences as a refusal to provide bodily fluid samples for mandatory testing; the employee shall be disqualified from covered service for nine months. Subpart D--Testing for Cause As proposed, FRA amends the heading of this subpart to differentiate now-mandatory reasonable suspicion from accident/incident and rule violation testing, which remain discretionary. This new heading also conforms to the classifications used in FRA's drug MIS. As discussed above, existing provisions will remain in effect to allow for a smooth transition. The effective date for the amendments to this subpart is January 1, 1995. Sec. 219.300 Mandatory reasonable suspicion testing. Noting that an alcohol test is not needed to confirm possession of alcohol or controlled substances, the AAR considered the proposed language in this paragraph overbroad since it would require alcohol testing whenever a supervisor suspected any violation of Sec. 219.101(a). FRA agrees, and has amended this paragraph to require testing only when use of alcohol is suspected. (The Act specifies that reasonable suspicion testing must be conducted for use of alcohol or controlled substances.) Therefore, a suspicion of possession alone does not require a reasonable suspicion test. FRA will not add long-time decline in job performance as a factor for the employer to consider when determining whether to conduct a reasonable suspicion test on an employee. Only one commuter line supported this proposal. The AAR thought this additional factor unnecessary, since a supervisor could only consider long-term performance in conjunction with specific contemporaneous observations indicative of substance abuse, and such observations would provide an independent basis for a now mandatory reasonable suspicion test. Rail labor agreed, commenting that employers have authority to conduct reasonable suspicion testing without this factor, and that long-term decline in job performance would more properly be handled by EAPs, since it is often caused by personal problems other than substance abuse. The NTSB did not comment on this issue. (For purposes of intermodal consistency, and to avoid any misimpression that a reasonable suspicion drug test could be required solely based on perceived long-term decline in performance without a contemporaneous manifestation clearly tied to drug use, FRA has deleted the following phrase: ``[s]uch observations may include indications of the chronic and withdrawal effects of drugs.'' Deletion of this phrase is not intended to limit a supervisor from taking into consideration any contemporaneous indication of drug use. But FRA cautions that chronic and withdrawal effects alone will seldom yield unambiguous contemporary indications in the case of either alcohol or drugs. Where unambiguous withdrawal effects are encountered, prompt medical attention will normally take precedence over occupational alcohol or drug testing. Where a long-term decline in performance occurs, referral under subpart E of this rule (``Identification of Troubled Employees'') is the indicated course of action. Paragraph (b) FRA had proposed a ``logic tree'' for situations where a supervisor reasonably suspects an employee of substance abuse, but is unable to determine whether the medium of abuse is alcohol or controlled substance(s). FRA continues to recommend, but will not require, the following process: a supervisor should conduct an alcohol test first whenever an employee's symptoms could be consistent with either alcohol or controlled substance use. If the breath test result is below .02, the supervisor should conduct a urine drug test to continue the search for an explanation. If the breath test result is .02 percent or more, drug urinalysis is optional. Because of polydrug abuse, Metra and SEPTA considered this logic tree unnecessary, suggesting that employers should be allowed to conduct both a breath alcohol test and a urine drug test whenever reasonable suspicion exists to test for alcohol. Although FRA agrees that employers need some latitude to inquire into suspected polydrug abuse, FRA recommends that supervisors decide step by step whether additional testing is needed, instead of automatically proceeding to test for both alcohol and controlled substances. Paragraph (c)(1) Reaction was cleanly split on FRA's proposal to prohibit a supervisor who makes a reasonable suspicion determination from also conducting the breath test for that employee. Railroads uniformly opposed this prohibition as both impractical and unnecessary. Carriers who conduct for cause testing programs under their own authority had experienced no problems with supervisory abuse of discretion. To document their lack of supervisory depth, several AAR member railroads submitted examples of locations where normally only one supervisor would be available. Commenters also noted that testing could only be conducted if a trained supervisor had an articulable basis for his or her suspicion. Additionally, the employee would observe the entire breath alcohol testing process and the NPRM required all breath testing devices to meet evidential standards. In contrast, rail labor supported the determining supervisor as BAT prohibition. The BLE and BRS noted that FRA's requirement of two supervisors (one of whom must be trained) to make a reasonable suspicion drug testing determination has been in place for several years. In theory, therefore, the same two supervisors would also be available for reasonable suspicion alcohol testing. More importantly, this requirement makes more likely that each breath alcohol test will have an independent witness. FRA believes that a compromise may be possible, particularly in light of the Department's decision to allow employers to use PBTs for screening tests. In the above-mentioned NPRM on blood testing procedures, OST proposes to allow the