ERIC EJ823785: Attendance Rates in a Workplace Predict Subsequent Outcome of Employment-Based Reinforcement of Cocaine Abstinence in Methadone Patients

2008·0.28 MB·English

by  ERIC

JOURNALOFAPPLIEDBEHAVIORANALYSIS 2008, 41, 499–516 NUMBER4 (WINTER2008)
ATTENDANCE RATES IN A WORKPLACE PREDICT SUBSEQUENT
OUTCOME OF EMPLOYMENT-BASED REINFORCEMENT OF
COCAINE ABSTINENCE IN METHADONE PATIENTS
WENDY D. DONLIN, TODD W. KNEALING, MICK NEEDHAM,
CONRAD J. WONG, AND KENNETH SILVERMAN
JOHNSHOPKINSUNIVERSITYSCHOOLOFMEDICINE
This study assessed whether attendance rates in a workplace predicted subsequent outcome of
employment-based reinforcement of cocaine abstinence. Unemployed adults in Baltimore
methadone programs who used cocaine (N 5 111) could work in a workplace for 4 hr every
weekday and earn $10.00 per hour in vouchers for 26 weeks. During an induction period,
participantsprovidedurinesamplesbutcouldworkindependentoftheirurinalysisresults.After
the induction period, participants had to provide urinalysis evidence of cocaine abstinence to
workandmaintainmaximumpay.Amultipleregressionanalysisshowedthatinductionperiod
attendancewasindependentlyassociatedwithurinalysisevidenceofcocaineabstinenceunderthe
employment-based abstinence reinforcement contingency. Induction period attendance may
measure the reinforcing valueof employment and could be used toguide the improvement of
employment-based abstinence reinforcement.
DESCRIPTORS: abstinence reinforcement, cocaine addiction, contingency management,
employment,methadone
_______________________________________________________________________________
Research in both the laboratory (Higgins et nitude increases. Although low-magnitude ab-
al., 2007; Nader & Woolverton, 1991) and the stinence reinforcement can have considerable
clinic (Dallery, Silverman, Chutuape, Bigelow, clinical utility (Peirce et al., 2006), the
& Stitzer, 2001; Higgins et al., 2007; Silver- substantial body of research on reinforcement
man, Chutuape, Bigelow, & Stitzer, 1999; magnitude suggests that for any population of
Stitzer & Bigelow, 1984) has shown that the drug abuse patients, the proportion of patients
effectiveness of abstinence reinforcement con- that respond to an abstinence reinforcement
tingencies increases as the reinforcement mag- intervention by initiating abstinence should
increase as the magnitude of reinforcement
Wendy D. Donlin is now at the Department of increases. Identifying practical ways to arrange
Psychology,UniversityofNorthCarolinaatWilmington; high-magnitude abstinence reinforcement con-
Todd W. Knealing is now at the Department of
tingencies is an important challenge for clini-
Psychology, Briar Cliff University, Sioux City, Iowa; and
ConradJ.WongisnowatEliLilly,NeuroscienceMedical cians and clinical researchers.
Research,Indianapolis, Indiana. To develop potentially practical ways of
This research was supported by Grants R01DA13107
arranging high-magnitude abstinence reinforce-
and T32-DA07209 from the National Institute on Drug
Abuse,Rockville, Maryland. ment contingencies, some researchers have
We thank Jacqueline Hampton for recruiting partici- sought to identify high-magnitude reinforcers
pantsandconductingintakeassessments, CarolynCarpe-
that are routinely dispensed in the community
nedo and Haley Brown for staffing the workrooms and
urinalysis laboratory, and Kylene Broadwater and Karly and that might be harnessed for therapeutic
Diemer for serving as study coordinators for the project purposes (Donlin, Knealing, & Silverman,
andforassisting inthe preparation of themanuscript.
