jaba-29-03-0305

5/24/2018 jaba-29-03-0305

1/15

305

JOURNAL OF APPLIED BEHAVIOR ANALYSIS 1996, 29, 305319 NUMBER3 (FALL 1996)

SEPARATE AND COMBINED EFFECTS OF METHYLPHENIDATE ANDA BEHAVIORAL INTERVENTION ON DISRUPTIVE BEHAVIOR IN

CHILDREN WITH MENTAL RETARDATION

NATHANJ. BLUM, JOYCE E. MAUK, JENNIFERJ. MCCOMAS,AND F. CHARLES MACE

UNIVERSITY OF PENNSYLVANIA

We investigated the separate and combined effects of a behavioral intervention and meth-ylphenidate (Ritalin) on disruptive behavior and task engagement in 3 children withsevere to profound mental retardation. The behavioral intervention involved differentialreinforcement of appropriate behavior and guided compliance. All 3 children demon-strated decreased disruptive behavior and improved task engagement in response to thebehavioral intervention. Two of the 3 children demonstrated similar improvement inresponse to methylphenidate. Although both interventions were highly effective for these2 participants, the relative efficacy of the interventions varied between the 2 children.There was no evidence of an additive or synergistic effect of the two interventions, but

the high efficacy of each intervention alone limited our ability to detect such effects.DESCRIPTORS: disruptive behavior, children with mental retardation, drug effects,

methylphenidate, differential reinforcement, guided compliance

Disruptive behavior in individuals withmental retardation is a common clinicalproblem that is treated by a variety of pro-fessionals including behavior analysts andphysicians. Physicians frequently prescribemedications to decrease disruptive behavior(Gadow, 1985), whereas behavior analysts

use a variety of behavioral procedures (Har-ris & Ersner-Hershfield, 1978; Matson &Gorman-Smith, 1986). Despite the wide-spread use of both pharmacologic and be-havioral approaches to treatment, with theexception of a few studies (Fisher, Piazza, &Page, 1989; Johnson, Handen, Lubetsky, &Sacco, 1994; Schell et al., 1986), it is un-common for physicians and behavior ana-lysts to collaborate in evaluating the efficacyof these interventions. Potential benefits ofcollaboration would include (a) more rigor-ous measures of behavioral effects and sideeffects than are usually found in pharmaco-logic studies, (b) use of experimental designsthat are capable of investigating treatment

Address correspondence and reprint requests to F.Charles Mace, Childrens Seashore House, 3405 CivicCenter Blvd., Philadelphia, Pennsylvania 19104.

effects on individual subjects, and (c) in-creased investigation of the relative efficacyof behavioral and pharmacologic treatments

when used alone and in combination(Schroeder, 1985).

Widespread use of neuroleptic medica-tions for individuals with mental retardation

has been criticized because of studies inwhich participants have shown behavioralimprovement when medication was discon-tinued (Heistad, Zimmermann, & Doebler,1982) and because of the potential of suchmedications to cause irreversible movementdisorders or other severe side effects (Whi-taker & Rao, 1992). Despite these concerns,neuroleptic medications continue to be thepsychotropic medication used most fre-quently in persons with mental retardation

(Baumeister, Todd, & Sevin, 1993; Gadow,1985). Given that overactive behavior is oneof the primary reasons for treatment withneuroleptics (Gadow, 1985), stimulant med-ications such as methylphenidate may be aneffective alternative (Aman, Marks, Turbott,

Wilsher, & Merry, 1991).Stimulant medications are used to treat

5/24/2018 jaba-29-03-0305

2/15

306 NATHAN J. BLUM et al.

attention deficit hyperactivity disorder(ADHD) in 1% to 2% of school-age chil-dren in the United States (Safer & Krager,

1988). Early studies of the use of stimulantsto treat persons with mental retardation sug-gested that they were ineffective (Aman,1982). However, more recent placebo-con-trolled studies suggested that children withmild to moderate mental retardation and hy-peractive behaviors respond to stimulants ina manner similar to children with normalintelligence and ADHD (Aman et al., 1991;Handen, Breaux, Gosling, Ploof, & Feld-man, 1990; Handen et al., 1992). Children

with mental retardation may have an in-creased risk of side effects such as tics andsocial withdrawal (Handen, Feldman, Gos-ling, Breaux, & McAuliffe, 1991; Helsel,Hersen, & Lubetsky, 1989). However, incontrast to the side effects of neuroleptics,the side effects of stimulants are usually re-versible when the medication is withdrawn.

Research on the effects of methylpheni-date on persons with severe to profoundmental retardation is limited. Aman et al.(1991) found that only 1 of 13 children and

adolescents with an IQ of less than 45 andADHD or conduct disorder had a beneficialresponse to methylphenidate. However, onlya 0.4 mg/kg dose of methylphenidate wasassessed in this study. In a placebo controlledstudy of 29 adolescents and adults with se-vere to profound mental retardation, Amanand Singh (1982) found no beneficial re-sponse to 0.3 mg/kg and 0.6 mg/kg dosesof methylphenidate on a variety of ratingscales and observations of ward behavior.However, the participants enrolled in this

study displayed a variety of disruptive be-haviors, including stereotypic behaviors,which one would not expect to respond tomethylphenidate. Moreover, the group ex-perimental design did not allow determina-tion of whether individual subjects may havebenefited from the medication. There are re-ports of beneficial effects of stimulants for

some individuals with severe to profoundmental retardation, but these reports havenot used experimental designs (Spencer,

