Use of Skin-Shock at the Judge Rotenberg Educational Center (JRC)

 

print this page

Treatment of Life-Threatening Vomiting and Rumination
with Contingent Electric Shock

Robert W. Worsham, Matthew L. Israel, Robert E. von Heyn, and Daniel A. Connolly

Behavior Research Institute

This paper describes the treatment of life-threatening ruminating and vomiting in a 12-year-old boy with profound mental retardation and pervasive developmental disorder. Body weight and behavioral data are presented for baseline and behavior modification treatment conditions over a period of two-and-a-half years. Treatment conditions which were held constant throughout included positive reinforcement schedules (DRO, DRA), a functional communication system, and skills training. Treatment conditions which varied consisted of contingent water spray, and four levels of contingent electrical stimulation (shock). The shock stimulus was administered via the Self-Injurious Behavior Inhibiting System (SIBIS) and the Graduated Electronic Decelerator (GED). Results showed initial partial success with SIBIS, followed by subsequent adaptation and behavioral recovery. This resulted in a life-threatening loss in body weight due to accelerated vomiting. Treatment with GED, using a stronger level and increased duration of shock, multiple electrodes, and varying electrode placement, resulted in reduction of ruminating and vomiting behaviors to near-zero, positive social side effects, and a substantial concomitant weight gain.

Rumination is a common problem among the developmentally disabled. It has been defined as "regurgitation of previously ingested food and its re-consumption and/or drooling from the mouth" (Singh, Manning & Angell, 1982). Estimates of the incidence of rumination in institutionalized developmentally disabled persons vary. For example, Singh (1981) has estimated the incidence of rumination at six percent, while Ball, Hendrickson, and Clayton (1974) have estimated it at 9.6 percent.

Starin and Fuqua (1987) have listed health problems that may arise from chronic rumination, some of the more serious being weight loss, malnutrition, and dehydration. At the extreme, chronic rumination and vomiting has lead to death in infants (Kanner, 1972).

Among behavioral treatment techniques, the use of contingent punishment is one of the most prevalent and effective methods (Davis & Cuvo, 1980; Starin & Fuqua, 1987). Punishment procedures that have been used to treat rumination and/or vomiting in developmentally disabled individuals include contingent electric shock (Bright & Whaley, 1968; Galbraith, Byrick, & Rutledge, 1970; Kohlenberg, 1970; Luckey, Watson, & Musick, 1968; Watkins, 1972; White & Taylor, 1967; Wright & Thalassinos, 1973), noxious tastes (Becker, Turner, & Sajwaj, 1978; Bright & Whaley, 1968; Hogg, 1982; and Marholin, Luiselli, Robinson, & Lott, 1980), pinch (Minness, 1980), oral hygiene (Foxx, Snyder, & Schroeder, 1979; Singh, Manning, & Angell, 1982), contingent exercise (Daniel, 1982; O’Neil, White, King, & Carek, 1979), and overcorrection (Azrin & Wesolowski, 1975; Duker & Seys, 1977). Contingent shock has also been used very successfully to treat life-threatening ruminative vomiting in normal infants, with subsequent weight gain and maintenance of effects over time (Cunningham & Linscheid, 1976; Linscheid & Cunningham, 1977).

In general, it is difficult to draw conclusions about the relative efficacy of various treatment approaches because, as Davis and Cuvo (1980) point out, the successfully reported treatment has rarely been the initially attempted treatment. The results of these initially attempted treatments, if unsuccessful, may go unreported.

This case study reports the results of the use of two remote-controlled contingent shock devices to reduce the life-threatening vomiting and ruminating of a 12-year-old boy with profound mental retardation. A comprehensive behavior analysis had been conducted to identify possible antecedents, maintaining consequences, and high-probability setting events and times of day for the behaviors. In this case, contingent shock was used as a treatment, following the failure of less intrusive techniques based on functional analysis. These included the differential reinforcement of other behavior (DRO), differential reinforcement of incompatible behavior (DRI), planned ignoring, contingent reprimands, modified functional communication training, academic instruction, and contingent water mist. The treatment was not designed in advance to be a controlled study.

Method

Client

On 6-2-89 when Brandon arrived at Behavior Research Institute, a private non-profit residential school for persons with severe behavior problems, he was 12 years old, and weighed 68 lbs (30.84kg). He had been diagnosed as having pervasive developmental disorder, profound mental retardation, seizure disorder, and tuberous sclerosis. He had scored at the 4- to 11-month level on the Bailey Scale of Infant Development two years prior. The Vineland Adaptive Behavior Scales indicated age equivalencies of 9 months, 20 months, and less than one month for the Communication Domain, the Daily Living Skills Domain, and the Socialization Domain, respectively. He was completely unable to work independently on educational tasks, and needed continual hands-on prompting to maintain any participation. Brandon had no expressive language other than reaching toward something he wanted, or leading someone by the arm to what he wanted. In terms of receptive language, he could not follow any verbal directions. His medication at the time of admission was phenytoin 150mg per day and haloperidol 6mg per day.