2008). In this vein, a few researchers have
AddresscorrespondencetoKennethSilverman,Depart-
ment of Psychiatry and Behavioral Sciences, Johns explored the possibility of using contingent
Hopkins University School of Medicine, 5200 Eastern access to employment to reinforce drug absti-
Avenue, Suite 142 West, Baltimore, Maryland 21224 (e-
nence (Cohen, Bigelow, Hargett, Allen, &
mail:[email protected]).
doi:10.1901/jaba.2008.41-499 Halsted, 1973; Crowley, 1986; Milby et al.,
499
500 WENDY D. DONLIN et al.
1996; Miller, 1975; Silverman, Svikis, Robles, reinforcement, Silverman et al. (2007) used a
Stitzer, & Bigelow, 2001). Employment set- workplace induction procedure in which partic-
tings have several features that might make ipants were hired and allowed to work in the
them ideal vehicles for arranging abstinence workplace for about a month prior to initiation
reinforcement contingencies (Silverman, 2004), of an employment-based abstinence reinforce-
but one of their most important features is that ment contingency. This workplace induction
they can control high-magnitude reinforcers, period was designed to allow participants to
most notably wages, that can be used to samplethereinforcersavailableintheworkplace
reinforce drug abstinence. Employment-based (Kazdin, 1982) and to allow them to establish
abstinence reinforcement can be arranged by critical work-related behaviors (e.g., following a
hiring persistent drug users and requiring them dailyroutineofwakingupatasettime,traveling
to provide drug-free urine samples to gain and totheworkplace,workingontrainingprograms,
maintain access to the workplace. earningvouchers,andexchangingearnedvouch-
A systematic program of research over the ers) prior to requiring that participants provide
past several years has shown that employment evidence of recent cocaine abstinence to work.
can be a useful vehicle for arranging high- The induction period also affords the
magnitude abstinence reinforcement (Silver- opportunity to assess individual differences in
man, 2004; Silverman et al., 2001, 2002, workplace attendance before arranging an
2007). That research, conducted primarily with employment-based abstinence reinforcement
methadone patients who continued to use contingency. This measure of workplace atten-
cocaine and heroin despite exposure to routine dance could be important because it may
counseling and daily methadone maintenance, predict the extent to which contingent access
has shown that employment-based reinforce- to the workplace will serve as a reinforcer for
ment can be effective in promoting abstinence individual participants. This expectation is
from cocaine and opiates in many patients based on research that has shown that the rate
(Silverman et al., 2001, 2002, 2007). In these at which an organism contacts a stimulus when
studies, participants were hired to work in a that stimulus is freely available is proportional
model therapeutic workplace where they could to the magnitude of the reinforcement effect
work 3 to 6 hr per day and earn between produced when that stimulus is presented
$35.00 and $60.00 per day for many months. contingent on some low-rate behavior (e.g.,
To promote abstinence, participants in the Premack, 1961, 1963).
therapeutic workplace were required to provide This principle was illustrated in an early
urinalysis evidence of drug abstinence to gain study that examined the reinforcement of lever
and maintain daily access to the workplace. pressing in rats with either contingent access to
Despitethehigh-magnitudereinforcementused a drinking tube or a running wheel (Premack,
in these studies, not all participants initiated 1963). In that study, rats were monitored in
abstinence when exposed to employment-based free-access assessment sessions in which they
abstinence reinforcement (Silverman et al., were given free access either to a running wheel
2001, 2002, 2007); one study even failed to (in morning sessions) or to a drinking tube that
demonstrate the overall effectiveness of an dispensed a sucrose solution (in afternoon
employment-based abstinence reinforcement sessions). The point of the free-access assess-
intervention(Knealing,Wong,Diemer,Hamp- mentsessionswastoassessthepercentageofthe
ton, & Silverman, 2006). sessions that the rats spent engaging in the
In an effort to increase the proportion of available behavior (running or drinking) when
participants who respond to employment-based each activity was freely available. The probabil-
PREDICTING EMPLOYMENT-BASED REINFORCEMENT 501
ities of running in the free-access running uate the effectiveness of employment-based
sessions and of drinking in free-access drinking abstinence reinforcement as a maintenance
sessions were varied by manipulating the force intervention to sustain abstinence over an
required to operate the wheel (low and high extended period of time. All participants in
force) and the sucrose concentration in the the larger trial were invited initially to attend
drinking tube (low, medium, and high). After the workplace for 6 months. Individuals who
the assessment sessions, conditioning sessions became abstinent from cocaine and opiates and
were arranged in which three presses on a lever acquired a minimal set of job and professional
produced 15-s access to running (again in skills during that initial 6-month period were
morning sessions) and separate conditioning invited to attend the workplace for an addi-
sessions in which lever presses produced 15-s tional year and participate in a randomized
access to the drinking tube (again in afternoon controlled study. This report describes data
sessions). The study showed that the amount of collected during the initial 6-month period of
lever pressing in the conditioning sessions was the larger study. The results of the randomized
directly related to the free-access probability of controlled trial will be published in a later
thecontingentresponseasmeasuredinthefree- article.