1970; Triantafillou, 1972).In contrast, there is an extensive behav-

ioral literature on the treatment of disruptivebehaviors in persons with mental retardation(Harris & Ersner-Hershfield, 1978). In thisliterature, differential reinforcement proce-dures (including differential reinforcementof alternative behavior; DRA) are the mostfrequently described intervention for de-creasing problem behaviors (Lennox, Mil-tenberger, Spengler, & Erfanian, 1988). Dif-ferential reinforcement has been shown to be

effective in decreasing stereotyped behavior(Mulhern & Baumeister, 1969), inappropri-ate verbalizations (Deitz & Repp, 1973;Konczak & Johnson, 1983), disruptive class-room behaviors (Deitz, Repp, & Deitz,1976), and aggressive and self-injurious be-haviors (Repp & Deitz, 1974). Although thepotential of differential reinforcement pro-cedures to decrease problem behaviors isclear, the success of this intervention de-pends on the identification of effective re-

inforcers and maintaining variables that canbe altered (Vollmer & Iwata, 1992). Fur-thermore, behavioral treatments may not al-

ways reduce the disruptive behaviors to zeroor near-zero levels (Repp & Deitz, 1974;Sewell, McCoy, & Sewell, 1973) and mayrequire very high frequencies of reinforce-ment (Mulhern & Baumeister, 1969) thatmay be difficult to implement in the homeand school environments.

Given the advantages and disadvantagesof pharmacologic and behavioral treatments,

it is important to investigate the relative ef-ficacy of these interventions and the possi-bility that combining behavioral and phar-macologic treatments will result in additiveor interactive effects (Pelham & Murphy,1986). Studies of groups of children with

ADHD provide some support for the addi-tive effects of methylphenidate and behav-

5/24/2018 jaba-29-03-0305

3/15

307BEHAVIORAL AND PHARMACOLOGIC TREATMENT

ioral treatments (Carlson, Pelham, Milich,& Dixon, 1992; Pelham, Milich, & Walker,1986), but the extent to which these effects

are seen in individual children is less clear(Schell et al., 1986). In this study we inves-tigated the separate and combined effects ofmethylphenidate and a behavioral treatmentinvolving differential reinforcement andguided compliance in decreasing the disrup-tive behaviors of 3 children with mental re-tardation.

METHOD

Subjects and Setting

Three children with severe to profoundmental retardation participated in this study.

All engaged in severe disruptive and aggres-sive behaviors and were admitted to aneight-bed inpatient hospital unit dedicatedto the treatment of severe behavior disordersin individuals with developmental disabili-ties. Treatment sessions were conducted bygraduate students or trained behavior ana-lysts who worked with the children individ-ually. Sessions were conducted in a dormi-

tory-style room (4.5 m by 6.0 m) that servedas the childrens living quarters. It wasequipped with a full bathroom, a table (180cm long), two to four beds, and three to fivechairs. Participants were enrolled in thestudy after obtaining written parental in-formed consent.

Ted. Ted was a 6-year-old boy with severemental retardation whose aggressive and dis-ruptive behaviors included hitting, kicking,and throwing or knocking over objects andfurniture. His mother reported that his dis-

ruptive and aggressive behaviors occurredseveral times every day at both home andschool, but were most evident at home. Shestated that he had been hyperactive sincehe was an infant. Ted had never taken med-ications for his behavior. He lived with hismother, stepfather, and two siblings. Hecommunicated using approximately seven

signs and a variety of gestures. He could fol-low one-step commands and eat with uten-sils, but did not independently use the toilet.

On the Revised Gesell DevelopmentalSchedules (Knobloch, Stevens, & Malone,1987), his adaptive and language skills wereestimated to be at 21 months.

Bill.Bill was an 11-year-old boy with pro-found mental retardation whose aggressiveand disruptive behaviors included pullinghair, hitting, and throwing objects at othersor off shelves. He was described by hismother as very active, often climbing onpeople and jumping on furniture. These be-haviors occurred several times per day at

home. Bills behavior had been treated withthioridazine in the past without improve-ment. He was taking no medications at thetime of his admission. Bill lived with his par-ents and five siblings, two of whom also hadmental retardation. He communicated pri-marily by using gestures and pointing. Heimitated single-syllable words by usinggrunts of various intonations and could fol-low simple one-step commands. Bill re-quired assistance with all personal care skills.

He could eat with his fingers, but not withutensils. He could help with dressing by put-ting his arms through the sleeves of a shirtand by pulling up his pants.

Art.Art was a 7.5-year-old boy with frag-ile X syndrome, severe mental retardation,and a seizure disorder. His anticonvulsant(valproic acid) dose was not changed duringthe study. He was admitted for the treatmentof aggressive and disruptive behaviors thatincluded hitting, biting, pinching, and spit-ting at his parents and others. Prior to his

admission to this hospital, he had been ad-mitted to an inpatient psychiatric unit fortreatment of these behaviors. During thepsychiatric hospitalization, his aggressive be-haviors had been treated with risperidone,but his parents were concerned that heseemed to be sedated. Art was weaned offthe risperidone before this investigation was

5/24/2018 jaba-29-03-0305

4/15


begun. Art communicated using multiwordsentences. On the Revised Gesell Develop-mental Schedules (Knobloch et al., 1987),

his adaptive skills were at 36 months. Onthe Peabody Picture Vocabulary Test (Dunn& Munn, 1981), his age-equivalent score

was 27 months (standard score less than 40).