determining supervisor to conduct the screening test on an EBT or a PBT (if an EBT is unavailable), but not the confirmation test, which must still be conducted on an EBT by another individual. This would provide testing flexibility for reasonable suspicion events that occur in remote locations, while still requiring all confirmation tests to be conducted by a person not in the employee's chain of command. Comments on this proposal should be submitted to OST's docket. Paragraph (c)(2) Commenters raised similar pros and cons for FRA's proposal to require two trained supervisors to make a determination to conduct a reasonable suspicion drug test. In response, FRA will instead retain its existing requirement of two supervisors, one of whom must be trained in the signs and symptoms of substance abuse (as provided for in Sec. 219.11(g)), to make a reasonable suspicion drug testing determination. Paragraph (d) Under FRA's original proposal, a major concern for commenters was the possibility that testing situations could frequently occur in remote locations where EBTs are unavailable. Where this occurred, a railroad could still enforce prohibitions against alcohol misuse under its own authority, but would be unable to conduct federally mandated testing. At a minimum, permitting the use of PBTs should enable railroads to conduct screening tests in most circumstances. As discussed above, FRA proposes to allow confirmation blood alcohol testing. Under current FRA procedures, however, if the screening test result indicates alcohol misuse, the supervisor must determine whether a confirmation test can be conducted on an EBT within eight hours of the screen. If an EBT is unavailable, the unconfirmed screening test will be considered a no test. However, once a supervisor suspects alcohol misuse, he or she must enforce the prohibitions in Subpart B even if the employee cannot be tested at that time. FRA requires the supervisor to comply with the prohibitions in Sec. 219.101(a)(4) and send the employee home for at least eight hours. As always, a railroad may also take independent enforcement action under its own authority. Sec. 219.301 Testing for reasonable cause. FRA has reformulated its proposed language in response to the ASLRA's comment that the phrase ``based on affirmative evidence of unsafe conduct'' implied a probable cause standard for testing. Instead, FRA will require a supervisor to have ``a reasonable belief, based on specific, articulable facts, that the employee's acts or omissions contributed to the occurrence or severity of the accident or incident''. This case law-derived language, while clearly a lesser standard than traditional probable cause, stresses that a supervisor must have more than a hunch or a guess upon which to base his or her determination. This restriction applies only to testing under FRA accident/incident authority, however. Sec. 219.302 Prompt sample collection; time limitation. Paragraph (b)(2) Contrary to the discussion in the common preamble, the eight hour time limit for completion of accident/incident for cause testing, runs from the time a responsible railroad supervisor receives notice of the event providing reasonable cause for conduct of the test. Paragraph (f) A railroad must prepare and maintain a report, similar to the post- accident delay report required in Sec. 219.209(c), whenever a reasonable suspicion test cannot be conducted within two hours of the observations or events that were the basis for the railroad's reasonable suspicion determination. The report does not have to be submitted, but must be made available upon request. Sec. 219.303 Alcohol test procedures and safeguards. As discussed earlier, FRA will allow PBTs to be used for screening tests in response to commenters' near unanimous demand for greater flexibility. By allowing the use of PBTs, however, FRA does not excuse railroads from making their best efforts to ensure the availability of EBTs for routine testing situations. Sec. 219.303 Breath test procedures and safeguards. Paragraph (c) As discussed above, FRA will adopt part 40 blood alcohol testing procedures upon publication of the Department's final rule. This paragraph is accordingly deleted. Subpart F-Pre-employment Tests Existing provisions will remain in effect to allow for a smooth transition. As discussed above, however, the effective date for implementation of pre-employment alcohol testing is January 1, 1995. Sec. 219.501 Pre-employment tests. As mentioned above, FRA adopts the proposed retest option only for final applicants for employment. An applicant whose test result indicates an alcohol concentration between .02 percent and .039 percent will not be allowed to perform safety-sensitive service until he or she has a later test result indicating an alcohol concentration of less than .02. Since rule G applies only to current railroad employees, allowing a retest option for applicants is compatible with traditional industry policy. This option also makes FRA pre-employment policy administratively consistent with the rest of the Department. Commenters expressed no interest in the proposed ``grandfathering'' provision, which would have allowed railroads to exempt a covered employee from pre-employment testing if a background check indicated that the employee had not had any violations of this part or of the alcohol and drug misuse rules of another DOT agency within the last six months. FRA has decided to delete this option, since the added flexibility it would have provided is unnecessary for the rail industry, which traditionally has a stable employee population and a low turnover rate (unlike trucking, for example). Subpart G-Random Testing Programs Sec. 219.607 Railroad random alcohol testing programs. Paragraph (a) All commenters on the issue favored combined alcohol and drug random testing programs. The