access assessment sessions.
The workplace induction period is analogous Settings and Materials
totheperiodinPremack’s(1963)studyinwhich The current study was conducted in the
heallowedfreeaccesstothedrinkingtubeorthe therapeutic workplace at the Center for Learn-
running wheel and allows us to measure the ingandHealth,atreatment-research unitatthe
extenttowhichparticipantscontactorattendthe Johns Hopkins Bayview Medical Center. De-
workplacewhenaccesstotheworkplaceisfreely tails of the therapeutic workplace setting,
available. The results obtained by Premack materials, and procedures are similar to proce-
suggest that the effectiveness of an employ- duresusedbySilvermanetal.(2007).Themost
ment-basedabstinencereinforcementcontingen- important procedures for this study, as well as
cy, when access to the workplace is made procedures that differ from those employed in
contingent on the provision of cocaine-negative the prior study, are described below.
urine samples, should be directly related to the On workdays, participants reported to the
rates that participants attended the workplace urinalysis laboratory in the therapeutic work-
during the induction period. place,wheretheysignedintotheworkplaceand
This study had two objectives. The primary provided observed urine samples when re-
objective was to determine if workplace atten- quired. The laboratory was near a central area,
dance during a workplace induction period where up to three workroom assistants moni-
predicted abstinence outcomes in a subsequent tored the activities of the participants in
condition under an employment-based absti- workrooms. Participants passed through this
nence reinforcement contingency. The second centralized area to enter one of three work-
purpose was to provide a systematic replication rooms, where all training was conducted.
of the effectiveness of employment-based rein-
Recruitment and Participant Selection
forcement of drug abstinence in methadone
patients (Silverman et al., 2007). The Western Institutional Review Board
approved this study. Participants were enrolled
from October, 2003, to March, 2006. The
METHOD
study was advertised at 16 methadone clinics in
Thisstudywasconductedasapartofalarger Baltimore. Research staff visited these clinics,
randomized controlled study designed to eval- and flyers and letters were distributed within
502 WENDY D. DONLIN et al.
these clinics to invite unemployed adults in criteria for cocaine dependence (American
methadone treatment to apply to enroll in a Psychiatric Association, 1994), were currently
study that provided monetary vouchers for receiving welfare benefits in Baltimore, and
participating in job skills training. achieved 80% correct on the reading assess-
WhentheycontactedtheCenterforLearning ment. Participants were excluded if they were
and Health, individuals were asked to complete considered to be at risk of suicide (reported on
a brief screening interview containing eight the Addiction Severity Index), reported audito-
questions to determine study eligibility. Ques- ry or visual hallucinations (reported on the
tions asked age, marital status, zip code of Addiction Severity Index), had a physical
residence, employment status, drug use, treat- limitation that prevented typing, were incarcer-
ment enrollment, income, welfare status, and ated in a halfway house or under constant
major medical conditions. Some of the ques- monitoring, or if they earned more than
tionswereunrelatedtostudyeligibilitybutwere $200.00 in the last month in under-the-table
included to obscure inclusion criteria. Individ- (unreported) income.
uals indicating that they were 18 years of age or
older, were unemployed, used cocaine or crack Experimental Design
within the last 30 days, were receiving welfare Eligible participants were invited to attend
benefits, and were enrolled in methadone the workplace for 6 months. Urine samples
maintenance in Baltimore were invited to the were collected every Monday, Wednesday, and
center to participate in a full screening Friday and were tested for cocaine and opiates.