Dependent Variables and Data Collection

Ted and Bill. The primary dependentmeasure for Ted and Bill was the percentageof time engaged in disruptive behavior dur-ing 10-min task sessions in which they wereasked to place plastic blocks or books into aspecific container. Disruptive behaviors weredefined as throwing objects, climbing on orunder furniture, knocking over furniture,swinging doors open and closed repeatedly,or attempting to destroy objects by ripping,stomping, or kicking them. Data were col-lected on laptop computers using PortableComputer Systems For Observational Usesoftware. The duration of disruptive behav-ior was measured from the onset of the dis-ruptive behavior until the child had not en-gaged in any disruptive behavior for 5 s. In-

terobserver agreement was obtained on anaverage of 38% (Ted) and 32% (Bill) of ses-sions across all phases and conditions of theexperiment. Total agreement was calculatedas the exact overlap in the length of time (inseconds) that two independent observersagreed on the presence of disruptive behav-ior plus the exact overlap in the length oftime that they agreed on the absence of dis-ruptive behavior divided by the length of thesession (600 s) and multiplied by 100%. ForTed, mean agreement (with ranges in paren-

theses) for disruptive behavior during base-line conditions was 96% (85% to 100%) onplacebo and 98% (87% to 100%) on 0.3mg/kg of methylphenidate. During theDRA and guided compliance intervention,mean agreement for disruptive behavior was89% (63% to 100%) on placebo and 99%(93% to 100%) on 0.3 mg/kg of methyl-

phenidate. For Bill, mean agreement for dis-ruptive behavior during baseline conditions

was 94% (87% to 99%) on placebo, 98%

(96% to 99%) on 0.3 mg/kg of methyl-phenidate, and 98% (97% to 100%) on 0.6mg/kg of methylphenidate. During theDRA and guided compliance intervention,mean agreement for disruptive behavior was96% (88% to 100%) on placebo, 98%(97% to 99%) on 0.3 mg/kg of methyl-phenidate, and 99% (97% to 100%) on 0.6mg/kg of methylphenidate.

Engagement with the task, defined as thechilds picking up or walking with a toy orbook in the direction of the container or de-

positing a toy or book in the container, wasalso measured. Engagement was coded fromthe onset of the behavior until these behav-iors were discontinued for at least 5 s. Cod-ing of engagement was also stopped if thechild exhibited disruptive behaviors, receivedphysical guidance from the therapist, or helda toy or book for longer than 5 s withoutattempting to deposit it in the container. In-terobserver agreement for engagement wascalculated in the same manner as described

for disruptive behavior. For Ted, meanagreement for engagement during baselineconditions was 99% (97% to 100%) on pla-cebo and 98% (86% to 100%) on 0.3mg/kg of methylphenidate. During theDRA and guided compliance intervention,mean agreement for engagement was 88%(33% to 100%) on placebo and 98% (89%to 100%) on 0.3 mg/kg of methylphenidate.For Bill, mean agreement for engagementduring baseline conditions was 93% (87%to 99%) on placebo, 88% (87% to 88%) on

0.3 mg/kg of methylphenidate, and 94%(91% to 100%) on 0.6 mg/kg of methyl-phenidate. During the DRA and guidedcompliance intervention, mean agreementfor engagement was 95% (85% to 98%) onplacebo, 94% (90% to 97%) on 0.3 mg/kgof methylphenidate, and 97% (95% to99%) on 0.6 mg/kg of methylphenidate.

5/24/2018 jaba-29-03-0305

5/15


Art. The primary dependent measure wasthe rate of disruptive behaviors during anumber or letter identification task. Disrup-

tive behavior was measured as a rate ratherthan as a percentage of time, because Artsdisruptive behaviors tended to be discreteevents of short duration. Disruptive behav-iors were defined as described above. How-ever, because Art had to sit at a table to en-gage in this task, getting up from the table

was also coded as a disruptive behavior. En-gagement was defined as holding or lookingat task materials. It was measured from theonset of the behavior until Art did not en-gage in the behavior for 5 s. If Art held thetask material for longer than 5 s withoutlooking at it, he was not considered to beengaged with the task. Data were collectedon laptop computers as described above. In-terobserver agreement was obtained on anaverage of 23% of the sessions across allphases of the experiment. For disruptive be-havior, interobserver agreement was calculat-ed on a point-by-point basis (House, House,& Campbell, 1981). Counts of disruptivebehaviors within 10-s intervals were com-

pared for two independent observers, andthe number of agreements was divided bythe total number of agreements plus dis-agreements and multiplied by 100%. Dur-ing baseline conditions, mean agreement fordisruptive behavior was 91% (76% to100%) on placebo, 96% (93% to 100%) on0.3 mg/kg of methylphenidate, and 100%on 0.6 mg/kg of methylphenidate. Duringthe DRA and guided compliance interven-tion, mean agreement for disruptive behav-ior was 95% (92% to 97%) on placebo,

91% (77% to 100%) on 0.3 mg/kg meth-ylphenidate, and 100% on 0.6 mg/kg meth-ylphenidate. For engagement, interobserveragreement was calculated by the observa-tional software as described above. Duringbaseline conditions, mean agreement for en-gagement was 92% (88% to 100%) on pla-cebo, 96% (93% to 100%) on 0.3 mg/kg of

methylphenidate, and 92% (agreement re-corded in only one session) on 0.6 mg/kg ofmethylphenidate. During the DRA and

guided compliance intervention, meanagreement for engagement was 93% (90%to 97%) on placebo, 86% (59% to 100%)on 0.3 mg/kg of methylphenidate, and 97%(agreement recorded in only one session) on0.6 mg/kg of methylphenidate.

Procedures

Baseline. During each baseline session forTed and Bill, the therapist pointed to a spe-cific block and the designated box and said,

Put the block in this box. If the childcomplied with the request, the therapist is-sued the next request upon completion ofthe task. If the child did not comply withthe request, the therapist repeated the re-quest 30 s later. Disruptive behaviors wereignored during this condition.

During baseline sessions with Art, thetherapist presented him with two letters orthree numbers and said, Point to the . . .(a letter or number). If he did not respond

within 5 s, the therapist repeated the direc-

tion. If he did not respond to this secondinstruction within 5 s, he was shown thecorrect answer. When Art pointed to a cardor after he was shown the correct answer,another trial with a different set of lettersand numbers began. No reinforcement fortask completion was provided. The therapistdid not respond to disruptive behaviors.However, if Art got up from the table, thetherapist brought him back.