His vomiting and ruminating were frequent throughout his childhood. Reports obtained from his previous educational placements indicate that ruminating was common following meals, and occurred continuously throughout the day. His father reported that he vomited five or six times per day at the age of 10.

Brandon also exhibited many other inappropriate behaviors. His stereotypic behaviors included hand-flapping, hand-gazing, and twirling. His severe self-injurious behaviors (SIB) included biting, scratching, pinching, and hitting himself; these began occurring when he was less than a year old. He has shown unprovoked aggression, property destruction, eating inedible objects, jumping in the air and landing on his knees, and running away from those supervising him. This is not an exhaustive list of his inappropriate behaviors. Haloperidol had been unsuccessful in remediating any of these inappropriate behaviors at his previous placement

Functional Analysis

Brandon’s records, behavioral observations during a baseline data recording period, interviews with staff and parents, and completion of several functional analysis checklists, suggested that vomiting and rumination were functioning to obtain both sensory stimulation and escape from demands.

The Motivational Assessment Scale (Durand and Crimmins, 1988) showed that the behaviors of vomiting and ruminating scored highest in the category of "sensory." A scatterplot (Touchette, McDonald, and Langer, 1985) of vomiting and ruminating frequencies showed that they occurred more often at times associated with eating; however, they also occurred throughout the whole day.

Behavioral observations and interviews with staff indicated that Brandon would sometimes purposely spit vomitus at them, apparently in an attempt to get them to move away from him when they were trying to carry out teaching protocols. He also vomited and ruminated when left completely alone without demands or the close proximity of any other people.

Thorough medical examinations (pediatric, neurological, and gastroenterological) did not identify any medical reasons for the occurrence of vomiting and ruminating. Brandon was examined each day by a nurse who closely monitored his health.

Treatment Description

Baseline — Phase 1. Data were collected in the form of frequency counts during all waking hours from the time of Brandon’s admission on 6-2-89 through 6-14-89. All inappropriate behavior topographies were recorded during this time by a one-to-one staff person assigned for this purpose. Demands were relatively low, although there were occasional probes in which tasks were presented. No planned consequences were in effect during this period for either appropriate or inappropriate behaviors. For the first several days after admission Brandon wore athletic socks over his hands to soften slaps to his face. At other times during baseline, if Brandon’s SIB appeared dangerous, staff either blocked the behavior, manually restrained his arms, or placed a hockey helmet with face guard on him. Vomiting and ruminating were ignored. The tapering of haloperidol was begun immediately upon admission, with the total daily dosage being reduced by 0.5mg each week until it was discontinued. This process was complete by late August, 1989.

All of Brandon’s inappropriate behaviors were targeted for treatment after approximately two weeks of a baseline period. At this time, vomiting and ruminating posed no severe health risk, his weight was stable, and his appetite was good. Although vomiting and ruminating had initially been targeted and charted as separate behavior topographies, in July of 1989 they were collapsed into the category of "major health dangerous behaviors," along with a number of other topographies. In February of 1990, however, ruminating developed into a major problem behavior, and separate recording and charting of its frequency was resumed. In August of 1990 vomiting had likewise become a major problem, and separate recording and charting of it was also resumed. Both vomiting and ruminating continued to be recorded and charted separately throughout the remainder of the study.

Until his vomiting and ruminating worsened significantly in February of 1990, the most serious behaviors that Brandon exhibited were hitting, pinching, and biting himself. Prior to entering BRI, Brandon had accumulated self-inflicted scars all over his body from these self-injurious behaviors during his life. Treatment had been primarily geared to gain control of these behaviors in Brandon’s first eight months at BRI. Significant incidents at BRI had included numerous occasions of biting his tongue, repeatedly vomiting directly on staff, in some cases causing infections and skin rashes, and self-injurious hitting and pinching which had resulted in numerous cuts, bruises, and swelling.

General Educative/Positive Programming Systems. After baseline, Brandon was started on a full program of teaching, as specified in his Individual Educational Plan. This consisted of a number of self-care tasks, receptive and expressive communication tasks (e.g., signing, receptive labeling, pointing), prevocational tasks, and computer pointing tasks. Task analyzed sequences with graduated prompting and reinforcement were used in teaching protocols.

Reinforcement systems were always in effect, and selecting the most powerful reinforcers was an ongoing process. Positive programming included verbal and physical social reinforcers, contingent on successful task performance (differential reinforcement of alternate behavior, DRA) and the absence of inappropriate behaviors, at the rate of approximately 20 or more in a 5-minute period. In addition to this, he earned reinforcers for the absence of certain major inappropriate behaviors for some fixed length of time (differential reinforcement of other behavior, DRO "contracts"). Early in treatment the length of these DRO intervals ranged anywhere from 30 seconds to 5 minutes, and were carried out by staff assigned to work with him one-to-one during all waking hours.