interview. The study employed a multiple baseline across
drugs design to evaluate the effectiveness of
Full Screening Interview employment-based abstinence reinforcement
The full screening interview included a in promoting abstinence from cocaine and
breath sample to measure alcohol use (not then opiates. After a workplace induction
reported here); a urine sample to test for drug period in which participants could attend the
use; and assessments to characterize drug workplace independent of their urinalysis
history, psychological status, and academic results, employment-based reinforcement of
skills. Observed urine samples were tested for cocaine abstinence was arranged. After a
cocaine, opiates, benzodiazepines, methadone, participant achieved 3 weeks of sustained
andamphetamines.Researchstaffalsofilledout cocaine abstinence, employment-based rein-
an assessment on the physical limitations of the forcement of cocaine and opiate abstinence
individual to exclude those who could not was arranged.
operate a computer keyboard. Additional Workplace induction period. During the
assessmentswerecompletedtoascertaincocaine workplace induction period, participants could
and heroin dependence (Cottler, 1991), psy- attend the workplace independent of their
chosocial functioning (McLellan et al., 1985), urinalysis results and could earn vouchers for
HIV risk behaviors (Navaline et al., 1994), and their attendance in the workroom and for their
reading skill levels (Wilkinson, 1993). Partici- performance on training programs. The vouch-
pants were paid $30.00 in vouchers for er system was adapted from a voucher system
completing the full interview. originally developed for the treatment of
Study eligibility required that participants primary cocaine dependence patients (Higgins
were at least 18 years old, were currently et al., 1991). Vouchers have monetary values
unemployed, were enrolled in a Baltimore City and can be exchanged for goods and services
methadone maintenance program, provided a that are purchased for participants by staff.
cocaine-positive urine sample at intake, met Vouchers are used instead of cash to reduce the
PREDICTING EMPLOYMENT-BASED REINFORCEMENT 503
chance that participants will use their earnings abstinence.Inaddition,basepaywasreducedto
to purchase drugs. $1.00perhourforthe1stdayworkedfollowing
Participants could earn a base pay of $8.00 a positive urine sample. Hourly base pay then
per hour for the time they spent in their increased by $1.00 per hour for each day that
assigned workrooms and approximately $2.00 the participant provided a cocaine-negative
per hour in productivity pay for working on urine sample and worked at least 5 min, until
their training programs. Participants earned the value reached the original $8.00 per hour.
5 min of paid break for every 55 min worked. Opiate and cocaine abstinence contingency.
Base pay and productivity pay accumulated in Afterthesuccessfulcompletionof3consecutive
the participant’s account as an electronic weeks of cocaine abstinence during the absti-
voucher until he or she chose to exchange the nence contingency, participants were moved to
pay for gift cards or other approved goods and a cocaine and opiate abstinence contingency.
services. The workplace induction period con- Under this condition, participants had to
tinued for at least 4 weeks and until the provide urinalysis evidence of recent abstinence
participant attended the workplace for at least from both cocaine and opiates on mandatory
5 min on 15 workdays. The requirement that urine days to gain access to the workplace and
participants attend the workplace on at least 15 to maintain maximum pay. Samples indicated
dayswas arrangedto ensure that allparticipants cocaine and opiate abstinence if the urinary
had a minimal amount of exposure to the concentration of the drug’s metabolite (benzoy-
workplace before being exposed to the employ- lecgonine and morphine, respectively) was #
ment-based abstinence reinforcement contin- 300 ng/ml or if the urinary metabolite concen-
gency. tration decreased by at least 20% per day since
Cocaine abstinence contingency. At the end of the last sample was submitted. If the sample
the induction period, participants were exposed failed to meet these criteria, workplace access
to an employment-based abstinence reinforce- was denied for that day, and base pay was
ment contingency in which they were required decreased as previously described. Workplace
toprovideevidenceofrecentcocaineabstinence access could be granted the following workday
to work and to maintain maximum pay. Under or any day thereafter by providing urinalysis
thiscontingency,mandatoryurinesampleswere evidence of recent cocaine and opiate absti-
collected every Monday, Wednesday, and nence.