BLE and BRS recommended this approach as a cost effective means of deterrence, since an employee could not know in advance whether he or she was being tested for alcohol, for drugs, or for both. Transtar felt that a single testing program would reduce both cost and confusion. Metra and APTA supported using the same selection to conduct alcohol and drug testing, while Amtrak wanted the flexibility to conduct alcohol and drug tests simultaneously for some employees, but at different times for others. FRA agrees that flexibility is important, and will allow railroads to combine plans/and or testing, if desired. A railroad may submit a separate alcohol testing plan or a combined testing plan for approval. The combined testing plan may be a previously approved random drug testing plan that has been modified by adding alcohol testing elements (any modifications should be specified and highlighted, however). FRA will, for the most part, duplicate the notice requirements and criteria for plan approval that were used to implement random drug testing, including a three-tier submission and implementation schedule that allows smaller carriers additional time to develop and execute random alcohol testing plans. Class I and commuter railroads must submit random alcohol testing plans for FRA approval within 6 months after publication of this rule and must implement random alcohol testing beginning on January 1, 1995; Class II railroads must submit plans within 12 months after publication and must implement testing beginning on July 1, 1995, and Class III railroads must submit plans within 18 months after publication and must implement testing beginning on January 1, 1996. (For the reasons discussed below, there will be no phase-in of the testing rate.) Paragraph (b)(2) Commenters (including, among others, the APTA, AAR, ASLRA, BLE, BRS and RLEA) were close to unanimous in supporting a 10 percent testing rate the first year, followed by an annual performance-based rate set according to each railroad's random alcohol positive rate for the preceding year. In support of this low initial rate, commenters pointed to the existence of an extensive peer prevention and employee assistance system, the addition of new types of mandatory (pre- employment and reasonable suspicion) alcohol testing, and supervisory knowledge of the signs and symptoms of alcohol, making alcohol misuse comparatively easy to detect. Commenters also cited the General Accounting Office's study and the results of the AAR's test project, as evidence that lowering testing rates does not affect deterrence. Support was also widespread for railroad-specific performance-based testing (for both alcohol and drugs), both as an incentive for carriers to try alternative methods of deterrence, and as a reward to those carriers who have viable peer referral and bypass programs. FRA agrees that the initial random rate for alcohol testing should be lower than the current 50 percent random drug testing rate, and that testing rates should then be tied to the preceding year's positive rate. However, FRA believes that a 10 percent testing rate is the absolute minimum level of effort required to sustain awareness and deterrence. Moreover, FRA is concerned that a railroad-specific performance-based system could create an incentive for some railroads to report and/or keep records dishonestly, thus requiring considerable enforcement effort to guard against such improper manipulation. Such a system would also be difficult to apply to smaller railroads, since even a small variation in the number of positives (e.g., 0 vs. 1) would result in high year-to-year volatility in each railroad's required testing rate. In other words, the small size of a given railroad's population and the resulting small number of tests could create variations in positive rates that do not truly reflect that railroad's incidence of employee misuse. Therefore, FRA (along with FAA, FHWA and FTA) will instead set a 25 percent initial rate for alcohol testing, with performance-based alcohol testing by industry. Each railroad is expected to achieve the required 25 percent testing rate at the inception of random alcohol testing. (FRA intends to discuss the random rate for drug testing in a separate NPRM). As discussed in Sec. 219.608, the Administrator will publish a minimum random testing rate, which will be determined by the overall violation rate for the rail industry. Paragraph (b)(3) Because of the short half-life of alcohol in the body, railroads may conduct random alcohol testing at any time the employee reports for work and at any time during the duty tour (except for when the employee is expressly relieved of any responsibility for the performance of safety-sensitive duties). Paragraph (b)(6) An employee may be tested only when he or she is on duty and either performs covered service or is immediately available to perform covered service. Therefore, the issue of commingled service is critical to plan approval, since off-duty consumption of alcohol is not proscribed (so long as the employee does not report for duty with a BAC of .02 or more). Plans must be carefully designed to subject commingled service employees to testing only to the extent that they perform or can reasonably be expected to perform covered service during a given duty tour. Industry commenters wanted the flexibility to conduct testing at different times during an employee's ``duty tour.'' FRA has no objections, so long as the date or time of the test is not predictable to the employee. For instance, a railroad could not cluster its tests exclusively at the beginning or end of the duty tour. (FRA will not adopt its proposal to limit a railroad's time in which to notify an employee to a period running from 30 minutes before to 30 minutes after the employee performs safety-sensitive functions, plus