The ASAM Principles of Addiction Medicine 5th Edition

62. Aversion Therapies

P. Joseph Frawley, MD and Matthew O. Howard, MD













Aversion therapy, or counterconditioning, is a powerful tool in the treatment of alcohol and other drug addiction. Its goal is to reduce or eliminate the “hedonic memory” or craving for a drug and to simultaneously develop a distaste and avoidance response to the substance. Unlike punishments (jail, firings, fines, divorce, hangovers, cirrhosis, and the like), which often are delayed in time from the use episode, aversion therapy relies on the immediate association of the sight, smell, taste, and act of using the substance with an unpleasant or “aversive” experience.

This treatment is not designed to appeal to the logical part of the individual’s brain, which often is all too aware of the negative consequences of alcohol and other drug use, but to the part of the brain where emotional attachments are made or broken through experienced associations of pleasure or discomfort. Aversion therapy provides a means of achieving control over injurious behavior for a period of time, during which alternative and more rewarding modes of response can be established and strengthened (1). It is first reported to be used in America by Benjamin Rush, a physician, in 1789 (2).

People need care—behavior needs modification. It is important not to confuse aversion with punishment. In punishment, it is the individual who receives the negative consequence, whereas in aversion therapy the negative consequence is only paired with the act of using a drug. This has a very important benefit to self-esteem. While the patient is engaging in positive recovery activities, he or she is receiving immediate positive support for a new way of behaving and thinking. It is only when the patient is engaging in an old behavior—alcohol or drug use—that he or she experiences immediate and consistent discomfort. Hence, self-esteem is rebuilt by separating the drug from the self (3).

In nonaddicted populations, hangovers have been cited as a significant reason to cut down or stop drinking (4). However, the hangover is delayed in time from the actual use of alcohol; thus, for the alcoholic, who drinks for the immediate euphorogenic effects of alcohol, a hangover often is ineffective in producing aversion because it is delayed in time from use of the substance whose immediate effect was experienced as pleasant or euphoric. Moreover, the discomfort of a hangover, though logically understood to be the result of drinking, is blamed on “drinking too much” (weakness) rather than drinking at all (disease). Even worse, alcohol may be used to cure the withdrawal (“hair of the dog”), which ensures that emotionally the patient perceives the alcohol as a solution, not a problem.

Contrary to popular belief, disulfiram (Antabuse) is not an aversion treatment. In aversion therapy for alcohol addiction, alcohol is not absorbed into the system (3). With disulfiram, alcohol must be absorbed and metabolism begun for it to produce its toxic effect (5). Aversion relies on safe but uncomfortable experiences that can be repeated, whereas disulfiram reactions can be life threatening, even in healthy persons. For this reason, patients today are not given alcohol at the same time that they are prescribed disulfiram. As a result, they have not actually experienced a disulfiram reaction. Thus, disulfiram does not change the way the addict feels about alcohol. He or she may fear the consequence of drinking, just as he or she fears being arrested for drinking and driving; nevertheless, he or she still retains the euphoric recall of past episodes of drinking alcohol and hence the craving for the alcohol itself. Aversion works to eliminate or reduce euphoric recall by recording new negative experiences with the drug (6).

In fact, spontaneous aversions are common, because the capacity to develop aversions is a biologic defense mechanism. DeSilva and Rachman (7) published a study of 125 students and hospital employees, 105 (84%) of whom had a history of natural aversions. There were an average of 3.5 aversions per person (females more than males), and 70.1% had been present since childhood.


Ivan Pavlov noted that the repetitive pairing of a bell with food soon led to a “conditioned response” in dogs, who salivated at the sound of the bell alone, even with no food present. This type of pairing or training is called classical conditioning.

B.F. Skinner expounded on the observation that the nervous system is so constructed that organisms will reduce or avoid behavior that is consistently paired with negative consequences and will increase behavior that is rewarded. This type of learning is called operant conditioning. Both types of learning can be shown to occur in addiction (8). Aversion therapy uses these principles in reversing the drug-rewarded learning and conditioned reflex to seek drugs.