DRA and guided compliance. During each

session for Ted and Bill, the therapist askedthe child to complete the task describedabove. If the child complied with the re-quest, the therapist provided a reinforcer,and the next request was made after thechild was given the reinforcer. If the childdid not comply with the request, he wasphysically guided to complete the task and

5/24/2018 jaba-29-03-0305

6/15


the next request was issued immediatelyupon completion.

Art was given the instructions described

above. If he pointed to either card, he wasgiven verbal praise. If he pointed to the cor-rect card, he also received an edible rein-forcer on a variable-ratio (VR) 3 schedule. Ifhe did not respond to the second verbal in-struction, hand-over-hand physical guidance

was used to assist him in pointing to thedesignated number or letter.

For Ted, verbal praise and physical atten-tion were initially used as reinforcers, but

when they appeared to be ineffective, the re-inforcers were changed to verbal praise andedible items. For Bill and Art, a forced-choice preference assessment (Fisher etal.,1992) was used to identify reinforcers.Ten stimuli (a combination of edible items,solitary play items, and interactive playitems) were presented two at a time, and thechild was allowed 30 s of access to the stim-ulus he selected. The pairings were presentedin a randomized order, and each stimulus

was paired with all other stimuli one time.The three stimuli selected most frequently

were chosen as reinforcers. For both Bill andArt, the three most highly preferred stimuliwere all edible items (small piece of a cookieor cracker, or a piece of cereal).

Methylphenidate.Each child was random-ly assigned to receive a placebo (A), a 0.3mg/kg dose of methylphenidate (B), and a0.6 mg/kg dose of methylphenidate (C), inone of the following four orders: ABACA(Bor C), ACABA(B or C), BACA(B or C), orCABA(B or C). The methylphenidate dose

was rounded to the nearest 2.5 mg, and

identical capsules containing methylpheni-date or a placebo were prepared by the hos-pital pharmacy. The child received one cap-sule at 8 a.m. and one at noon. All sessions

were conducted 1 to 3 hr after the child re-ceived the capsule. All members of the treat-ment team, participants in the study, andfamily members were blind to the order of

medication administration during the firstfour phases of the study. At the end of thefourth phase, one physician was unblinded

in order to instruct the pharmacy aboutwhat medication to prepare for the finalphase. This was done to allow within-subjectreplication of the most effective methylphen-idate dose as determined during the firstfour phases of the investigation. All othermembers of the treatment team, studygroup, and family remained blind to themedication that the child received duringthese phases.

Teds medication schedule was random-ized in the ACABA(B or C) order, but he

had side effects to the medication (tics) onthe first day that the 0.6 mg/kg dose ofmethylphenidate was given. When the ticsoccurred, one physician was unblinded tomedication conditions and instructed thepharmacy to continue with the study usingan ABAB design. All other members of thetreatment team remained blind to medica-tion conditions. Bills and Arts medicationschedules were both randomized in theBACA(B or C) order.

Experimental DesignThe baseline sessions and the DRA and

guided compliance sessions were conductedeach day of the study. They were counter-balanced in a multielement design across allmedication phases. The order of the sessionseach day was randomized. Methylphenidateand the placebo were alternated in a reversaldesign for the most effective methylpheni-date dose.

RESULTS

Ted

Figure 1 shows the percentage of disrup-tive behavior during each session. Initially,verbal praise and physical attention wereused as reinforcers in the DRA intervention,but when they appeared to be ineffective, the

5/24/2018 jaba-29-03-0305

7/15


Figure 1. Percentage of each task session engaged in disruptive behaviors (Ted, Bill) or number of disruptive

behaviors per hour during each session (Art). Baseline conditions (open squares) were counterbalanced with adifferential-reinforcement-of-alternative-behavior and guided compliance (DRA GC) intervention across allmethylphenidate (MPH) and placebo phases. The open circles indicate that praise and attention were thereinforcers, and the filled circles indicate that verbal praise and an edible item were the reinforcers.

5/24/2018 jaba-29-03-0305

8/15


reinforcers were changed to verbal praise andfood for the remainder of the study. DuringPhase 2 on placebo with the DRA and guid-

ed compliance intervention, disruptive be-havior occurred during an average of 11.3%of each session. In baseline sessions duringthis phase, disruptive behavior occurred dur-ing 61.4% of each session.

On the 0.6 mg/kg dose of methylpheni-date (Phase 3), only two data points wereobtained because Ted developed tics, and he

was tearful or crying much of the day. Theseside effects were resolved on the 0.3 mg/kgdose of methylphenidate, and thus only theefficacy of the 0.3 mg/kg dose of methyl-phenidate was further assessed.

During Phase 4 on the 0.3 mg/kg dose ofmethylphenidate, disruptive behavior oc-curred during an average of 1.7% of eachsession with the DRA and guided compli-ance intervention and during an average of44.6% of each session in baseline. Duringthe return-to-placebo conditions, disruptivebehavior increased to an average of 82.3%of each session in baseline, but there was noreversal of disruptive behavior in the sessions

with the behavioral intervention. DuringPhase 6, on methylphenidate at the 0.3mg/kg dose, disruptive behavior occurredduring 65.2% of each session under baselineconditions and remained at near-zero levels

with the DRA and guided compliance in-tervention. Although the behavioral inter-vention was clearly effective, the 0.3 mg/kgdose of methylphenidate did not have a clin-ically significant effect.

Engagement data are shown in Figure 2.Although engagement initially seemed to

improve on methylphenidate, this effect wasnot replicated during the reversal phase.During Phase 2, the DRA and guided com-pliance intervention resulted in engagementduring an average of 31.2% of each session.In both phases that combined the DRA in-tervention with methylphenidate, mean en-gagement improved to over 90% of each ses-

sion. However, even when methylphenidatewas withdrawn, the DRA and guided com-pliance intervention continued to result in

similarly high levels of engagement.