Because Brandon’s problem behaviors were probably motivated by more than one variable, part of the general educative treatment strategy was to design approaches which would be successful regardless of the function of the behaviors, and modify these strategies if they later seemed to be contraindicated. This was accomplished by: (1) ensuring that Brandon had a high rate of social attention contingent on appropriate behavior and not inappropriate behavior; (2) providing a functional communication token system to allow Brandon an appropriate means to escape from demands, seek attention, or request a tangible item, along with instruction and prompting in exchanging these for what he wanted (Carr & Durand, 1985); (3) minimizing staff attention contingent on inappropriate behaviors; (4) never permitting escape from demands to be contingent on inappropriate behavior (Iwata et al., 1990).

Due to the constant effort to fine-tune the details of these general positive programming treatment systems, it would be impossible to report on and indicate the causal contributions of all of these manipulations. The best that can be said is that there was an ongoing effort to maximize reinforcement for alternative behavior and the absence of inappropriate behavior, adjust the level of demands to Brandon’s individual tolerance level, and select tasks on which he could be successful. This process occurred throughout all phases of treatment.

Contingent Water Mist — Phase 2. During this treatment phase, contingent on vomiting or ruminating, a three-second application of water mist, applied with compressed air, was administered to the face. The compressed air was supplied through a hose and was under a pressure of 310,264 N/m2 (45 psi), which was the lowest pressure setting that would pull the water out of the holding container located at the end of the hose. It was applied in a back-and-forth fashion at a point below the bridge of the nose. The water mist procedure used by Dorsey et al. (1980), of which this procedure is an extension, had been applied with a hand-operated plant sprayer.

Contingent Electrical Stimulation #1 — Phase 3. By February of 1990 ruminating had become a serious problem. Treatment was resumed with contingent electrical stimulation, and separate charting for ruminating was also begun at this time. Vomiting was consequated in the same manner, but was still part of the behavior category "major health dangerous behavior," and was not charted separately. Separate charting for vomiting resumed in August, 1990.

The Self-Injurious Behavior Inhibiting System (SIBIS; commercially available from Human Technologies, Inc., 300 3rd Avenue North, St. Petersburg, FL 33701) was used as the potentially decelerating consequence. Staff members used a hand-held remote triggering device to activate the SIBIS. Other topographies consequated with SIBIS included major self-injurious and destructive behaviors, and aggression.

The SIBIS is described in more detail elsewhere by Linscheid, Iwata, Ricketts, Williams, and Griffin (1990). Linscheid et al. (1990) list SIBIS as delivering a 3.5 mA stimulation through a 24K ohm fixed resistor. This actually consists of 16 rectangular-wave pulses, separated by an equal off-time, throughout a shock "stimulation" period of 0.2 s. Oscilloscopic testing of our SIBIS through a 24K ohm resistor revealed a peak current of 5.4 mA; this is approximately 1.5 times the specification value. Our device had a rectangular-wave pulse duration of .0063 s, with an equal off-time for each cycle (50% duty cycle).

We prefer to measure the shock stimulus in terms of "average current" during the stimulation period, since subjective pain is a function of both current and time (Brull and Silverman, 1995) . Due to the fact that, during a SIBIS stimulation, the current is on for half the time and off for half the time, we will refer to the SIBIS intensity level as being an "average current" of 2.7 mA (50% of 5.4 = 2.7) through a 24K ohm fixed resistor. Subsequently specified levels of electrical stimulation will be referenced to this same standard of measurement, average current through a 24K ohm fixed resistor.

Mild Verbal Reprimand — Phase 4. During this condition the use of SIBIS was terminated because it was not functioning to suppress the targeted behaviors. Instead, vomiting, ruminating, and all other major inappropriate behaviors were consequated with a mild verbal reprimand (e.g., "No vomiting!"). For his own protection, Brandon required restraint equipment virtually 24 hrs per day. These were used as needed, and consisted of a hockey helmet with full face guard, cuffs which held his wrists at his waist, gloves or mitts, and a specially designed mouth device to prevent the biting of his tongue, part of which he had bitten off in July of 1990.

After approximately seven weeks under these conditions he required a hospital admission to treat life-threatening weight loss, dehydration, and fecal impaction arising from vomiting and ruminating. During this hospital admission, testing and evaluation failed to disclose any medical cause, nor indicate any medical treatment for these behaviors. After Brandon’s condition was stabilized, he was discharged back to BRI.

Contingent Electrical Stimulation #2 — Phase 5. BRI, in conjunction with a consulting electrical engineer, designed its own remote-controlled electrical stimulation device called the Graduated Electronic Decelerator (GED). The technical details of this device are reported in Israel, von Heyn, Connolly, & Marsh (1992). The GED had been in development and was not ready for use until the beginning of Phase 5. During this phase of the treatment, vomiting and ruminating, along with major self-injurious and aggressive topographies, were consequated with an electrical stimulation of 2.1 mA average current (8.3 mA peak, 25% duty cycle, through 24K ohm fixed resistor), for a duration of 0.2 s. This level was selected because it was believed to be somewhat similar to the intensity of a SIBIS stimulation.

The area of the body to which stimulations were applied were varied within each day by periodically changing the position of the electrode placement, and included the inner bicep, inner thigh, and stomach.