The development of an aversion can be very specific. Inadequate treatment can occur when aversion is developed only to one type of alcoholic beverage (9,10). In professional alcohol addiction treatment, for example, 50% of trials may be with the addict’s favorite brand or type of liquor, but the other trials include a range of alcoholic beverages (3,11).

Repetition is an essential part of training and conditioning (12). Adequate trials are needed to develop an aversion (11) and to maintain and reinforce it to prevent extinction (13,14).

Addicts already have been conditioned by the drug prior to entry into treatment. Studies have shown that alcoholics increase the number of swallows and amount of salivation in response to the sight of alcohol, as compared to nonalcoholics (15). Studies of smokers seeking to quit show that those who are least likely to quit have a much larger conditioned drop in pulse (presumably to compensate for the increase in pulse rate caused by smoking) when presented with a cigarette (16). Cocaine-dependent addicts experience progressively steeper drops in skin temperature and increased galvanic skin response (a sign of arousal) when viewing progressively more intense and explicit pictures of cocaine use. These responses can be shown to decay in strength as time away from the drug increases.

The presence of these phenomena suggests that one of the consequences of addiction is that the body becomes conditioned to drink or use drugs in the presence of certain stimuli. This may contribute to the sensation of physical craving experienced by addicts. The availability of drugs such as heroin or cocaine in the environment also influences craving (1719).


Aversion Therapy in Smoking Cessation

It is estimated that more than 80% of smokers wish to quit but feel compelled to continue smoking because of the difficulty of stopping; indeed, the annual spontaneous recovery rate for smokers is less than 5% (20). Sachs (20,21) reviewed modern smoking cessation treatments and concluded that programs that use rapid smoking aversion or satiation had superior outcomes. Rapid smoking or satiation involves smoking cigarettes with inhalations every 6 seconds. Sessions last an average of 15 minutes, and the subject smokes an average of five cigarettes (22). The treatment sessions are usually daily for 5 days with a tapering frequency of booster treatments after that. When compared to the physical effects of normally paced smoking, clients undergoing rapid smoking experience increased burning in the lungs, palpitations, facial flush, headache, and feeling faint or weak (23). Hall et al. (24) and Lando (25) reported in separate studies that the best results were reported by programs in which aversion was combined with several other modalities, including relapse prevention, relaxation training, written exercises, contract management, booster sessions of aversion, and group support. Hall et al. (26) found that skills training (cue-produced relaxation training, commitment-enhanced training, and relapse prevention training) had more of an effect than the type of aversive smoking on outcome. Conversely, in a study of 18 patients with cardiopulmonary disease, Hall et al. (22) reported that those in a waiting list control group had no abstinence as compared with those treated with satiation aversion, 50% of whom achieved 2-year abstinence. Their study of 18 patients with cardiopulmonary disease who underwent satiation treatment also found no myocardial ischemia or significant arrhythmia in this group. Five patients with ischemic changes on the treadmill did not experience the changes during the satiation treatment.

Much of the research on aversion therapy for smoking cessation has focused on improved outcomes with aversive smoking (puffing or inhaling smoke from the cigarette in a rapid manner to induce nicotine toxicity, often including nausea) (27). Though nicotine is taken into the system during this treatment, the aversion developed to smoking is adequate to prevent relapse despite the transient presence of nicotine in the bloodstream during treatment. A Cochrane Review of 25 trials of aversions therapy for smoking reported that “the existing studies provide insufficient evidence to determine the efficacy of rapid smoking, or whether there is a dose response to aversive stimulation. Milder versions of aversive smoking seem to lack specific efficacy. Rapid smoking is an unproven method with sufficient indications of promise to warrant evaluation using modern rigorous methodology” (28). Faradic aversion (mild electrical stimulus applied to the forearm) has been used commercially for smoking cessation since 1972 (29). With faradic aversion, the smoke is not inhaled but merely puffed. Inhaling during faradic aversion may lead to early relapse because of maintenance of the nicotine dependence (30).