Friday, and participants were allowed access to
the workplace only if the urine samples General Workplace Procedure
indicated recent cocaine abstinence. Recent The therapeutic workplace employed stan-
cocaine abstinence was assumed if a partici- dard procedures that were applied with all
pant’surinarybenzoylecgonineconcentration(a participants. It was operated through a Web-
metabolite of cocaine) was # 300 ng/ml or if based therapeutic workplace application pro-
the urinary benzoylecgonine concentration gram (Silverman et al., 2005, 2007). Partici-
decreased by at least 20% per day since the last pants could attend the workplace from 10:00
sample was submitted. a.m. to 12:00 p.m. and 1:00 p.m. to 3:00 p.m.
Under the cocaine abstinence contingency, a every weekday except holidays and weather
participant who submitted a cocaine-positive emergencies.
urine sample was denied access to work on that On mandatory urine days, participants were
day. Following a positive urine sample, a required to submit samples in a bathroom in
participant could regain access to the workplace the presence of a same-sex research staff
only if he or she showed evidence of cocaine member using well-established procedures that
504 WENDY D. DONLIN et al.
were designed to ensure valid specimen collec- out the day, participants sat at their cubicles to
tion (Silverman et al., 2007). The sample was work on computerized typing, keypad, and
tested for opiates and cocaine using the Abbott data-entry training programs.
AxSYMsystem.Benzoylecgonineandmorphine Typing, keypad, and data-entry training
concentrations (the metabolites of cocaine and programs. Training programs were designed in
heroin, respectively) were quantified, using such a way that participants with no previous
serial dilution procedures if necessary. typing skills could become fluent typists (see
The therapeutic workplace software deter- Silverman et al., 2007, for detailed description
mined work eligibility and printed a urine of the trainingprograms). Briefly,theprograms
feedback graph. The feedback graph provided a were divided into small steps through which
graphical and numerical display of benzoylec- participants could progress sequentially. Partic-
gonine and morphine concentrations, indicated ipantscouldearnvouchersforcorrectresponses,
whether or not the results indicated that the lose vouchers for incorrect responses, and earn
participanthadusedopiatesorcocainerecently, bonuses for passing steps. The programs were
and indicated whether or not the participant designed so that participants could earn ap-
was granted access to the workplace. If all proximately $2.00 per hour in productivity pay
abstinence requirements were met or if no for their work on the training programs.
abstinence requirements were in effect, partic- Staff routinely monitored participant perfor-
ipantsweregiventheiridentification(ID)cards, mance on training programs. If participants
which they presented to a workroom staff were having trouble passing particular steps,
member in the common area of the workplace. their training programs were modified to
If abstinence requirements were not met, the attempt to improve performance.
participants were given their feedback graphs, Voucher system. Each participant’s computer
wereinformedthattheycouldnotworkonthat displayed a home page that continuously
day,wereinformedthattheirbasepaywouldbe tracked and updated earnings (an electronic
reset to $1.00 per hour, and were instructed on voucher).Theelectronicvoucherdisplayedbase
how and when their base pay could increase pay and productivity pay, hours worked, paid
again to the maximum level of $8.00 per hour. hours, accuracy, and earnings for the training
When participants had no abstinence require- programs.Theelectronicvouchercouldbeused
ments on a mandatory urine sample day (i.e., to purchase goods or services (gift cards, bills
during the induction period), the samples were paid, etc.). Voucher purchases had to be for the
testedlaterintheday,andfeedbackgraphswere participants or for the participants’ immediate
given to those individuals at their desks after family. Vouchers could not be used to purchase
testing. cigarettes or alcohol, and gift certificates could
General workroom procedure. Once partici- not be obtained from stores that returned more
pants received their ID cards, they presented than $5.00 in cash back. Many gift cards were
their IDs to a workroom assistant who signed stocked continuously and could be obtained on
them into the workroom. Throughout the the same day. Other purchases requested by the
workday, participants’ ID cards were swiped endofthedayonMondayorWednesdaycould
throughabarcodereadereachtimetheyentered be obtained by the next Wednesday or Friday,
or left the workroom. This system allowed respectively.
