59778 Federal Register / Vol. 57, No. 241 / Tuesday, December 15, 1992 / Proposed Rules

DEPARTMENT OF TRANSPORTATION

Federal Aviation Administration

14 CFR Part 121

Federal Highway Administration

49 CFR Parts 382 and 391

Federal Railroad Administration

49 CFR Part 219

Federal Transit Administration

49 CFR Part 653

Research and Special ProgramsAdministration

49 CFR Part 199

Coast Guard

46 CFR Part 16

JOST Docket No. 48498, Notice 92-291

RIN 2105-AB94

Random Drug Testing Program

AGENCY: Office of the Secretary, FederalAviation Administration, FederalHighway Administration, FederalRailroad Administration, FederalTransit Administration, Research andSpecial Programs Administration, andthe United States Coast Guard, DOT.ACTION: Advance notice of proposedrulemaking (ANPRM).

SUMMARY: In response to public, comments received during the

President's regulatory moratorium,petitions submitted by industry, and ontheir own initiative, the five operatingadministrations that currently requirerandom drug testing-the FederalAviation Administration (FAA), theFederal Highway Administration(FHWA), the Federal RailroadAdministration (FRA), the Research andSpecial Programs Administration(RSPA) and the United States CoastGuard (USCG) (collectively referred toas "DOT" or "the agencies")- areconsidering modifying the random drugtesting program. [The Federal TransitAdministration (FTA) -also a part ofDOT-published elsewhere in today'sFederal Register a drug testing NPRMthat proposes to adopt, among otherthings, the 50 percent random testingrate that is currently used by the otheroperating administrations. This ANPRMis not intended to affect the randomdrug testing program to be establishedby that rulemaking. Rather, this ANPRMwill review the general issue. FTA joinsthe other operating administrations in

issuing this random drug testingprogram notice requesting publiccomment and data concerning whetherthere are less costly alternatives to thecurrent random testing program that canmaintain an adequate level of deterrenceand detection of illegal drug use.DATES: Comments are due February 16,1993. A public meeting on technical andscientific issues will be held onFebruary 1, 1992, from 8:30 a.m. to 5p.m. and on February 2, 1992, from 8:30a.m. to 12:30 p.m. Seating is limited to150, which will be offered on a first-come, first-served basis.ADDRESSES: Comments should be sent toDocket 48498, Office of DocumentaryServices (C-55), U.S. Department ofTransportation, room 4107, 400 SeventhStreet, SW., Washington, DC 20590-0001. To provide a copy for eachoperating administration's docket and tofacilitate the Department's review, werequest that an original and sevenadditional copies of each comment besubmitted.

The technical meeting will be held atthe Holiday Inn--Capitol, 550 C St.,SW., Washington, DC 20024. (202) 479-4000. The hotel is offering participantsa special rate, if they mention themeeting when they reserve their room.FOR FURTHER INFORMATION CONTACT: Dr.Donna Smith, Acting Director, Office ofDrug Enforcement and ProgramCompliance, (202) 366-3784.SUPPLEMENTARY INFORMATION: TheDepartment of Transportation seekscomment on the effectiveness of thecurrent random drug testing program fortransportation industry employees. Inthe April 1992 "Report to the President:Review of Regulations," DOT SecretaryCard committed to review this issue anddetermine whether adequate deterrenceand detection of illegal drug use couldbe achieved at a lower rate of randomtesting and at a lower cost. In addition,several petitions have been filed withvarious operating administrationsseeking to simply lower the rate ofrandom drug testing.

The purpose of this notice is to seekdata and ideas on additional strategiesthat would ensure the continuedeffectiveness of the Department's anti-drug program while reducing its cost.The Department is reluctant to proposea specific change in random testingrequirements at this time because of thelack of data for transportation industryworkers and because most of theoperating administration programs haveonly been in place for a short time.Other than the experience of themilitary services that was relied onwhen the operating administrationsoriginally set their random testing rates

at 50 percent, we are unaware of anylong-term reliable data indicating therelative deterrent effectiveness ofdifferent random testing rates. However,as described later in this document,DOT is gathering data that may help onthis issue. Although the anti-drugregulations have been promulgated byvarious DOT agencies, we are issuing aDepartmental ANPRM because of thecommonality of the issues. We invitecomment and supporting data onwhether different programs for differentindustries would be appropriate. Wehave numbered specific questionswithin brackets throughout the ANPRM,and would appreciate it if commenterswould reference those numbers in theirresponses.Regulatory Background

DOT agencies have been involved indrug testing since the mid-1980s. TheUSCG has tested its uniformedpersonnel for drug use since 1982. DOT

egan testing certain of its civilianemployees in September 1987.