One advantage of this form of treatment is that less medically sophisticated staff can supervise the administration of the treatment. In both forms of treatment, patients personally administer the aversive agent (rapid smoking or electrical stimulus) to themselves, while the therapist serves as a coach. In the case of faradic aversion, each time a patient brings a cigarette toward his or her lips, a mild electrical stimulus is administered automatically by a 9-V battery. The stimulus is activated by a string attached to the smoker’s wrist. The therapist also instructs the patient in relapse prevention methods and behavior and dietary changes that help maintain abstinence and achieve comfort during the initial period after smoking cessation. Smith (29) contacted 59% of 556 patients treated with this method in a commercial program and found that 52% had achieved continuous abstinence at 1 year.

Aversion Therapy for Alcohol Addiction

The spontaneous recovery rate for alcoholism is influenced by a variety of factors, including severity of problems, age, and the presence of a co-occurring psychiatric disorder (31). Nevertheless, the annual spontaneous recovery rate estimated by Vaillant in his study of the natural history of alcoholism was 2% to 3% (32). In 1949, Voegtlin and Broz (33) published a 10.5-year follow-up of 3,125 patients treated with aversion therapy. One-year abstinence rates of 70% were reported using chemical aversion therapy with minimal counseling.

There are three well-conducted controlled trials of aversion therapy for alcohol dependence. Boland et al. (34) evaluated the 6-month abstinence rates for 50 lower socioeconomic alcoholics, using lithium as a chemical aversive agent. Twenty-five patients given emetic aversion with lithium had 36% total abstinence, as compared to 12% for the 25 patients given control treatment (p < 0.05). Cannon et al. (35) divided 20 Veterans Administration patients into three groups and found that there was little difference in outcome between seven patients given chemical aversion and those given control treatment (both groups, however, had extremely high abstinence rates: 170 to 180 days for the chemical aversion group vs. 158 to 180 days for the control group). Both did better than seven patients receiving faradic aversion. However, the groups were small, there may have been some ceiling effect, and the subjects had to drink some alcohol during the actual testing sessions, which could counteract any aversion being developed.

In the private sector, Smith et al. (14) compared 249 inpatients receiving aversion therapy as part of a multimo-dality treatment program with 249 inpatients from a large treatment registry of more than 9,000 patients receiving multimodality treatment but without aversion therapy. All were matched on 17 baseline characteristics. Of the patients receiving aversion therapy, 84.7% had total abstinence from alcohol at 6 months, compared with 72.2% in the control group (p < 0.01); at 1 year, 79% of those treated with aversion had maintained abstinence, versus 67% of those without such treatment (p < 0.05; Table 62-1). The group showing the greatest benefit from aversion therapy was the daily drinkers (84% vs. 67%; p < 0.001).



From Smith JW, Frawley PJ, Polissar L. Six- and twelve-month abstinence rates in inpatient alcoholics treated with either faradic aversion or chemical aversion compared with matched inpatients from a treatment registry. J Addict Dis1997;16(1):5–24.

Though the majority of patients in the study were treated with chemical aversion therapy, a subsample of 28 patients received faradic aversion instead. The decision to prescribe faradic aversion instead of chemical aversion within clinical practice usually is based on the patient’s medical condition. If the patient has a medical contraindication to chemical aversion therapy, then faradic is prescribed. Jackson and Smith (36) reported that patients selected in this way had nearly identical abstinence rates. In a more recent study, the same results were found (37). Blake (38) found that alcoholics treated with relaxation training, motivational arousal, and faradic aversive conditioning had a 6-month follow-up of 100% and an abstinence rate of 54% and a 12-month follow-up rate of 100% and an abstinence rate of 52%.

Neueberger et al. (39) found 1-year abstinence rates of 50% to 56% in patients treated with chemical aversion therapy. As a replication study, Wiens and Menustik (40) found total abstinence rates of 63% in patients treated with chemical aversion therapy. A meta-analysis by Thurber (41) of six studies on chemical aversion therapy found support for chemical aversion that uses nausea as the unconditioned stimulus (emetine and lithium). These had a significant effect size but not paralysis-inducing drugs (succinylcho-line chloride or suxamethonium). He found that aversion accounted for 9% of the variance in outcomes. Caddy and Block (42) review the use of vertigo to produce a nausea aversion, but no outcome study using this method in clinical practice is available.