Bill

The results in Figure 1 show that the 0.6mg/kg dose of methylphenidate (Phase 3)

was more effective in decreasing disruptivebehaviors than the 0.3 mg/kg dose (Phase1). During Phase 2 on placebo, Bill was dis-ruptive an average of 56.6% of the time dur-ing baseline sessions, compared to an averageof 20.2% of the time during sessions withthe DRA and guided compliance interven-

tion. During Phase 3, treatment with the 0.6mg/kg dose of methylphenidate suppresseddisruptive behavior to an average of 2.7% ofeach session under baseline conditions andan average of 1.5% of each session with thebehavioral intervention. During return tothe placebo, disruptive behavior increased toan average of 31.4% of each session duringbaseline conditions, but remained at a lowlevel during sessions with the behavioral in-tervention. Reintroduction of methylpheni-

date at the 0.6 mg/kg dose replicated thedecrease in disruptive behavior during base-line sessions. Medication and the DRA andguided compliance intervention were notmore effective than the medication alone.During the first placebo phase, the DRA andguided compliance intervention alone wasnot as effective as it was during the secondplacebo phase. However, during the secondplacebo phase, the DRA and guided com-pliance intervention alone was approximate-ly as effective as the 0.6 mg/kg dose of meth-

ylphenidate alone. Although both treatmentswere effective when used alone, there was noevidence of an additive effect when the treat-ments were used together.

Engagement data are shown in Figure 2.During Phase 2 while on placebo, engage-ment occurred during an average of 16.5%of each session under baseline conditions

5/24/2018 jaba-29-03-0305

9/15


Figure 2. Mean percentage of each session engaged in the task for the baseline condition and the differentialreinforcement and guided compliance (DRA GC) conditions across all methylphenidate (MPH) and placeboconditions. Bars indicate ranges.

5/24/2018 jaba-29-03-0305

10/15


and 45.9% of each session with the DRAand guided compliance intervention. Treat-ment with the 0.6 mg/kg dose of methyl-

phenidate did not further improve engage-ment during sessions combined with the be-havioral intervention, but did improve en-gagement in baseline sessions.

Art

The results in Figure 1 indicate that the0.6 mg/kg dose of methylphenidate (Phase3) was more effective than the 0.3 mg/kgdose (Phase 1) in suppressing disruptive be-havior. In the first placebo condition, dis-ruptive behaviors occurred an average of39.6 times per hour during baseline sessionsand 3.4 times per hour in sessions with theDRA and guided compliance intervention.During the next phase of the study on the0.6 mg/kg dose of methylphenidate, disrup-tive behavior did not occur in any baselinesession and occurred at an average rate of1.5 per hour during sessions with the be-havioral intervention. When the medication

was withdrawn, disruptive behaviors oc-curred at an average rate of 105.6 per hour

during baseline sessions and 15.0 per hourduring sessions with the behavioral interven-tion. Reintroduction of the medication sup-pressed disruptive behaviors to near-zero lev-els under both conditions. Methylphenidatealone and the behavioral treatment alone

were very effective at decreasing disruptivebehaviors. During the second placebo phase,the behavioral treatment was not as effectiveas it was during the first placebo phase.Thus, in general, it appears that the medi-cation alone may have been somewhat more

effective than the DRA and guided compli-ance intervention alone.Engagement data are shown in Figure 2.

During the first placebo phase, Art was en-gaged with the task during an average of91.4% of each session with the DRA andguided compliance intervention and 42.2%under baseline conditions. Treatment with

the 0.6 mg/kg dose of methylphenidate in-creased engagement to over 90% underbaseline conditions. There was no additional

effect of the behavioral intervention.

DISCUSSION

This study demonstrates the use of single-subject experimental methodology to inves-tigate the separate and combined effects ofa behavioral and pharmacologic interventionconcurrently. The multielement design al-lows comparison of baseline and behavioraltreatment alone conditions during placebophases and comparison of methylphenidatealone and combined treatment during meth-ylphenidate phases. Methylphenidate andplacebo phases were compared using a with-in-subject reversal design. Methylphenidatehas a rapid onset of action (within 30 min)and short duration of action (3 to 4 hr) thatallow for rapid within-subject reversals(Greenhill, 1995).

This is one of only a few studies (Johnsonet al., 1994; Schell et al., 1986) that haveused single-subject experimental designs that

allow a controlled comparison of baselineconditions, a behavioral intervention alone,methylphenidate alone, and the combina-tion of methylphenidate and a behavioral in-tervention for the treatment of disruptivebehavior in individual subjects. It is the onlystudy to provide a relatively large number ofdata points within each medication condi-tion, thus allowing for a more complete eval-uation of within-subject effects and an as-sessment of the stabilization of data withinmedication conditions. Furthermore, it is

the only study to use a single-case experi-mental design to investigate these effects inchildren with severe to profound mental re-tardation.

For all 3 participants in this study, theDRA and guided compliance interventionreduced disruptive behaviors and improvedengagement with the task. For Ted, the low

5/24/2018 jaba-29-03-0305

11/15


dose of methylphenidate (0.3 mg/kg) had noclinically significant effect on disruptive be-havior or engagement, and the high dose of

methylphenidate (0.6 mg/kg) was discontin-ued due to side effects. For Bill and Art, thehigh dose of methylphenidate decreased dis-ruptive behavior and improved engagement

with the task. In both cases, the high dosewas much more effective than the low dose.

For Bill, high-dose methylphenidate aloneinitially seemed to be more effective thanDRA and guided compliance alone, but dur-ing the second placebo and high-dose meth-ylphenidate phases, the two interventionsappeared to be equally effective in decreasingdisruptive behavior. Both interventions werealso effective in improving engagement withthe task. However, despite the effectivenessof each intervention alone, there was no ev-idence that the two interventions had an ad-ditive or synergistic effect in decreasing dis-ruptive behaviors or improving engagement

with the task. The effectiveness of each in-tervention alone may have limited the pos-sibility of detecting an additive or synergisticeffect in decreasing disruptive behavior, but

this would not explain the failure to dem-onstrate such an effect on engagement withthe task. In Arts case, high-dose methyl-phenidate alone was slightly more effectivethan the DRA intervention alone in decreas-ing disruptive behavior and improving en-gagement with the task. Methylphenidate

was so effective that the potential for a syn-ergistic effect cannot be evaluated because ofa floor effect for disruptive behavior and aceiling effect for engagement.