precise measurement of the duration of time On occasions when the therapeutic work-
that a participant spent in the workplace. All place was closed due to holidays or severe
participants had their own cubicle workstations weather emergencies, participants received
within one of the three workrooms. Through- workplace closing pay. The closing pay was
PREDICTING EMPLOYMENT-BASED REINFORCEMENT 505
the mean of the previous and following day’s contingency. This resulted in a precontingency
earnings, unless there was a reset on either day. (induction or baseline) percentage of negative
If their base pay was reset either day, partici- samples and postcontingency (intervention)
pants were ineligible to receive closing pay. percentage of negative samples for each indi-
Trainee instruction. At the beginning of the vidual. Each of the baseline and intervention
study and at every transition to different periods included 12 urine samples for cocaine
conditions, participants were given instructions and 10 urine samples for opiates. The opiate
detailing the procedures of the therapeutic baseline and intervention periods included only
workplace.Originalinstructionsexplainedwork 10 urine samplesbecause some participants had
rules, testing procedures, and the voucher only 10 urine sample days between the onset of
procedures. Transitional instructions provided the cocaine abstinence contingency and the
information about the changes in contingencies onset of the opiate abstinence contingency. The
(e.g., the introduction of the cocaine abstinence percentage of samples testing negative in the
contingency). Instructions were read out loud induction and intervention periods were com-
by a workroom assistant, and throughout the pared for both cocaine and opiates using a
instructions participants were required to an-
Wilcoxon signed ranks test and paired t tests.
swer multiple-choice questions. Participants
The Wilcoxon signed tanks tests were consid-
could earn $0.20 for every correct answer and
eredtheprimaryanalyses.Forallanalyses,urine
$0.10 for each corrected error. Quizzes were
samples were considered negative for cocaine
administered several days later to ensure
and opiates if benzoylecgonine and morphine
retentionandunderstandingoftheinstructions.
concentrations,respectively,were#300 ng/ml.
Each participant kept a copy of the instructions
To account for missing urine samples, two
at his or her workstation.
separate analyses were conducted that varied in
howvaluesformissingsampleswerederived.In
Standard Treatment Services
oneanalysis,missingurinesamplesweretreated
All participants were recruited from Balti-
as positive samples (missing positive). In the
more methadone clinics, where they received
other analysis, values for missing samples were
routine drug abuse counseling concurrent with
interpolatedbasedonthevaluesbeforeandafter
study participation. Participants reported mean
the missing sample (missing interpolated). In
methadonedosesof91 mg(SD525 mg),and
the missing interpolated analysis, a missing
allurinesampleswerepositiveformethadoneat
sample or missing group of samples was
study enrollment. Participants had access to
considered positive if a urine sample before or
referrals to community services throughout the
after the missing sample or missing group of
study and were offered referrals to employment
samples was positive.
services prior to being discharged from the
To determine if rates of attendance during
study.
the induction period predicted cocaine absti-
Data Analysis nence during the intervention period, correla-
Both visual inspection and statistical analyses tions and multiple regression analyses were
were used to evaluate the effects of the calculated. Data for the induction period were
contingencies. For each drug (cocaine and based on the first 4 weeks (20 days) of possible
opiates), dichotomous outcome measures (e.g., workplace attendance. Because baseline rates of
cocaine-positive vs. cocaine-negative urine sam- cocaine and opiate abstinence have been
ples) were aggregated within participant for associated with cocaine abstinence outcomes
approximately 1 month prior to and 1 month under voucher-based abstinence reinforcement
following the introduction of each abstinence (e.g., Silverman et al., 1998), we also assessed
506 WENDY D. DONLIN et al.
whether the percentage of cocaine-negative urine samples with positive results would
urinesamplesandopiate-negativeurinesamples increase the correlation between the induction-
in the induction period predicted the percent- period drug use measures and the measures of
age of cocaine negative samples in the inter- attendance.