Garcia and Koelling (43) compared stimulus conditioning in rats, comparing lithium and radiation that produced unconditioned nausea in rats to electric shock to the paw. Rats conditioned with lithium and radiation developed conditioned nausea to a gustatory stimulus but not to an auditory stimulus, whereas rats conditioned with shock to the paw developed conditioned avoidance to an auditory/visual stimulus but not to a gustatory stimulus. These preclinical findings suggest that humans and other organisms may be biologically predisposed to form long-lasting conditioned aversions to consumables such as alcohol and foodstuffs whose consumption is followed by nausea and vomiting.

Nausea Aversion

Smith (3) has report of nausea as an aversion technique. Before a typical treatment session, the patient is kept NPO except for clear liquids for 6 hours, which reduces the risk of solid stomach contents being aspirated. After informed consent about the procedure is obtained, the patient is escorted to a small treatment room that has bookcases filled with bottles of various alcoholic beverages and wall hangings of alcohol-related advertising. This provides sufficient visual cues related to drinking alcohol. After the patient sits in a comfortable chair with an attached large emesis basin, he or she receives a pilocarpine and ephedrine injection to elicit arousal of the autonomic nervous system. The patient is given an oral dose of emetine. Prior to the expected onset of emesis (5 to 8 minutes), the patient drinks two 10-oz glasses of lightly salted warm water (to counteract electrolyte loss), which creates a bolus of easily vomited fluid. The clinician doing the procedure pours a drink of the patient’s preferred alcoholic beverage, diluting it 1:1 with warm water, and presents it to the patient just prior to the expected onset of nausea. The patient smells the drink, takes a small mouthful to swish around, and then spits it out into the basin. The “sniff, swish, and spit” procedure provides a differentiated visual, olfactory, and gustatory experience that will be linked to the patient’s preferred beverage before the aversive stimulus of nausea begins. After nausea and vomiting begin, the patient can then “sniff, swish, and swallow” the preferred beverage mixture, but it stays down for only a short time, with little absorption, before it is vomited up. This first phase of intensive conditioning, which lasts 20 to 30 minutes, is followed by a second, longer phase back in the hospital room, where the preferred alcohol-containing beverage, which contains an oral dose of emetine and tartar emetic, is drunk after another 30 minutes. This precipitates residual nausea for up to 3 hours duration. Five treatments are given on an every other day basis to a typical patient over 10 days.

Howard (44) has provided a detailed review of the production of nausea and vomiting as treatment progresses. He showed that both nurses and patients reported on a 0 to 3 scale progression from 2.22 and 2.30, respectively, after the first treatment to 2.99 and 3.00, respectively, at the end of the fifth treatment. By the third treatment, more than 95% of patients and nurses reported the production of nausea and vomiting.

Faradic Aversion Therapy for Alcohol

For a faradic session, Smith (3) suggested that a pair of electrodes is placed 2 inches apart on the forearm of the patient’s dominant hand. An electrostimulus machine provides 1 to 20 mA of direct current through the electrodes, and the therapist begins test stimuli in increasing amplitude to determine the stimulus that the patient will experience as aversive. This must be done on each session as there is wide variation by day within and between patients that requires individual thresholds set for each patient on each day.

The intervention pairs the visual, olfactory, and gustatory stimuli of alcoholic beverages with the predetermined amplitude of electrostimulation that the patient will find aversive. The clinician directs the patient to pour some of a bottled alcoholic beverage into a glass and to taste it but not to swallow it. During this forced-choice phase, from reaching for the bottle to tasting the drink, the onset of the aversive stimulus occurs at random intervals, with varying numbers (from 1 up to 8) of aversive stimuli among trails. A free-choice phase of 10 supplemental trials provides negative reinforcement, such that if the patient selects a nonalcoholic choice such as fruit juice, the aversive stimulus is removed. The clinician monitors the patient’s compliance with the directive that he or she must not swallow any alcohol at any time during the treatment session.