With the beginning of phase 5 another behavior was added for treatment, which was referred to as "refuse-to-swallow." Brandon had developed the behavior of holding medications, food, and even ruminated material in his mouth for extended periods. When staff observed that he was holding material in his mouth, he was given the verbal instruction, "Swallow." If he did not, he received a GED stimulation. He learned to swallow very quickly, and did not receive many GED stimulations for this behavior. Throughout the remainder of the study, whenever vomiting and ruminating were treated with electrical stimulation, "refuse-to-swallow" was also.

Contingent Electrical Stimulation #3 — Phase 6. The same treatment conditions were in effect during this time, with two exceptions. The electrical stimulation was increased to 3.1 mA average current (12.5 mA peak, 25% duty cycle, through 24K ohm fixed resistor). Brandon also wore from three to five electrodes at any one time, which were located on different areas of his body. Although only one electrical stimulation was applied, contingent on a targeted behavior, the electrode through which the stimulation would come was completely unpredictable. This presumably helped to prevent any amelioration of the stimulation by his preparatory body responses prior to the receiving of a stimulation. Electrode placement sites included the stomach, inner upper and lower arms, inner thighs, the calf, and the bottom of the foot. Exact electrode placements on body areas were varied within a day to prevent any physical side effects.

Mild Verbal Reprimand — Phase 7. This treatment condition is the same as previously described. All behaviors which had been consequated with the GED received a mild reprimand. These included vomiting, ruminating, major self-injurious behaviors, and aggression.

Contingent Electrical Stimulation #4 — Phase 8. During this treatment condition vomiting, ruminating, and major self-injurious behaviors were consequated with the GED at a level of 4.2 mA average current (16.7 mA peak, 25% duty cycle, through 24K ohm fixed resistor) and a duration of 2.0 s. Aggressive and destructive behaviors continued to receive mild reprimands. These conditions were maintained throughout the remainder of treatment reported here, through December of 1991.

Medication Changes During Study. Medication manipulations were not made for the express purpose of controlling behavior; however, the following information is presented in order to provide a complete picture of variables that might have had an effect on behavior.

When Brandon was admitted he was taking phenytoin (150 mg/day) for a seizure disorder, and haloperidol (6 mg/day) for behavior. Reduction of haloperidol began immediately, with decreases in daily dosage of 0.5 mg each week for 11 weeks, until it was completely discontinued.

The following changes in phenytoin occurred: 7-27-90, increased to 200 mg/day; 8-9-90, reduced to 175 mg/day; 2-13-91, began tapering off with decreases in daily dosages of 25 mg/day each week, for six weeks until completely discontinued.

Carbamazepine (100 mg/day) was started for the treatment of seizures on 9-24-90. Dosage changes occurred as follows: 9-27-90, 200 mg/day; 10-9-90, 400 mg/day; 10-24-90, 800 mg/day; 12-15-90, 1000 mg/day; 1-9-91, 1200 mg/day; 1-21-91, 1600 mg/day; 2-7-91, 1800 mg/day; 2-9-91, 1600 mg/day, where it remained throughout the rest of the study.

Metoclopramide hydrochloride, a medication used for the treatment of gastroesophogeal reflux, nausea, and vomiting, was administered at the dosage of 6 mg/day from 11-10-90 through 1-10-91.

Safeguard of Human Rights

During Brandon’s treatment at BRI, the use of all restrictive or aversive procedures followed strict ethical and legal guidelines. Informed consent for all restrictive procedures was given by Brandon’s parents, along with permission to report the results of treatment to the professional community. Procedures were reviewed and approved by human rights and peer review committees. Final authorization for the use of all restrictive and aversive procedures was obtained from a judge’s order following a "substituted judgment" hearing in a Massachusetts probate court.

Results

Figure 1 shows graphs for weight, ruminating, and vomiting in Panels A, B, and C, respectively. They are plotted as a function of time in weeks, and cover from the time of his admission in June of 1989 through December 1991. Panels B and C plot frequencies of ruminating and vomiting following the conventions of the Standard Behavior Chart (Lindsley, 1971; Pennypacker, Koenig, & Lindsley, 1972). Basically these conventions require that a certain proportional relationship be maintained between the x- and y-axes, and that the frequency of the dependent variable be plotted on a 6-cycle logarithmic y-axis. A relatively small upward or downward movement on this type of chart represents a relatively large change proportionally. The underlying assumption of this charting methodology is that behavior rates change in a multiplicative or divisive fashion, rather than additive or subtractive. Weight is plotted on a standard-axis graph in Panel A.

Figure 1. Body weight (Panel A), rumination frequency (Panel B), and vomiting frequency (Panel C), as a function of weeks. Some weeks are missing because ruminating and vomiting had not been tracked as separate behavioral topographies. Note that there is a break in the abscissa between July of 1989 and February of 1990. The behavior frequencies on Panels B and C are plotted on a logarithmic scale following the format of the Standard Behavior Chart. 