The Department's civilian employeeprogram is tightly controlled, centrallyadministered by headquarters staff, andmonitored daily. Employee awarenessand the visibility of the program aremaintained through training programsconducted by regional drug programcoordinators. Specimens are collectedby a single contractor service, whichoperates under a uniform standard ofprocedures that provides for consistentand reliable collections.

The random testing program wasphased-in and, by September 1988, DOTwas testing a population of nearly33,000 employees (primarily air trafficcontrollers, safety inspectors andindividuals with high securityclearances) at a testing rate of at least 50percent each year for illegal drug use.The annual rate of positive random testshas declined from about 0.83 percent toas low as 0.21 percent over the last fiveyears. Over the past three years, the rateconsistently stayed well below 0.5percent. The reports indicated that inthis homogeneous, skilled, and stablepopulation, there was no distinction inthe percentage of positive testing resultsbased on geography, age, etc. As a resultof the apparent deterrent effect of thetesting program as demonstrated bycarefully-maintained recordkeeping,long experience, and the decreasingnumber of positive results, theDepartment lowered its federalemployee random testing rate this year.Effective March 1, 1992, the Departmenthas been conducting random testing ata rate of at least 25 percent annually.The positive rate continues to remain ata similarly low level. The Department

Federal Register I Vol. 57, No. 241 / Tuesday, December 15, 1992 J Proposed Rules

will continue to evaluate the data andmay adjust the random testing rate, ifnecessary, beck to 50 percent. TheDepartment estimates that the reductionin the testing rate will saveapproximately 20 percent of the annualcollection and laboratory testing costs.

The Federal Railroad Administration(FRA) has the longest experience withdrug testing programs applicable totransportation industry workers. In1986, railroads began pre-employment,post accident, and reasonable cause/suspicion testing. as required by theFRA.

In 1988, the Department ofTransportation issued six final rulesmandating anti-drug programs forcertain transportation workers in theaviation, interstate motor carrier.pipeline, maritime, rail and transitindustries. The rules includedrequirements for education, training,testing and sanctions. The testingcomponent of each program includedpre-employment, post-accident,reasonable cause, periodic (for thosesubject to periodic medicalexaminations) and random drug testingfor approximately four million workersin safety-sensitive positions. Based onextensive experience and success intesting military and other populations,the Department imposed widescalerandom testing requiremeats becauseunannounced random drug testing isgenerally regarded as the best method ofdeterring illeval drug use.

The operating administrations' rulesimposed a random testing rate of at least50 percent per year. This means that ifan employer has 400'coveredemployees, the employer must conductat least 200 tests per year. Selection fortesting must be random, with everyemployee in the random pool having anequal chance of being chosen each timea selection is made. Because of therandomness, some employees could betested more than once in a given year,while others might not be tested foryears. However, every coveredemployee would know that he or shehad one chance in two of being testedeach year. Employers were alloved tophase-in random testing at a rate of 25percent for the first year, but had toincrease to at least a 50 percent rate afterone year.

After the final rules were Issued,lawsuits delayed implementation of therules for three of the six DOT agencies.As of today, only transit workers are notcovered by the testing regulations. The1988 final rule adopted by the FederalTransit Administration (formerly calledthe Urban Mass TransportationAdministration) was vacated by theUnited States Court of Appeals for the

District of Columbia Circuit because ofa lack of statutory authority. Legislation(the FTA provisions of the OmnibusTransportation Employee Testing Act of1991, Pub. L 102-143, Title V, October28, 1991) was subsequently enacted toremedy this problem as well as addressother significant concerns with alcoholabuse and illegal drug use byindividuals in various transportationindustries who perform safety-sensitiveduties. An NPIM proposing to cover,transit employees is publishedelsewhere In today's Federal Register.As proposed, transit employees wouldbe subject to a random testing rate of atleast 50 percent. As noted above, thisANPRM involves a longer-term revieWof random testing program issues and isnot intended to effect the random ratedecision that must be made as part ofFTA's proposed rulemaking. Sevoealother DOT agency-regulated industriesor industry sectors only recently beganrandom testing at a 50 percent rate peryear.