The data demonstrate intersubject vari-

ability in the relative efficacy of the behav-ioral and pharmacologic interventions. Tedresponded well to the DRA and guidedcompliance intervention and had intolerableside effects on the high dose of methylphen-idate. Bill responded similarly to both inter-ventions, and Art responded better to meth-ylphenidate than to the behavioral interven-

tion. These findings support other studies ofindividual children with mental retardation,some of which report that behavioral inter-

ventions are more effective than methyl-phenidate (Shafto & Sulzbacher, 1977), andothers of which report that methylphenidateis more effective than a behavioral interven-tion in decreasing defiant, hyperactive, ordisruptive behaviors (Johnson et al., 1994;Schell et al., 1986). The use of reinforcerassessments to identify effective reinforcersfor Bill and Art may help to explain the ef-fectiveness of the behavioral interventions inthis study. However, it is noteworthy thatdespite the use of a reinforcer assessment in

developing the behavioral intervention forArt, the 0.6 mg/kg dose of methylphenidatewas more effective than the behavioral treat-ment. The use of functional analysis and re-inforcer assessment to develop behavioral in-terventions to study in combination withpharmacologic treatments is a potential areafor greater collaboration between physiciansand behavior analysts.

Other studies of children with mental re-tardation also have failed to show an additive

or synergistic effect of methylphenidate andbehavioral treatment on disruptive behavioror engagement with a task. Christensen(1975) found that adding 0.3 mg/kg ofmethylphenidate to a behavioral treatmentinvolving token and verbal reinforcementplus extinction did not improve the class-room behavior of 16 children with mentalretardation. Schell et al. (1986) assessed theseparate and combined effects of a behavior-al intervention and a 0.3 mg/kg dose ofmethylphenidate in a child with mild mental

retardation. They did not find an additiveeffect of methylphenidate and the behavioralintervention on defiant behavior, but theydid find an additive effect of the two inter-ventions on correct responses to an academictask. Similarly, in 3 children with mild men-tal retardation, Johnson et al. (1994) foundadditive effects of methylphenidate and a be-

5/24/2018 jaba-29-03-0305

12/15


havioral intervention on task accuracy, butnot on measures of fidgety or defiant behav-iors. Thus, although we did not find an ad-

ditive or synergistic effect of methylpheni-date and the DRA and guided complianceintervention, we cannot rule out that suchan effect may have been found had we in-vestigated other behaviors or these behaviorsin other settings.

In contrast to the predominately single-case studies in children with mental retar-dation discussed above, group studies ofchildren with ADHD and average intelli-gence have found that behavioral treatmentsin combination with low doses of methyl-phenidate are more effective than either in-tervention alone (Carlson et al., 1992; Pel-ham et al., 1986; Pelham & Murphy, 1986).However, some studies suggest that a highdose of methylphenidate alone is as effectiveas combined treatments (Carlson et al.,1992; Gittelman et al., 1980). The degreeto which these group studies characterize theresponse of individual children to these in-terventions is not known. This gap in theliterature highlights the need for further

studies of these effects using single-subjectexperimental designs.Interestingly, during the initial phase with

the DRA and guided compliance interven-tion and placebo, Ted and Bill exhibited rel-atively high rates of disruptive behaviors. Inboth cases, it was during treatment withmethylphenidate that disruptive behaviorsdecreased to near-zero levels. However, a re-turn to the behavioral intervention alone didnot result in a reversal in the frequency ofdisruptive behavior. In contrast, the frequen-

cy of disruptive behavior did reverse in thebaseline condition. These results suggests thepossibility that for these 2 participants,methylphenidate facilitated learning of thebehavioral contingencies. Although a facili-tative effect of methylphenidate on learningbehavioral contingencies has been suggestedby others (Christensen & Sprague, 1973), it

has not been clearly demonstrated. This ex-planation for the results of our study mustbe regarded as tentative, because this study

was not designed to investigate this hypoth-esis.

This is the first report of the beneficialeffects of stimulants in children with severeto profound mental retardation that has usedan experimental design to document the ef-fects. It is consistent with previous case re-ports of stimulant medications resulting inimproved behavior in this population (Spen-cer, 1970; Triantafillou, 1972). However, intwo group studies (Aman, Kern, McGhee,& Arnold, 1993; Aman et al., 1991), only

1 out of a total of 18 children with an IQof less than 45 had a beneficial response tomethylphenidate. In both studies, only a 0.4mg/kg dose of methylphenidate was as-sessed. Given that the response to the 0.6mg/kg dose was much greater than the re-sponse to the 0.3 mg/kg dose in our study,it is possible that the low response rate inthe previous studies was related to the rela-tively low dose of methylphenidate that wasused. Aman and Singh (1982) used both 0.3

mg/kg and 0.6 mg/kg doses of methylphen-idate in a group of adolescents and adultswith severe to profound mental retardationand reported no benefit for either dose, butthe group experimental design may not haveallowed detection of individuals who bene-fited from the medication. Alternatively, theresponse to medication in adults with severeto profound mental retardation may be dif-ferent from the response in children.