At the time of admission, Brandon's weight was 68 lbs. (30.84 kg), and his baseline frequencies per week for each of the behaviors, ruminating and vomiting, were well over 1000. Panels A and B Phase 2 show that contingent water mist had a significantly reductive effect on both ruminating and vomiting; however, the frequencies still remained extremely high at between 500-800 per week.

Phase 3 in Panel B shows that the use of SIBIS had significantly lowered ruminating to an average of about 60 per week during the first 15 weeks of its use. Compared to baseline, this represents a 96% reduction, or, in Standard Behavior Chart terminology, a "divide-by-23." This reduction remained stable until June of 1990 when the frequency jumped up drastically. Although there was considerable variability, the overall trend from February to September of 1990 was one of acceleration, at the rate of "times 1.6" per month, meaning that the frequency multiplied by a factor of 1.6 from month to month.

Even though rumination levels were still far below baseline, the overall accelerating trend, together with the fact that SIBIS was not adequately decelerating the other major self-injurious behaviors (not shown here), led to the decision to terminate the use of SIBIS, and to design a more effective electrical stimulation device to take its place. The results of terminating the SIBIS consequence are shown in Phase 4 of Figure 1.

During the two weeks prior to terminating the use of SIBIS, the behavior of vomiting was removed from the category of "major health-dangerous behaviors" and charted separately. Phase 3 of Panel C shows that vomiting was occurring an average of about 350 times per week. Brandon's weight had dropped to 58 lbs. (26.31kg) during the last week of Phase 3.

For the first three weeks of Phase 4, ruminating and vomiting did not change significantly from where they were near the end of Phase 3. His weight increased about five lbs (2.27 kg).

During the fourth week of Phase 4, ruminating and vomiting frequencies jumped down, then subsequently drastically accelerated for the remainder of Phase 4. Brandon's weight dropped to 52 lbs (23.59 kg). The data do not show it, but subjectively it appeared that the quantity that was being vomited during Phase 4 was of larger volume. He would not keep down even sips of water. During the last week in November, 1990, he ruminated approximately 1600 times and vomited close to 5000 times. After seven weeks of being off treatment with SIBIS, Brandon required hospitalization for dehydration and bowel impaction. X-rays revealed no blockage, and the impaction was likely due to lack of available fluids for moving the stool through his bowel. Figure 2 (left) shows a picture of Brandon taken when he was at his lowest weight. One can see his ribs protruding through his skin and that his legs are extremely thin.

Figure 2. Pictures of Brandon at his lowest weight (left), and after approximately three weeks into treatment of ruminating and vomiting with contingent electrical stimulation provided by the GED.

Phase 5 shows that the introduction of the GED at 2.1 mA caused an immediate reduction in rumination by a factor of 60 (a divide-by-60 jump-down), followed by a slight increase over the next three weeks. At the same time vomiting showed a reduction by a factor of 200 (divide-by-200 jump-down), and maintained at around 10 per week. Figure 2 (right) shows him approximately three weeks into Phase 5 after treatment of ruminating and vomiting with contingent electrical stimulation provided by the GED. His ribs were no longer visible through his skin. Because vomiting showed no further deceleration, rumination was showing a slight accelerating trend, and the other major self-injurious behaviors (not shown) were showing a clear-cut acceleration, the intensity of the GED electrical stimulation was raised to 3.1 mA in Phase 6.

Phase 6 shows further deceleration of ruminating and vomiting with the increased GED intensity (decelerations of divide-by-3.2 and divide-by-8, respectively). Despite these improvements in Brandon’s condition, the increased GED intensity still did not bring down the other major health-dangerous behaviors. Brandon continued to require protective devices a good deal of the time, which prevented freedom of movement of his arms and hands, and required that his head be encompassed in a helmet. Accordingly, the intensity and duration of the electrical stimulation were subsequently increased in Phase 8.

Prior to increasing the intensity and duration of the GED consequence, Brandon was taken off all contingent electrical stimulation. The purpose of this was to allow any adaptation effects to electrical stimulation to subside, thus increasing the contrast between two adjacent treatment phases and increasing the chances for success.

Phase 7 shows the resulting dramatic jump-ups for both ruminating ("times 250") and vomiting ("times 125") during the time that the consequence was changed back to a mild reprimand.

In Phase 8, when the GED intensity level was increased to 4.2 mA, and the duration was increased to 2.0 s, rumination showed a divide-by-85 jump-

down, followed by a divide-by-3 monthly deceleration through April. Vomiting showed a divide-by-160 jump-down, followed by a divide-by-2.1 monthly deceleration through April. Panel B shows that ruminating stayed at 3 or less per week during March and April, with the exception of one week in which the frequency was 10 and a second week in which the frequency was seven. Panel C Phase 8 shows that vomiting decelerated even further than Phase 6 levels. From March through December, 20 of the 42 weekly points were at zero, and 19 of the weekly points were at a frequency of either one or two; the highest weekly point was at seven, during the first week of May, and the remaining two were at frequencies of four. Some data points are missing for late November and early December due to Brandon going on a home visit.