The Federal Railroad Administrationphased-in random testing in threegroups: Large railroads, medium-sizerailroads, and short line railroads. InJanuary 1990, large railroads begantesting at 25 percent, medium-sizerailroads began testing at 25 percent inJuly 1990 and short line railroads begantesting st 25 percent in November 190.Random testing at a 50 percent ratebegan one year after these dates for eachof the three categories.

In the aviation industry, the 25percent rate was Instituted for large aircarriers in December 1989, for medium-size carriers in April 1990, and for thesmallest carriers in August of 1990.Testing at 50 percent began one yearafter the initial phase-in. Testing ofcontractor employees (such as repairstation personnel or security screeners)began one year after the carriers thatthey worked for or supported.

Testing of pipeline personnel begannext. Phase-in (25 percent testing) beganin April 1990 for large operators and inAugust 1990 for small operators, withthe 50 percent rate implemented oneyear later by each group.

Random and non-suspicion-basedpost-accident drug testing in the motorcarrier industry vere enjoined by courtorder, although the other types of testingwere implemented on December 21.1989. After the injunction was lifted,random testing by large motor carriersbegan in November 1991 at a 25 percentrate and testing by small motor carriersbegan in January 1992 at a 25 percentrate. One year after these dates, the rateincreases to 50 percent. (The currentrule covers just interstate motor carriers,but a proposed rule in today's Federal

Register would extend coverage to allholders of commercial driver's licenses,including employees of intrastate motor.carriers and school bus drivers.)

The USCG rule regarding randomtesting of merchant seamen wasenjoined by court order in December1989. Other types of testing werephased-in commencing in June 1989. InJuly 1991, the USCG Issued a revisedrule addressing the court's concerns andjustifying the categories of employeessubject to random testing. in October of1991, the maritime industry begantesting at a 25 percent random rate witha requirement to Increase to a 50 percentrate one year after implementation.There -was no distinction between largeand small maritime employers for thisimplementation of random testing.Only the Federal Aviation

Administration and the Federal RailroadAdministration require their regulatedemployers to report testing statistics tothem. The Federal HighwayAdministration, the U.S. Coast Guardand the Research and Special ProgramsAdministration review recordsmaintained by covered employers, butdo not have composite data for allpositive test reeults in their industries.Separate NPRMs published in today'sFederal Register would require that theemployers (or industries) regulated byall six operating administrations submituniform data conoening drug testing onan annual basis to thoseadministrations. (Data from the motorcarrier industry wouid be gathered on asurvey basis.)

Purpose of Random Testing

The primary purpose - theDepartment's anti-drug rules is toensure saiety by deterring drug use,with detectiom an important collateralbenefit. The integrity of the randomselection process, the timing of thecollection, the use of correct collectionprocedures, andihe credibility of theMRO (medical review officer)verification actions must be consideredwhen assessing the ofectivenes of anyrandom testing program in deterringdrug use and detecting drug users.When -drug users to undetected becauseof errors in the selection, collection, orreview processes. the deterrent -value forthat employee and others is stronglydiminished.

Random testing to achieve deteneeis a form of drug use prevention.Prevention generally is carried out usingone or mnor of four strategies:Education. persuasion, motivation orfacilitation. Many companies have anexisting drug education program.Employer policies, supervisor training,and peer-identification programs are

59779

57, No. 241 / Tuesday, December 15, 1992 / Proposed Rules

examples of a persuasion strategy. Drugtesting is usually considered a tool fora motivational strategy In drug useprevention. Comprehensive EmployeeAssistance Programs are a majorcomponent of a facilitation model. Inour view, workplace drug useprevention programs should consist of amix of the four prevention strategies.These strategies must maintain orincrease the visibility of the programand, therefore, the employees'awareness of it. Different employers anddifferent industries may find differentmixes of education, persuasion,motivation, and facilitation moreeffective and efficient in deterring druguse.