The findings of this study show the idi-osyncratic character of response to methyl-

phenidate and to behavioral treatment inchildren with severe to profound mental re-tardation. The findings suggest that individ-uals who conduct interventions for disrup-tive behaviors in this population should con-sider both behavioral and pharmacologic ap-proaches. When considering the use ofmethylphenidate, one should remember that

5/24/2018 jaba-29-03-0305

13/15


children with mental retardation have beenreported to have a higher incidence of sideeffects than is found in the general popula-

tion (Aman et al., 1991; Handen et al.,1991). This may be especially true of social

withdrawal, which is a commonly reportedside effect in children with mental retarda-tion (Handen et al., 1991; Helsel et al.,1989), but is not frequently found in chil-dren of average intelligence (Ahmann et al.,1993; Barkley, McMurray, Edelbrock, &Robbins, 1990). Due to this higher rate ofside effects and the difficulty in predicting

which childrens behavior will improve on

stimulants, it is necessary to carefully mon-itor for efficacy and side effects when chil-dren with mental retardation are givenmethylphenidate. It is likely that this can bebest accomplished if physicians and behavioranalysts collaborate on the clinical manage-ment of these individuals.

REFERENCES

Ahmann, P. A., Waltonen, S. J., Olson, K. A., Theye,F. W., Van Erem, A. J., & LaPlant, R. J. (1993).

Placebo-controlled evaluation of Ritalin side ef-fects.Pediatrics, 91, 11011106.Aman, M. G. (1982). Stimulant drug effects in de-

velopmental disorders and hyperactivitytowarda resolution of disparate findings. Journal of Au-tism and Developmental Disorders, 12, 385398.

Aman, M. G., Kern, R. A., McGhee, D. E., & Arnold,L. E. (1993). Fenfluramine and methylphenidatein children with mental retardation and ADHD:Clinical and side effects. Journal of the AmericanAcademy of Child and Adolescent Psychiatry, 32,851859.

Aman, M. G., Marks, R. E., Turbott, S. H., Wilsher,C. P., & Merry, S. N. (1991). Clinical effects ofmethylphenidate and thioridazine in intellectually

subaverage children.Journal of the American Acad-emy of Child and Adolescent Psychiatry, 30, 246256.

Aman, M. G., & Singh, N. N. (1982). Methylphen-idate in severely retarded residents and the clinicalsignificance of stereotypic behavior. Applied Re-search in Mental Retardation, 3, 345358.

Barkley, R. A., McMurray, M. B., Edelbrock, C. S.,& Robbins, K. (1990). Side effects of methyl-phenidate in children with attention deficit hy-

peractivity disorder: A systemic, placebo-con-trolled evaluation.Pediatrics, 86, 184192.

Baumeister, A. A., Todd, M. E., & Sevin, J. A.(1993). Efficacy and specificity of pharmacologi-

cal therapies for behavioral disorders in personswith mental retardation. Clinical Neuropharma-cology, 16, 271294.

Carlson, C. L., Pelham, W. E., Milich, R., & Dixon,J. (1992). Single and combined effects of meth-ylphenidate and behavior therapy on the class-room performance of children with attention-def-icit hyperactivity disorder. Journal of AbnormalChild Psychology, 20, 213232.

Christensen, D. E. (1975). Effects of combiningmethylphenidate and a classroom token system inmodifying hyperactive behavior.American Journalof Mental Deficiency, 80, 266276.

Christensen, D. E., & Sprague, R. L. (1973). Reduc-tion of hyperactive behaviors by conditioning pro-

cedures alone and combined with methylpheni-date (Ritalin).Behaviour Research and Therapy, 11,331334.

Deitz, S. M., & Repp, A. C. (1973). Decreasing class-room misbehavior through DRL schedules of re-inforcement. Journal of Applied Behavior Analysis,6, 457463.

Deitz, S. M., Repp, A. C., & Deitz, D. E. D. (1976).Reducing inappropriate classroom behaviour ofretarded students through three procedures of dif-ferential reinforcement. Journal of Mental Defi-ciency Research, 20, 155169.

Dunn, L. M., & Munn, L. M. (1981). Peabody picturevocabulary testrevised.Circle Pines, MN: Amer-ican Guidance Service.

Fisher, W., Piazza, C. C., Bowman, L. G., Hagopian,L. P., Owens, J. C., & Slevin, I. (1992). A com-parison of two approaches for identifying rein-forcers for persons with severe and profound dis-abilities. Journal of Applied Behavior Analysis, 25,491498.

Fisher, W., Piazza, C. C., & Page, T. J. (1989). As-sessing independent and interactive effects of be-havioral and pharmacologic interventions for a cli-ent with dual diagnoses. Journal of Behavior Ther-apy & Experimental Psychiatry, 20, 241250.

Gadow, K. D. (1985). Prevalence and efficacy ofstimulant drug use with mentally retarded chil-dren and youth. Psychopharmacology Bulletin, 21,291303.

Gittelman, R., Abikoff, H., Pollack, E., Klein, D. F.,Katz, S., & Mattes, J. (1980). A controlled trialof behavior modification and methylphenidate inhyperactive children. In C. K. Whalen & B.Henker (Eds.), Hyperactive children: The socialecology of identification and treatment (pp. 221243). New York: Academic Press.

Greenhill, L. L. (1995). Attention-deficit hyperactiv-ity disorder: The stimulants. Child and AdolescentPsychiatric Clinics of North America, 4, 123168.

5/24/2018 jaba-29-03-0305

14/15


Handen, B. L., Breaux, A., Gosling, A., Ploof, D. L.,& Feldman, H. (1990). Efficacy of methylphen-idate among mentally retarded children with at-tention deficit hyperactivity disorder. Pediatrics,

86, 922930.Handen, B. L., Breaux, A., Janosky, J., McAuliffe, S.,

Feldman, H., & Gosling, A. (1992). Effects andnoneffects of methylphenidate in children withmental retardation and ADHD. Journal of theAmerican Academy of Child and Adolescent Psychi-atry, 31, 455461.

Handen, B. L., Feldman, H., Gosling, A., Breaux, A.,& McAuliffe, S. (1991). Adverse side effects ofmethylphenidate among mentally retarded chil-dren with ADHD. Journal of the American Acad-emy of Child and Adolescent Psychiatry, 30, 241245.