From May to November of 1991, ruminating had accelerated to an average of 24 per week for a number of weeks, but subsequently decelerated beginning in October. This may have been due to the increased demands of fire-drill training at Brandon’s residence. Vomiting continued to remain low during this period.

Panel A of Figure 1 shows that weight increased with the onset of the use of GED to treat ruminating and vomiting in Phase 5. It increased further when the GED parameters were increased to 4.2 mA, 2.0 s duration. His weight continued to increase into the low 80’s (approximately 37kg), as shown in Phase 8 of Panel A. This is a 56% increase from his lowest weight of 52lbs (23.6kg) in late October and early November of 1990. Some weight data are missing from August and September of 1991.

Brandon presently requires several types of protective devices for his hands, to prevent tissue damage from the self-injurious behaviors of rubbing or pressurizing his finger joints on surfaces of objects. However, his general health is good, he never requires a helmet, and he no longer has the smell of vomitus about him, as many persons do who ruminate. This makes him much more pleasant to be around, and others do not shun physical contact with him.

In terms of positive social side-effects, Brandon’s mood has been good, once effective treatment of inappropriate behaviors was achieved. He is happy, is exhibiting communicative gestures and pre-speech sounds, and seeks and accepts affectionate hugging and touching without attempts at self-restraint. He has also been much better at accepting some simple training demands, such as self-feeding with a fork, dressing, basic hygiene, and a simple operant response to computer generated shapes. However, he still has a long way to go in order to achieve functionally significant daily living skills.

Present data show that he is judged as being in a happy mood during approximately 50 15-minute intervals in a day, and irritable or angry during approximately five 15-minute intervals. Graphs were not presented for these data because mood ratings were not done until after treatment with GED began. However, subjective reports of staff who worked with him shortly after admission indicated that he had been in an irritable mood virtually all the time.

His other major health-dangerous behavior topographies (not shown in Figure 1) occurred at an average of 10.4 per day through November and December of 1991. This compares to approximately 160 per day upon admission.

Discussion

The data presented in this case study showed that life-threatening weight loss, arising from ruminating and vomiting of behavioral etiology, was reversed when these behaviors were treated with an effective level of contingent electrical stimulation. The effects of four different intensity/duration levels of electrical stimulation were shown, and the results demonstrated that the best suppression was achieved at the highest levels of intensity and duration.

These results are basically similar to those of Bright & Whaley (1968), Galbraith, Byrick, & Rutledge (1970), and Watkins (1972), in that ruminating and/or vomiting were reduced to zero or near-zero when treated with contingent electrical stimulation, and that a weight increase was observed in the person treated.

The results also demonstrate the recovery and subsequent slow acceleration of a behavior that was not reduced to near-zero levels when treated with the lowest intensity of contingent electrical stimulation (Figure 1, Panel B, Phase 3). This kind of phenomenon is consistent with laboratory research studies of the effects of mild, moderate, and severe punishment on suppression and recovery of responding maintained by variable-interval schedules of reinforcement (Azrin, 1960; Appel & Peterson, 1965). Briefly, if a punishing consequence is not strong enough to suppress a behavior to near-zero levels, the rate of responding recovers until, in some cases, it goes back to the original baseline level. In the present study, a higher intensity and duration of electrical stimulation proved effective where weaker and shorter ones had not.

Our finding of positive social side-effects is consistent with Bright & Whaley (1968) and Luckey, Watson, & Musick (1968). However, Galbraith, Byrick, & Rutledge (1970) and White & Taylor (1967) reported negative side-effects in the form of increased undesirable behaviors and increased emotional behaviors, respectively.

Regarding medication changes made during the course of the study, one of the possible side effects of carbamazepine is nausea and vomiting (Physicians’ Desk Reference, 43rd ed., 1989). This raises the possibility that the onset of carbamazepine therapy, beginning at the second graph point in Phase 4, might have caused the ruminating and vomiting. This is an extremely unlikely possibility for the following reasons.

Brandon’s attending physician (E. A. Sassaman, personal communication, April 8, 1992) reported that treatment with carbamazepine would not have caused vomiting to the extent that Brandon was exhibiting it, and that if it were the cause, he would not have adapted to its effects. Therapy was started at a low dosage and titrated upward to therapeutic levels in order to minimize such reactions. He presently continues on carbamazepine at a dosage of 1600 mg/day with no apparent side effects, no vomiting, and very little rumination.

The acceleration of ruminating seen in Phase 4 appears to be a continuation of the acceleration seen in Phase 3. The extreme and immediate reductions seen in Phase 5 would be highly unlikely if these behaviors had arisen from organic causes. The extreme jump-ups in ruminating and vomiting frequencies in Phase 7, and their subsequent jump-downs in Phase 8, strongly suggest that the behaviors were under environmental, rather than organic, control.

Institution of treatment with metoclopramide hydrochloride prior to the eighth graph point in Phase 4 was followed by two subsequent weeks of increases in both ruminating and vomiting. After discontinuation of this medication, prior to the third graph point in Phase 6, both ruminating and vomiting continued to decelerate. Both of these effects are contrary to what would be expected if this anti-emetic medication were exerting control on an organic condition.