Sessions last 20 to 45 minutes, depending on the individual patient. After the aversion conditioning session in the treatment room, the patient returns to his or her room, listens to a relaxation tape, makes a list of positive changes with sobriety, and contrasts them with the negative consequences of continued use.

Covert Sensitization

Elkins (45,46) has published on the use of covert sensitization to demonstrate that conditioned nausea responses can be trained in alcoholic patients through the use of imagination and verbal suggestion without the use of an emetic drug. In covert sensitization, patients are helped to imagine personally relevant drinking scenes that emphasize the motivational, sensory, and behavioral precursors and concomitants of alcohol ingestion. The drinking scenes then are paired repeatedly with verbally induced nausea. Most cooperative participants can learn to experience genuine and intense nausea reactions by focusing on the therapist’s noxious verbal suggestions; these suggestions prompt recipients to remember and recreate prior feelings and thoughts that have been prominent in their former nausea experiences. Such verbally induced nausea is designated as demand nausea. Repeated presentations of the drinking scenes (i.e., conditioned stimulus or CS) followed by episodes of verbally induced demand nausea (i.e., unconditioned stimulus or US) can, over extended conditioning trials, produce conditioned aversions to alcohol in a majority of participants. Elkins (45) described behavioral and psy-chophysical indices that can be used to define an individual subject’s transition from demand nausea to conditioned nausea, the goal of treatment. Conditioned nausea is nausea as an automatic consequence of the patient’s focusing on a drinking scene without any attempted therapist or self-induction of nausea.

Fifty-two patients were entered. Thirty-three were able to develop verbally induced nausea after imagined drinking scenes. It took an average of four CS–US pairings over an average of 1.8 sessions to develop this demand nausea. Of these patients, 23 were able to develop conditioned nausea to either the desire for alcohol or other alcohol-related physical stimuli. It took an average of 13.43 scenes over 4.83 sessions for them to develop conditioned aversion. Of note, the 10 patients who did not develop conditioned nausea had the same number of training sessions compared to those who did (11.79 vs. 11.64) but had more scenes (48.48 vs. 37.82).

Those who developed conditioned nausea had an average of 13.74 months of total abstinence as compared to 4.52 months for those who failed to progress beyond the demand nausea stage or treatment (p < 0.05). Elkins suggests that some patients are resistant to developing conditioned nausea to alcohol, despite the ability to develop demand nausea to verbal prompting.

Aversion Therapy for Marijuana Dependence

The spontaneous recovery rate from marijuana dependence is not known and, like that for alcohol and nicotine, probably depends on multiple factors. Chemical aversion using emetine has been used for marijuana dependence. In clinical practice, aversion therapy for marijuana uses faradic aversion. The protocol for faradic aversion is similar to that of the treatment for alcohol, except that it uses a variety of bongs, drug paraphernalia, and visual imagery. An artificial marijuana substitute and marijuana aroma are used in treatment. A 1-year abstinence rate of 84% was reported after 5 days of treatment, combined with three weekly group sessions on self-management techniques (47).

Aversion Therapy for Cocaine-Amphetamine Dependence

The spontaneous recovery rate from cocaine or amphetamine dependence is not known. Rawson et al. (48) followed 30 patients who had requested information about stopping cocaine but had not used treatment for an 8-month period; 47% were reported at the follow-up point to be using cocaine at least monthly. No total abstinence figures are available. Frawley and Smith (49) reported the use of chemical aversion for the treatment of cocaine dependence. In this treatment, an artificial cocaine substitute called articaine was developed from tetracaine, mannitol, and quinine. Patients snorted this substance and paired it with nausea induced by emetine. Of those so treated, 56% were continuously abstinent and 78% currently abstinent (i.e., for the prior 30 days) at 6 months after treatment; at 18 months, 38% were continuously abstinent and 75% currently abstinent. For those treated for both alcohol and cocaine, 70% were continuously and currently abstinent from cocaine at 6 months, 50% were continuously abstinent, and 80% were currently abstinent at 18 months after treatment.