As the Department noted during thedrug testing rulemakings and as FRAfound when it added random testing toan existing testing program, randomtesting does deter use. FRA and themilitary services noted reductions in thepositive rates for other types of drugtesting after starting random testing. TheDepartment seeks comment and data onthe minimum effective random testingrate.

[1] Does increasing or decreasing thetesting rate affect an individual'sperception of the chances of being"caught" and increase or decreasedeterrence? Why?

[2] Is there a mathematicalrelationship between the testing rateand deterrence, e.g., does doubling therate double the deterrence?

[31 How does the increase or decreasein the rate affect the costs of testing,considering constant costs that existregardless of the testing rate?

[4] To the extent the rate of positivetests is not expected to increase with arandom testing rate reduction, pleaseexplain why.

[51 What is the lowest testing rate atwhich the rate of positive tests will notincrease over the current positive testrate and how is it determined?

Random testing also helps to identifyand remove drug users from safety-sensitive positions. Sanctions againstdetected drug users serve as effectivedeterrents for others. With any givenprevalence of drug use, the higher therandom testing rate, the higher thenumber of drug users who will bedetected. At present, DOT's regulationscall for each covered employee to besubject to random testing at the samerate as all other covered employees ofthat employer. Testing must be based onrandom selection, and employers arenot permitted to target individuals or

employee groups for selection forostensibly "random" testing.

To the extent that detection is animportant goal, lowering the randomtesting rate reduces the percentage ofthe workers tested over a given periodand, therefore, reduces the number ofdrug users that would be identified. Forexample, if one assumes four testingperiods per year and testing takes placeover three years, then the followingfractions of a workforce would be testedat least once during the three years atthe following different annual testingrates.

Percent ofAnnual testing rate (percent) woddorcetested in 3

years

100 ...................................................... 96 .875 ...................................................... 91.750 ...................................................... 79.925 ...................................................... 53.910 ...................................................... 26.2

Available DataThe Department would appreciate

public comment in identifying dataconcerning the effectiveness of randomtesting rates. The following summarizesthe data currently available to theDepartment concerning the results ofrandom testing in the regulatedindustries, the Department's civilianworkforce, and the U.S. Coast Guarduniformed service.

AVIATION

1990 1991

Total number of randomtests ............................... 84,481 169.240

Number of postives ............ 446 1,232Percent positive.................. 0.53 0.73

Data on 1991 . 00 kwcun of oontOwdso andson. uetg at , 50 peroaK rate.

RAILROADS

Percent

1987 1988 189 1990 1991

5.1 5.6 3.0 3.0 11

For similar purposes, the positiverates for reasonable cause testing areprovided; they are as follows:

Percent

1987 1988 1989 1990 1991

5.4 4.7 3.6 1.8 1.9

FRA is currently conducting anexperimental study of eight railroads,comparing testing results at the 25percent and 50 percent random testingrates. The study began in July 1991 andis expected to be completed in July1993. Data from the first year of thestudy are now being analyzed and willbe released to the public shortly. Initialindications are that the FRA data do notshow an increase in positive rates forthe railroads conducting random testingat the 25 percent rate. However, all ofthese railroads had implemented testingat a 50 percent rate for some period andthen lowered it for the experimentalprogram.

Motor Carriers

In audits of 8,384 motor carrier drugtesting programs by FHWA in FY 1992,records indicated that 13,612 randomtests were conducted. There were 289verified positive results (2.12 percent).The audits represent less than 5 percentof motor carriers subject to the FHWArule. The FHWA selects motor carriersfor these general complianceinvestigations by determining factorssuch as a safety rating or priorcompliance problem. These complianceinvestigations do not offer scientific,statistically unbiased samplingmethods.