Harris, S. A., & Ersner-Hershfield, R. (1978). Be-havioral suppression of seriously disruptive behav-

ior in psychotic and retarded patients: A review ofpunishment and its alternatives.Psychological Bul-letin, 85, 13521375.

Heistad, G. T., Zimmermann, R. L., & Doebler, M.I. (1982). Long-term usefulness of thioridazinefor institutionalized mentally retarded patients.American Journal of Mental Deficiency, 87, 243251.

Helsel, W. A., Hersen, M., & Lubetsky, M. A. (1989).Stimulant medication and the retarded.Journal ofthe American Academy of Child and Adolescent Psy-chiatry, 28, 138139.

House, A. E., House, B. J., & Campbell, M. B.(1981). Measures of interobserver agreement:Calculation formulas and distribution effects.Journal of Behavioral Assessment, 3, 3757.

Johnson, C. R., Handen, B. L., Lubetsky, M. J., &Sacco, K. A. (1994). Efficacy of methylphenidateand behavioral intervention on classroom behaviorin children with ADHD and mental retardation.Behavior Modification, 18, 470487.

Knobloch, H., Stevens, F., & Malone, A. F. (1987).Manual of developmental diagnoses.Houston, TX:Developmental Evaluation Materials, Inc.

Konczak, L. P., & Johnson, C. M. (1983, April). Re-ducing inappropriate verbalizations in a sheltered

workshop through differential reinforcement ofother behavior. Education and Training of theMentally Retarded,pp. 120124.

Lennox, D. B., Miltenberger, R. G., Spengler, P., &Erfanian, N. (1988). Decelerative treatment prac-

tices with persons who have mental retardation: Areview of five years of the literature. AmericanJournal of Mental Retardation, 92, 492501.

Matson, J. L., & Gorman-Smith, D. (1986). A re-view of treatment research for aggressive and dis-ruptive behavior in the mentally retarded.AppliedResearch in Mental Retardation, 7, 95103.

Mulhern, T., & Baumeister, A. A. (1969). An exper-imental attempt to reduce stereotypy by reinforce-

ment procedures.American Journal of Mental De-ficiency, 74, 6974.

Pelham, W. E., Milich, R., & Walker, J. (1986). Theeffects of continuous and partial reinforcement

and methylphenidate on learning in children withattention deficit disorder.Journal of Abnormal Psy-chology, 95, 319325.

Pelham, W. E., & Murphy, H. A. (1986). Behavioraland pharmacologic treatment of attention deficitand conduct disorders. In M. Hersen (Ed.),Phar-macologic and behavioral treatment: An integrativeapproach(pp. 108148). New York: Wiley.

Repp, A. C., & Deitz, S. M. (1974). Reducing ag-gressive and self-injurious behavior of institution-alized retarded children through reinforcement ofother behaviors.Journal of Applied Behavior Anal-ysis, 7, 313325.

Safer, D. J., & Krager, J. M. (1988). A survey ofmedication treatment for hyperactive/inattentive

children. Journal of the American Medical Associa-tion, 260, 22562258.

Schell, R. M., Pelham, W. E., Bender, M. E., Andree,J. A., Law, T., & Robbins, F. R. (1986). Theconcurrent assessment of behavioral and psycho-stimulant interventions: A controlled case study.Behavioral Assessment, 8, 373384.

Schroeder, S. R. (1985). Workshop summary: Issuesand future research directions of pharmacotherapyin mental retardation. Psychopharmacology Bulle-tin, 21, 323326.

Sewell, E., McCoy, J. F., & Sewell, W. R. (1973).Modification of an antagonistic social behavior us-ing positive reinforcement for other behavior.ThePsychological Record, 23, 499504.

Shafto, F., & Sulzbacher, S. (1977). Comparing treat-ment tactics with a hyperactive preschool child:Stimulant medication and programmed teacherintervention. Journal of Applied Behavior Analysis,10, 1320.

Spencer, D. A. (1970). Ronyl (pemoline) in overac-tive mentally subnormal children. British Journalof Psychiatry, 117, 239240.

Triantafillou, M. (1972). Pemoline in overactive men-tally handicapped children. British Journal of Psy-chiatry, 121, 577.

Vollmer, T. R., & Iwata, B. A. (1992). Differentialreinforcement as treatment for behavior disorders:Procedural and functional variations. Research inDevelopmental Disabilities, 13, 393417.

Whitaker, A., & Rao, U. (1992). Neuroleptics in pe-diatric psychiatry.The Psychiatric Clinics of NorthAmerica, 15, 243276.

Received October 17, 1995Initial editorial decision December 16, 1995Revisions received February 12, 1996; April 13, 1996Final acceptance April 23, 1996Action Editor, Patrick C. Friman

5/24/2018 jaba-29-03-0305

15/15


STUDY QUESTIONS

1. What are some disadvantages of operant and pharmacological approaches to the treatmentof behavior disorders, and what potential benefits might result from an integrative approach?

2. What were the target behaviors and how were they measured?

3. Describe the contingencies in effect during the baseline and the DRA plus guided complianceconditions.

4. Briefly describe the medication dosages that were evaluated and the general method ofadministration.

5. What types of experimental designs were used to evaluate the effects of the behavioral andpharmacological interventions?

6. What were the effects of each treatment separately on disruptive behavior and task engage-ment for Ted, Bill and Art, and which treatment was more effective for each subject?

7. The authors indicated that there was no evidence of an interactive or synergistic effectresulting from combined treatment. What type of outcome would have suggested an inter-action effect, and what feature of the results may have obscured such an effect?

8. The authors mention that they used a preference assessment to aid in the identification ofreinforcers used during treatment. What additional analysis may have been helpful in de-veloping the behavioral interventions or in interpreting the results if the behavioral inter-vention had been less effective than it was?

Questions prepared by SungWoo Kahng and Michele Wallace

Documents

jaba-29-03-0305