applicable travel time.) Sec. 219.608 Administrator's determination of random alcohol testing rate. For at least three calendar years (because of staggered implementation dates, the minimum amount of time needed for all classes of railroads to have one year of alcohol testing data), railroads will conduct random alcohol testing at a 25 percent annualized rate. Subsequently, FRA will use employer data obtained from MIS reports to determine the industry's violation rate, which will in turn determine the minimum testing rate to be authorized by the Administrator. Any random rate change indicated by industry performance will occur at the beginning of the calendar year. Railroads will remain free to test at higher rates provided they treat all employees alike. Subpart H--Procedures and Safeguards for Urine Drug Testing and for Breath Alcohol Testing As noted above, the urine drug testing split sample procedures specified in part 40 are mandatory beginning on August 15, 1994 as are the amendments to this subpart concerning urine drug testing. Sec. 219.708 Employee requests for testing. Comments were cleanly divided as to how much time should be allowed for an employee to request a test of his or her split sample. Carriers uniformly supported the 72 hour period proposed by OST, while labor wanted to duplicate the 60 day period currently allowed in FRA's retest option. Although FRA agrees with labor that the retest option has worked well in the past, FRA will no longer allow retests of primary samples. Instead, to be consistent with the rest of the Department, FRA will allow split sample testing only. An employee will have 72 hours (the minimum contemplated by the Act) within which to request a test of his or her split sample, running from when the employee receives actual notice from the medical review officer (MRO) that the employee's test result has been verified as positive. As an additional safeguard, the MRO must inform an employee who has a confirmed test result of his or her right to request a split sample test at the time of the MRO's initial contact with the employee, even though the employee's 72 hours do not begin to run until after the MRO has completed the verification process. The employee may also present information to the MRO if he or she believes that unusual circumstances (e.g., an intervening holiday, the unavailability of the MRO) prevented him or her from requesting a split sample test within 72 hours. Sec. 219.715 Alcohol testing procedures. As discussed, FRA will permit railroads to use EBTs that qualify for NHTSA's CPL but do not meet part 40 requirements (lacking sequential numbering and printout capability, for example) as PBTs for screening tests. This allows those railroads that already conduct alcohol testing under their own authority (such as Amtrak and the CNW), to recoup the full useful life of previously purchased EBTs, while allowing other railroads the flexibility to reduce costs. An EBT test is still required for confirmation, however. Subpart I--Annual Report Sec. 219.801 Reporting of alcohol misuse prevention program results in a management information system. FRA recently published a final rule implementing a management information system to obtain and analyze employer drug misuse program data. Included in this rule are comparable information collection requirements for alcohol misuse program data. Appendices D3 and D4, attached as exhibits to this rule, are respectively the standard and ``EZ'' forms for reporting of employer alcohol misuse program data. (These forms are similar to Appendices D and D2 in the drug MIS.) FRA will not ask for information on operational tests and inspections or supervisory training, since reporting on these combined programs is already required under the drug MIS. Like the drug MIS, the alcohol MIS asks for information on for cause alcohol testing (although these data elements appear repetitive, they are not identical since the charts have been modified to reflect the two-tier or bifurcated system). Once alcohol testing has produced a year's worth of data, FRA will republish a combined MIS form for alcohol and drug program data. Subpart J--Recordkeeping Requirements Secs. 219.901 and 219.903 Retention of breath alcohol testing records and urine drug testing records. FRA adopts the retention requirements as proposed. Although several commenters felt that these proposed requirements were too burdensome, these sections conform with the departmental requirements published in part 40 today. In addition, FRA retains Sec. 219.713's current requirement for railroads to maintain for at least five years a summary record of each covered employee's test results. Sec. 219.905 Access to facilities and records. FRA proposes to replace Sec. 219.713's language on access to records and adopt 49 CFR Sec. 40.81, which allows an employer to release an employee's alcohol and drug test records only with the written consent of the employee, with the exceptions of when access is requested by a DOT agency or the NTSB or when access is requested by a decisionmaker in a legal proceeding relating to a benefit sought by the employee. Executive Order 12866 and DOT Regulatory Policy
Alcohol Testing; Amendments to Alcohol/Drug Regulations; Final Rule
FRA issues a final rule to conform its regulations on control of alcohol and drug use in railroad operations to the requirements of the Omnibus Transportation Employee Testing Act of 1991 (Act). Among other changes, FRA now requires pre-employment and random alcohol testing of safety-sensitive employees, and makes the reasonable suspicion component of for cause testing mandatory for both alcohol and drugs. FRA also amends its procedures to incorporate split sample testing and to incorporate the departmental alcohol testing procedures published elsewhere in today's Federal Register.