19 I 199 US. DOT EmD1ovees

Total number of randomtests ................................. 35228 50,436

Number of positives ............ 365 447Percent positive................. 1.04 0.88

FRA's random testing regulationswere issued in November 1988, with thefirst testing, as noted earlier, starting inJanuary 1990. FRA has kept records ofpost-accident drug testing for the lastfive years. For purposes of analyzingany effect from the issuance of therequirement or the implementation ofthe testing, the positive rates for post-accident testing are provided; they areas follows:

In the Department's Federal employeetesting program, the random testing rateof at least 50 percent was phased-in overthe first year of the program andachievedat the end of FY 1988. Atesting rate of at least 50 percent wasmaintained in FY 1989-1991. In FY1992, the figures reflect testing over thefirst five months with a rate of at least50 percent, followed by seven months oftesting with a rate of at least 25 percent.(As noted earlier, DOT decided to lowerits testing rate in 1992.) The followingtable summarizes DOT Federalemployee random testing data.

59780 , Federal Register / Vol.

Federal Register i Vol. 57, No. 241 / Tuesday. December 15, 1992 / Proposed RaIes

FY88 FY0 FY0 FY91 FY92

Total nu..mber of randorn est .... .... .... ........................................... .. .................................. ............ 5,047 17,926 1%,103 18,671 12Z454

Number of positives ............................................. . ................................. . ............................................... 42 92 43 401 39Percent posti ......................................................................................................................................... 0.83 0.51 0.23 0.21 0.31

As noted earlier, the USCG has been testing rate as a requirement at the appropriated. The percentage of positiveconducting random drug tests on its beginning of the fiscal year. the USCG results for random tests in each fiscalactive duty and reserve uniformed conducts the maximum number of tests year and the approximate testing ratepersonnel. Rather than setting a specific possible from the funds that are was as follows:

Percent 1067 1N 1 190 10 101 1062

Percent positive ............................................................ .............. . . . ............................. 1.57 1.31 0.68 0.41 0.41 0.78Testing rate .................................... .............................................................................. 120 951 95 951 8, 5

Testing Rates in Various FederalAgencies

Executive Order 12564, "Drug-FreeFederal Workplace", signed byPresident Reagan in September 1986,required random drug testing of safety-and security-related Federal employeesin 135 Federal executive branchagencies. According to a 1991 report ofthe General Accounting Office("Employee Drug Testing: Status ofFederal Agencies' Programs: Report tothe Chairman, Committee onGovernmental Affairs," U.S. Senate,(May 1991). GAO/FFD-91-70. 14-19,)there is a wide variation in the randomdrug testing rates at the 18 agencies thatGAO reviewed.

AlternativesThere are a number of alternatives to

the current 50 percent random testingrate that DOT could consider. Theyinclude:

(1) Making an across-the-boardmodification of the rate for all DOT anti-drug programs;

(2) Modifying how the random testingrate is implemented (e.g., frequency oftesting, etc.)

(3) Making a selective modification ofthe rate by

(a) Operating administration (e.g.,FAA or FRA could modify its rate);

(b) Job category (e.g., pilots, trainengineers);

(c Any other category that warranteda different rate based on drug useprevalence or other factors (e.g., age orgeographic region);

(4) Establishing a performancestandard program (as discussed later);

(5) Permitting employers who takespecified additional steps to deter druguse to reduce their random testing rate;

(6) Modifying the random testing ratefor all operating administration rides fora specific time period. subject toreconsideration after the results amanalyzed;

(7) Conducting demonstrationprograms in each operatingadministration before further action istaken; or

(8) Combining some of thealternatives.

[6] DOT requests comment on thesealternatives and information on anyother possible alternatives.

Modifying the random testing rateacross-the-board in each operatingadministration would result in thesimplest rule and program changes andwould be the easiest to enforce.Lowering the rate could reduce costssignificantly with an undeterminedimpact on deterrence, but it wouldunavoidably result in less detection ofdrug users. Some argue that what dataDOT has show very low positive ratesand, therefore, the random testing ratecan be reduced.

[7] Is the rate of positive test resultslow because the 50 percent testing rateis the minimum effective deterrent?

[81 Does a low rate of positives alonewarrant a change? Does this make therate "acceptable" for safety purposes?

[9] Is detecting fewer users acceptableeven though the positive rate is thesame?