If Brandon’s behavioral treatment had been confined strictly to the use of a commercially available electrical stimulation device specifically designed for use in human behavior modification, success would very likely not have been achieved. His life-threatening weight loss, by necessity, would have been treated exclusively by medical methods. According to his physician, this would most likely have involved chronic full restraint, heavy sedation, and administration of nourishment and fluids through a gastro-tube surgically implanted in the stomach. Treatment with the GED resulted in a reversal of Brandon’s life-threatening weight loss, without having to resort to such an extreme medical solution.

This report has concentrated on the behaviors of ruminating and vomiting because of their life-threatening nature for Brandon; however, these were only two of many behaviors being treated. The entire treatment of all his major inappropriate behaviors was extremely complex, due to the myriad numbers of self-injurious, destructive, aggressive, noncompliant, and disruptive topographies. Also, Brandon’s treatment was a 24-hours-per-day, 365-days-per-year endeavor. His was not a simple case of a client with one predominant topography of dangerous behavior. Overall success with him would never have been achieved by confining treatment to only one of these topographies, or even to only one overall category of problem behaviors.

Several variables warrant further investigation regarding the use of contingent electrical stimulation as a punisher. These involve the location of electrode placement on the body, the varying of the location of electrode placement, duration of electrical stimulation, and the number of electrodes that are in place at any one time through which a stimulation might occur. Anecdotally, we believe that the manipulation of these variables helped in Brandon’s case, especially in October of 1991; however, the study was not designed in advance to provide evidence of this conclusion. It is quite possible that the manipulation of one or all of these variables could bring about a successful treatment in cases that have not responded to the basic one-dimensional use of contingent electrical stimulation.

References

Appel, J.B., & Peterson, N.J. (1965). Punishment: Effects of shock intensity on response suppression. Psychological Reports, 16, 721-730.

Azrin, N.H. (1960). Effects of punishment intensity during variable-interval reinforcement. Journal of the Experimental Analysis of Behavior, 3, 123-142.

Azrin, N.H., & Wesolowski, M.D. (1975). Eliminating habitual vomiting in a retarded adult by positive practice and self-correction. Journal of Behavior Therapy and Experimental Psychiatry, 6, 145-148.

Ball, T. S., Hendricksen, H., & Clayton, J. (1974). A special feeding technique for chronic regurgitation. American Journal of Mental Deficiency, 78, 486–493.

Becker, J.V., Turner, S.M., & Sajwaj, T.E. (1978). Multiple behavioral effects of the use of lemon juice with a ruminating toddler-age child. Behavior Modification, 2, 267-268.

Bright, G.O., & Whaley, D.L. (1968). Suppression of regurgitation and rumination with aversive events. Michigan Mental Health Research Bulletin, 11, 17-20.

Brull, S.J. & Silverman, D.G. (1995). Pulse width, stimulus intensity, electrode placement, and polarity during assessment of neuromuscular block. Anesthesiology, 83(4), 702-709.

Carr, E.G., & Durand, M. (1985). Reducing behavior problems through functional communication training. Journal of Applied Behavior Analysis, 18, 111-126.

Cunningham, C.E., & Linscheid, T.R. (1976). Elimination of chronic infant ruminating by electric shock. Behavior Therapy, 7, 231-234.

Daniel, W.H. (1982). Management of chronic rumination with a contingent exercise procedure employing topographically dissimilar behavior. Journal of Behavior Therapy and Experimental Psychiatry, 13, 149-152.

Davis, P. K., & Cuvo, A. J. (1980). Chronic vomiting and rumination in intellectually normal and retarded individuals: Review and evaluation of behavioral research. Behavior Research of Severe Developmental Disabilities, 1, 31–59.

Dorsey, M.F., Iwata, B.A., Ong, P., & McSween, T.E. (1980. Treatment of self-injurious behavior using a water mist: Initial response suppression and generalization. Journal of Applied Behavior Analysis, 13, 343-353.

Duker, P.C. & Seys, D.M. (1977). Elimination of vomiting in a retarded female using restitutional overcorrection. Behavior Therapy, 8, 255-257.

Durand, V.M., & Crimmins, D.B. (1988). Identifying the variables maintaining self-injurious behavior. Journal of Autism and Developmental Disorders, 18, 99-117.

Foxx, R.M., Snyder, M.S., & Schroeder, F. (1979). A food satiation and oral hygiene punishment program to suppress chronic rumination by retarded persons. Journal of Autism and Developmental Disorders, 9, 399-412.

Galbraith, D. A., Byrick, R. J., and Rutledge, J. T. (1970). An aversive conditioning approach to the inhibition of chronic vomiting. Canadian Psychiatric Association Journal, 15, 311–313.

Hogg, J. (1982). Reduction of self-induced vomiting in a multiply handicapped girl by "lemon juice therapy" and concomitant changes in social behavior. British Journal of Clinical Psychology, 21, 227-228.