vention period. The percentage of cocaine- Similaranalyseswereconductedtodetermineif
negative and opiate-negative samples were dailyearnings,likepercentageofminutesattend-
calculated by dividing submitted cocaine-nega- ed,duringtheinductionperiodpredictedcocaine
tive and opiate-negative urine samples, respec- abstinence during the intervention period. For
tively, by total scheduled urine samples during those analyses, induction pay per day was
the induction period. Percentage of minutes calculatedbydividingthetotalearningsforthe4
worked was calculated as total minutes worked weeks of the induction period by 20 days. Daily
during the first 4 weeks of the induction period earnings in the induction period were highly
divided by 4,800 min, the maximum possible correlatedwiththepercentageofminutesattended
numberofminutesinthatperiod(240 minper intheinductionperiod(r50.924,p,.001;P5
day times 20 days). The percentage of cocaine- 0.917,p,.001),andtheresultsofthemultiple
negative samples during the intervention period regressionanalysisweresimilartotheanalyseswith
encompassed all possible workdays after the thepercentageofminutesattended.Asaresult,the
datafordailyearningsintheinductionperiodasa
onset of the cocaine contingency and was
predictorwillnotbepresentedhere.
calculated by dividing the total number of
cocaine-negative urine samples by the number
of total scheduled urine sample collections. RESULTS
Both Spearman’s (r) and Pearson’s (P)
A total of 128 participants were enrolled in
correlation coefficients were calculated to assess
theworkplace;however,17ofthoseparticipants
the relation between each of the three induc-
did not continue attending long enough to be
tion-period measures (i.e., percentage of co-
exposed to the cocaine abstinence contingency.
caine-negative urine samples, percentage of
All participants who reached the cocaine
opiate-negative urine samples, and percentage
abstinence requirement (n 5 111) were includ-
of minutes worked) and the primary interven-
ed in the analysis to examine the effects of the
tion-period outcome measure (i.e., the percent-
cocaine abstinence reinforcement contingency
age of cocaine-negative samples in the interven-
oncocaineabstinence.Participantswhoreached
tion period). To determine if the percentage of
the opiate and cocaine abstinence requirement
minutes attended in the induction period was
(n 5 62) were included in the analysis to
independently associated with the percentage of
examine the effects of the opiate and cocaine
cocaine-negative urine samples in the interven-
abstinence contingency. Table 1 describes the
tion period, a multiple regression analysis was
characteristics assessed at intake of the 111
performed, with the percentage of minutes
individuals who were exposed at least to the
attended and the two induction-period drug
cocaineabstinencecontingency.Atintaketothe
use measures (i.e., percentage cocaine-negative
study, virtually all participants were living in
and percentage opiate-negative urine samples) poverty, most of their income came from
aspredictorvariables.Theoutcomemeasurefor welfare benefits, and much of their income
the multiple regression analysis was the per- was spent purchasing drugs.
centageofcocaine-negativeurinesamplesinthe
interventionperiod.Intheseanalyses,valuesfor Response to the Abstinence Contingencies
missing urine samples were interpolated as Participants showed very high rates of opiate
described above, because replacing missing abstinence during and after the induction
PREDICTING EMPLOYMENT-BASED REINFORCEMENT 507
Table 1
that contingency, we conducted secondary
Demographic Characteristics ofStudy Participants
analyses on two sequential cohorts of partici-
Assessed atIntake(N5 111)
pants within the total sample. The two cohorts
Age,M(SD),yearsa 43(6.4) were defined by the two major waves of
Femalea 77.5% recruitment that occurred in the study. The
Black/white/othera 92.8%/6.3%/0.9%
Marrieda 9.4% first cohort of participants was recruited from
HIVpositiveb 13.5% October 8, 2003, to August 11, 2004, and
HighschooldiplomaorGEDa 56.8%
Usuallyunemployedpast3yearsa 73.9% included67participants.Aftera5-monthbreak
Past30daysincome,M(SD)a in recruitment, the second cohort was recruited
Employment $5.60($25.80)
from January 10, 2005, to March 8, 2006, and
Unemployment $1.40($14.20)
Welfare(DSS) $346.90($189.90) included 44 participants. Data on the effects of
Pension,benefits,orsocialsecurity $32.17($128.24)
the cocaine abstinence reinforcement contin-
Mate,family,friends $45.54($128.56)
Illegal $9.32($60.70) gency were examined for the two cohorts to
Livinginpovertyc 94.5% provide a replication of the effect across two
Daysused,past30days,M(SD)a
Cocaine 18.1(10.6) different periods of calendar time.