Changes in how random testing isimplemented in the workplace may beanother approach to maintainingdeterrence at a lower testing rate. Forexample, increasing the number of timesrandom testing is conducted during theyear may serve to increase awarenessand visibility of the testing program.thus maintaining effective deterrenceeven though the actual number ofemployees tested is decreased.Procedures that ensure each personrandomly selected is actualy tested(e.g., if a selected employee is notavailable, the name would be keptconfidential until the employee isavailab for testing) as opposed torandom testing prooedures that allowselected indivkuals to "be excused"because of tempoary unevailebibty.

may also serve to maintain effectivedeterrence by making it more difficultfor drug users to avoid detection. Apolicy that requires a re-collection of aspecimen from a selected individualwhen a random test has been cancelledor the results invalidated would' helpdeter individuals from adulterating theirspecamens or otherwise obstructing the,testingprocess.

Authorizing variable random ratesamong operating administrations orwithin a workplace is another option toconsider in improving deterrence anddetection while reducing the overallrandom rate. Employers could testcertain categories of employees at higherrates based on prevalence or incidencedifferences in the-population. Forexample, if prevalence of drug use isgreater among mechanics than amongflight attendants, the employer couldtest mechanics at a 50 percent rate andflight attendants at a 20 percent rate.Variable rates enable deterrence anddetection efforts to be targeted wheredrug use is most likely. A variable ratestrategy, however, should be based onprevalence and incidence data, notindividualized suspicion or personalcharacteristics. Such an approach wouldrequire employers to maintain severalrandom pools and systematicallydetermine drug use prevalence andincidence data, something especiallydifficult for small firms. This approachmay present compliance andenforcement difficulties.

Another of the options DOT isexploring is use of a "performancestandard" for establishing randomtesting rates for a specified group suchas an industry, a job category, or anindividual employer. The group'soverall rate of testing would bedetermined by the success of the groupin deterring drug use as measured by itsrate of positive random tests. Forexample, if a group's positive rate is lessthan one peroest ver a givea period oftime, the group could be permitted to

5iw781

59782 Federal Register / Vol. 57, No. 241 / Tuesday, December 15, 1992 / Proposed Rules

reduce the random testing rate (e.g., to25 percent or 10 percent.) If the positiverate increased under a reduced randomtesting rate, the group would berequired to return to a higher randomtesting rate. The advantage of thisapproach Is that it could lower costs tolarge segments of the transportationindustry, if groups could maintain lowpositive rates. It would also provide anincentive to achieve low positive rates,the ultimate objective of DOT's program,and would reward those groups witheffective programs. Such an approachwould also encourage such groups touse whatever additional steps other thantesting that are appropriate in theirsituation, perhaps resulting in moreeffective programs. For example, somegroups may find increased educationhelpful in increasing the effectiveness oftheir anti-drug programs.

There are many potential obstacles toimplementing such an approach. Itwould require collection of additionaltesting data from each group that mightoffset savings in testing costs. Althoughthis may not be a difficulty for someoperating administrations, it may bemore problematic in others. In the motorcarrier industry, for example, even if thereports could be submitted, it is notclear that FHWA could effectivelyreview and respond to reports fromhundreds of thousands of motor carrierson millions of employees. From theDepartment's point of view, aperformance standard program wouldbe harder to administer because theoperating administrations would have tomonitor both positive rates and randomtesting rates for every affected group toensure proper compliance. In addition,it may be difficult to effectively apply alow random testing rate in smallcompanies. (For example, if a companywith one or two employees had to testat only 10 percent, employees might notbe deterred if they knew the next testmight not be for five or six years-andDOT would not get any reports todetermine whether the rate of positivetest results was being kept low.) Finally,there would be less detection of drugusers at a lower rate. The Department isalso concerned with whether this typeof performance standard would be anincentive to cheating. Employers mightbe tempted to falsify test results thatwere positive in order to maintain areduced random testing rate.

As another approach, the Departmentcould consider reducing the randomtesting rate but requiring additionalsteps such as increased education. Or, itcould offer a lower rate to thoseemployers that took extra steps toincrease deterrence.

Another approach would be toconduct a demonstration program ineach operating administration. FRA iscurrently in the middle of such aprogram. These studies could includepilot projects that explore positive ratesin relationship to different randomtesting rates. Demonstration programsthat include other deterrent orprevention initiatives in addition torandom testing could explore therelationship among education, trainingand awareness strategies and therandom testing rate. Pilot studies havethe advantage of examining specificvariables in relationship to deterrenceand detection outcomes in a controlledstudy environment.