Israel, M. L., & von Heyn, R. E., Connolly, D.A., & Marsh, D. (1992). A portable electric shock device for behavior modification. Unpublished manuscript.

Iwata, B.A., Pace, G.M., Kalsher, M.I., Cowdery, G.E., Edwards, G., & Cataldo, M.F. (1990). Experimental analysis and extinction of self-injurious escape behavior. Journal of Applied Behavior Analysis, 23, 11-27.

Kanner, L. (1972). Child psychiatry. Springfield, IL: Charles Thomas.

Kohlenberg, R. J. (1970). The punishment of persistent vomiting: a case study. Journal of Applied Behavior Analysis, 3, 241–245.

Lindsley, O.R. (1971). An interview. Teaching Exceptional Children, 3, 114-119.

Linscheid, T.R., & Cunningham, C.E. (1977). A controlled demonstration of the effectiveness of electric shock in the elimination of chronic infant rumination. Journal of Applied Behavior Analysis, 10, 500.

Linscheid, T.R., Iwata, B.A., Ricketts, R.W., Williams, D.E., & Griffin, J.C. (1990). Clinical evaluation of the Self-Injurious Behavior Inhibiting System (SIBIS). Journal of Applied Behavior Analysis, 23, 53-78.

Luckey, R. E., Watson, C. M., and Musick, J. K. (1968). Aversive conditioning as a means of inhibiting vomiting and rumination. American Journal of Mental Deficiency, 73, 139–142.

Marholin, D., Luiselli, J.K., Robinson, M., & Lott, I.T. (1980). Response-contingent taste-aversion in treating chronic ruminative vomiting of institutionalised profoundly retarded children. Journal of Mental Deficiency Research, 24, 47-56.

Minness, P.M. (1980). Treatment of compulsive hand in mouth behavior in a profoundly retarded child using a sharp pinch as the aversive stimulus. Australian Journal of Developmental Disabilities, 6, 5-10.

O’Neill, P.M., White, J.L., King, C.R., & Carek, D.J. (1979). Controlling childhood rumination through differential reinforcement of other behavior. Behavior Modification, 3, 355-372.

Pennypacker, H.S., Koenig, C.H., & Lindsley, O.R. (1972). Handbook of the Standard Behavior Chart. Kansas City, Kansas: Precision Media.

Physicians’ Desk Reference (43rd ed.). (1989). Oradell, NJ: Medical Economics Company Inc.

Singh, N. N. (1981). Rumination. In N. R. Ellis (Ed.), International review of research in mental retardation (Vol. 10, pp. 139–182). New York: Academic Press.

Singh, N. N., Manning, P. J., & Angell, M. J. (1982). Effects of an oral hygiene punishment procedure on chronic rumination and collateral behaviors in monozygous twins. Journal of Applied Behavior Analysis, 15, 309–314.

Starin, S. P. & Fuqua, R. W. (1987). Rumination and vomiting in the developmentally disabled: A critical review of the behavioral, medical, and psychiatric treatment research. Research in Developmental Disabilities, 8, 575–605.

Touchette, P.E., MacDonald, R.F., & Langer, S.N. (1985). A scatter plot for identifying stimulus control of problem behavior. Journal of Applied Behavior Analysis, 18, 343-351.

Watkins, J. T. (1972). Treatment of chronic vomiting and extreme emaciation by an aversive stimulus: case study. Psychological Reports, 31, 803–805.

White, J. C. Jr. & Taylor, D. J. (1967). Noxious conditioning as a treatment for rumination. Mental Retardation, 5, 30–33.

Wright, L. & Thalassinos, P. A. (1973). Success with electroshock in habitual vomiting: report of two cases in young children. Clinical Pediatrics, 12, 594–597.

Author Notes

The authors express appreciation to Thomas Kearney for his review of the shock, rumination, and vomiting literature, to J. Michael Rock and Peter Smith for overseeing the many details of program implementation, and to Bettina Briggs and John Daignault for providing support and encouragement.

Special thanks go to Ogden Lindsley for his service as a senior consultant for overall programming issues.

Requests for reprints should be addressed to Robert W. Worsham, The Judge Rotenberg Center, 240 Turnpike St., Canton, MA 02021.

The first author organized the data and prepared the final manuscript. All authors assisted in the editing of the manuscript. The second author clinically supervised the treatment, the third author conducted the functional analyses, and the fourth author assisted in data analysis and chart preparation.

The authors express appreciation to Thomas Kearney for his review of the literature, and to J. Michael Rock and Peter Smith for overseeing the many details of program implementation. Special thanks go to Ogden Lindsley for his invaluable suggestions in the conduct of the treatment and general programming issues. Thanks also to Bettina Briggs and John Daignault for their encouragement, to Eric McLeish for successful advocacy for the treatment plan, and to the late Hon. Ernest Rotenberg, former Chief Justice of the Bristol County, Massachusetts Probate Court, who authorized the treatment plan under the "substituted judgment" hearing process.

Back to Papers on Skin Shock
Back to Main Page