Heroin 3.29(5.77) Figure 1 shows the percentage of the partic-
$spentondrugs,past30days,M(SD)a $294.95($478.53)
Illegalactivityformoney,past30daysa 1.2 ipants who provided cocaine-negative urine
Currentlyonparoleorprobationa 13.5% samples across the entire 6-month period for
Felonyconvictioninlifea 64.9%
Cocainedependenced 99.0% the two cohorts of participants and for the total
Opiatedependenced 97.3% sample. Because participants were not exposed
aFromthe Addiction SeverityIndex. to the cocaine contingency until they had been
bFromthe Risk Assessment Battery. in the study for at least 1 month and had
cBased on U.S. Census Bureau Poverty Thresholds attendedtheworkplaceonatleast15days,they
2003. Weighted average thresholds for one person under
were exposed to the contingency at different
65 years 5 $9,573 and 65 years and over 5 $8,825.
timepointsthroughoutthe6-monthstudy.The
Annual income per participant was calculated by taking
solid line represents the percentage of partici-
thesum ofthe employment, unemployment,welfare.
dFrom the Composite International Diagnostic Inter- pants cumulatively exposed to the cocaine
view. abstinence contingency. For each cohort and
for the total sample, the percentages of cocaine-
period, which precluded the possibility of negative urine samples were stable during the
demonstrating effects of the abstinence rein- induction period and increased roughly in
forcement contingency on opiate abstinence. proportion to the cumulative percentage of
Specifically,96%and97%oftheurinesamples participants exposed to the cocaine abstinence
were negative for opiates (based on the missing reinforcement contingency. Although the two
interpolated analysis) during the induction and cohortsofparticipantsshoweddifferentlevelsof
intervention periods, respectively. This com- cocaine abstinence during the induction period
promised our ability to demonstrate the effects (Cohort 1 had higher rates of cocaine-negative
of the employment-based abstinence reinforce- urine samples than Cohort 2), each cohort and
ment contingencies using a multiple baseline the total sample showed similar and large
design across drugs. To provide additional increasesinthepercentageofcocaineabstinence
evidenceof experimentalcontrol, weconducted with the onset of the cocaine abstinence
further analyses of the effects on cocaine contingency.
abstinence.Specifically,inadditiontoanalyzing Figure 2showsthepercentageofparticipants
the effects of cocaine abstinence reinforcement abstinent from cocaine for the 12 samples (1
on the total sample of participants exposed to month) prior to and the 12 samples following
508 WENDY D. DONLIN et al.
Figure2. Thepercentageofparticipantswhosubmit-
tedcocaine-negativeurinesamplesrelativetotheonsetof
the cocaine abstinence contingency. The filled circles
represent the percentage of participants from Cohort 1
(top,n 567),Cohort 2(middle, n544), andthe total
sample (bottom, n 5 111) who submitted a cocaine-
negative urine for the 12 samples prior to and the 12
samples following the onset of the cocaine contingency
(represented at the 0 point on the x axis). All missing
values werecoded aspositive.
Figure 1. Percentage of cocaine urine samples sub-
mitted over the entire 6-month experiment. The filled
the onset of the cocaine abstinence contingency
circles represent the percentage of participants who
submitted a cocaine-negative urine sample for each for the two cohorts and the total sample. For
consecutive urine sample for Cohort 1 (top, n 5 67), the month prior to the contingency, the
Cohort2(middle,n544),andthetotalsample(bottom, percentage of participants that tested negative
n 5 111). The solid line in each panel represents the
for cocaine was extremely stable for each cohort
percentage of participants cumulatively exposed to the
and for the total sample. For each cohort and
cocaine abstinence contingency. All missing values were
codedaspositive. for the total sample, the percentage of partic-
ipants that provided cocaine-negative urine
samples increased abruptly after the introduc-
tion of the cocaine abstinence contingency,