As discussed earlier, there are manyalternative methods in addition totesting to deter illegal drug use,including education.

110] Which methods do commentersbelieve to be the most effective? What isthe basis for your belief?

[1.1] Should the Department mandateuse of one or more of these methods inaddition to current requirements in theanti-drug rules, or should it leave it upto the discretion of employers as asupplement to a testing program? Why?

In an attempt to explore the relevantresearch, random testingimplementation strategies, and drugdeterrence methodologies, to help it inthis rulemaking, the Department willconvene a conference on workplacerandom testing and drug deterrence.The conference will be open to thepublic. Details on the conferenceagenda, date and location will be offeredin a subsequent announcement.Information gathered from theconference will be considered alongwith comments received on thisANPRM in determining the nextappropriate action.

Additional QuestionsThe Department would appreciate

commenters responding to the followingadditional questions and providingwhatever data or studies might supporttheir position.

[12]Should other factors apply to anydecision to adjust the random testingrate besides deterrence, detection, andcosts? What are they and why are theyrelevant?

[131 What data would justify achange? Are low positive rates alonesufficient? How do we determine ifpositive rates are low because the 50percent testing rate is acting as adeterrent?

[14] What is the effect of the qualityof the selection and collection processon the positive rates that are currentlybeing produced? As the selection and

collection processes improve withexperience or enforcement, will thepositive rates be higher? How shouldany potential for increased detectionwith improved collection affect anydecision concerning the random testingrate?

[15] Can performance-based programsbe developed, operated, and auditedwithin the currently operating consortiaprograms? If so, how? Does participationin a consortium exacerbate the problemsassociated with implementation andenforcement of a performance standardfor random rates?

[16] If analysis of positive rates isnecessary before a decision can be madeto reduce the random testing rate, forhow long a period must DOT have data?What should DOT do about thoseindustries for which it currently haslittle or no data? Should operatingadministrations that have reliable,accurate data consider taking actionnow? Elsewhere in today's FederalRegister we are proposing extensivedata collection for all six operatingadministrations. This data would not,however, be available for some time.

[17] What if positive rates were toincrease after DOT lowered the randomtesting rate? Is some increase in the rateof positive tests acceptable? How muchand why? Are there other reasons therate could rise beside the lower randomtest rate?

[18] What cost reductions could beexpected if the random testing rate isreduced to 25 percent? 10 percent?What is the basis for these estimates?Even if some increase in positive ratesis expected with a lower random testingrate, is a reduction in the random testingrate nevertheless justified based onreduced costs? Please explain.

[19] Are there offsetting cost increasesthat could occur at a reduced randomtesting rate (e.g., consortium cost pertest, more accidents)?

[20] If a change is made, should itapply equally to all of the DOToperating administrations' programs, oronly as justified by operatingadministration specific data? Ifparticular industries, segments ofindustries or particular employersachieve very low positive rates, shouldthey be allowed to have different testingrates? Could this be effectivelyimplemented? Should random testingrates be linked to drug use in particulargroups based on prevalence orincidence data available for such groups(i.e., age groups, occupationalcategories, specific types of worksites?)

[21] What additional means ofincreasing deterrence are available tothe government and industiy tosupplement random drug testing? How

Federal Register / Vol. 57, No. 241 / Tuesday, December 15, 1992 / Proposed Rules

would these strategies work? Whatwould they cost? What is the evidencethey work? -

[22] How might DOT restructure itsrequirements for drug use preventionprograms to maximize the efficiencyand effectiveness of prevention withinthe transportation industry?

[231 Could a "performance standard"work? Would there be too muchincentive for some groups to cheat? Howcould DOT effectively implement orenforce such a standard, especially formotor carriers? Is the data collectionthat would be required more trouble andexpense than conducting a higher rate ofrandom tests? Would the trouble andexpense vary by operatingadministration? Please comment on thedifficulties in administering such aprogram and the increase inenforcement oversight that would berequired. How could it effectively beimplemented for small companies?Would it work if implemented Industry-by-industry or among segments of agiven industry? Could it be effectivelyimplemented and enforced employer-by-employer?

[24] Any reduction in the randomtesting rate will result in less detection.The lower the rate, the less thedetection. Are there alternativeapproaches apart from random testingthat could offset potential reductions indetection? What is their projectedefficiency and effectiveness, comparedwith current programs? What wouldthey cost? What is the basis of thenumbers? Are there safety implications?

1251 Are there any additional data orstudies concerning rates of drug testingand deterrence that may be relevant tothe Department's consideration?

[26] Should each operatingadministration conduct a demonstrationprogram to gather data on the relativeeffectiveness of different rates specificto its regulated industry? What outcomemeasures would be appropriate in suchpilot programs?

[27] Should one or more operatingadministrations reduce the randomtesting rate for a specific time periodand then analyze the results? Whatperiod of time is adequate fordetermining the impact on deterrence?

Regulatory Analyses and NoticeDOT Regulatory Policies and Procedures

The ANPRM is considered to be asignificant rulemaking under DOTRegulatory Policies and Procedures, 44FR 11034, because of the substantialpublic and Congressional interest in thissubject. It is difficult to estimatepotential costs or benefits at this timebecause the Department is notproposing specific options.

Executive Order 12612This ANPRM has been analyzed in

accordance with the principles andcriteria contained in Executive Order12612, and It has been determined thatit does not have sufficient federalismimplications to warrant the preparationof a Federalism Assessment.

Executive Order 12630This ANPRM has been analyzed in

accordance with the principles andcriteria contained in Executive Order12630, and it has been determined thatany potential modification in the

random drug testing program does notpose the risk of a taking ofconstitutionally protected privateproperty.

Regulatory Flexibility ActDepending on what, if any, action is

ultimately proposed and adopted, apotential modification in the randomdrug testing program could have asignificant economic impact on asubstantial number of small entities.The Department specifically seekspublic comment on the effect, if any, ofpotential changes in the program onsmall entities as well as any suggestedalternative approaches. Further reviewwill be conducted based on commentsreceived on this notice and when, andif, a notice of proposed rulemaking isissued.

Paperwork Reduction ActThere are a number of reporting or

recordkeeping requirements associatedwith DOT-mandated drug testing.Because the purpose of this notice is tosolicit information, no specific changesare being proposed at this time. TheDepartment notes that effectiveimplementation of possible alternatives

and the gathering of data necessary tojustify any changes, comparealternatives, or permit theimplementation of some approachesmay require recordkeeping andreporting by the affected industries. Weare proposing substantial data collectionconcerning drug testing elsewhere intoday's Federal Register.

National Environmental Policy Act

The Department has determined thatthis rulemaking is not a major Federalaction significantly affecting the qualityof the human environment and that anenvironmental impact statement is notrequired.

Authority: FAA-49 U.S.C. App. 1354(a),1355, 1356, 1357, 1401, 1421-1430, 1472,

1485,1502; 49 U.S.C. 106(g) revised, Pub. L.97-449, January 12, 1983.

FHWA--49 U.S.C 504 and 3102; 49 CFR1.48.

FRA- 45 U.S.C. 431,437, 438, as amended;Pub. L 100-342, Pub. L 102-143; and 49CFR 1.49(m).

FTA-49 CFR 1.51, Pub. L 102-143.RSPA-49 App. U.S.C. 1672, 1674a, 1681,

1804, 1808, 2002, 2040; 49 CFR 1.53.USCC--46 U.S.C. 2103, 3306, 7101, 7301,

7701; 49 CFR 1.46.

Multi-modal ANPRM on RandomDrug Testing Program.

Issued on December 2, 1992, inWashington, DC.Andrew H. Card, Jr.,Secretary.Thomas C. Richards,Administrator, Federal AviationAdministration.Thomas D. Larson,Administrator, Federal HighwayAdministration.Gilbert E. Carmichael,Administrator, Federal RailroadAdministration.Brian W. Clymer,Administrator, Federal TransitAdministration.Douglas B. Ham,Acting Administrator, Research and SpecialPrograms Administration.Admiral J. William Kim.,Commandant, United States Coast Guard.[FR Dc. 92-29690 Filed 12-10-92; 10:00am)ILtJW CODE *1-42-61

59783