Frawley and Smith (50) reported a 53% 1-year continuous abstinence rate in 156 cocaine- and/or amphetamine-dependent patients treated with aversion. This was based on a 73% follow-up rate. Outcomes for chemical and faradic aversion were not significantly different. The report also compared patients with both alcohol and cocaine dependencies who were treated with aversion to alcohol only with a later group when aversion was available for cocaine also. The addition of the aversion for cocaine produced statistically significantly improved abstinence rates in this population. The increase in cocaine abstinence in the later group compared to the first (55% vs. 88%) is greater than the decrease on follow-up from the first to the second group (84% vs. 64%). Because of the lower follow-up rate in the second study, this research needs replication.

Elkins et al. (51) reported a well-designed experimental evaluation of three aversion therapy treatments for cocaine dependence. Volunteer participants from the Augusta VA Medical Center (VAMC) Substance Abuse Treatment Program (SATP) were assigned to one of three aversion therapy treatments or to one of two control groups. All accepted participants satisfied the stringent medical criteria that must be met for participation in emetic therapy, the most physically demanding treatment. The additional two experimental treatments were faradic therapy and covert sensitization therapy. Participants were randomly assigned to one of the three aversion therapy groups or to one of the two control conditions—milieu treatment or relaxation therapy. Milieu therapy was the baseline control treatment that was received by all SATP participants. Emetic and faradic therapy participants used realistic placebo cocaine products during their assigned aversion interventions. The covert sensitization subjects imagined cocaine usage in conjunction with verbal nausea induction. The placebo cocaine products included simulated “rocks” of crack cocaine, simulated snortable cocaine, and Psychem, an essence that contains oils that mimic the odor of street cocaine. Each “rock” of smokeable placebo cocaine was prepared from benzocaine, baking soda, and water. Snortable placebo cocaine consisted of 97% mannitol, 1% quinine, and 2% tet-racaine in powder form. The quinine was added to simulate the bitter aspect of cocaine, and the tetracaine was added to mimic the “nose-deadening” anesthetic properties of snorted cocaine (49). All efforts were made to provide paraphernalia for snorting and smoking the placebo cocaine that resembled each patient’s customary paraphernalia. All aversion therapy sessions were completed without any serious side effects. Posttreatment abstinence performance was rigorously tracked during a 6-month follow-up period through telephone queries of the participants and their designated collateral contacts and via in-person contacts that included urine drug screens. Patients lost during follow-up were classified as relapsed.

A major positive outcome of this study is found in the emetic therapy participants’ significantly elevated post-treatment abstinence performance. The 57.9% 6-month follow-up abstinence finding for the emetic therapy recipients significantly exceeded the 26.5% 6-month abstinence finding for the milieu control participants. These emetic therapy data are quite encouraging and are consistent with the reported 56% 6-month abstinence rate from the previously discussed clinical trial of emetic therapy for cocaine dependence (49). Covert sensitization produced a significant therapeutic benefit, but its effect did not extend beyond 3 months after treatment. The second major positive outcome is found in the emetic therapy participants’ uniquely reported total loss of cravings for cocaine by the end of treatment (discussed in more detail in section “Effect of Aversion on Cocaine Craving,” below).

Aversion Therapy for Heroin Addiction

Copemann (52) employed a unique approach to aversion therapy by pairing aversive stimuli to cognitive images of heroin use. Patients were asked to verbalize only after they had conjured up a strong mental image. A second part of the treatment asked addicts to conjure images of socially appropriate behavior, involving employment, education, or nondrug entertainment. Latency to verbalization was measured. Copemann found that, at baseline, addicts could rapidly conjure up positive thoughts about heroin use but had significant delays in conjuring up thoughts about rewarding nondrug activities. Subjects were in a halfway house for heroin addicts and received group therapy in conjunction with relaxation therapy in addition to the aversion treatment. A faradic stimulator was used. Once addicts had conjured up drug images, faradic aversion was applied. At other times, addicts were given 15 seconds to conjure images of nondrug socially appropriate behavior to prevent aversion from being applied. With this training over an average of 15 sessions (range, 5 to 25), latency for drug-related images increased, while that for socially appropriate images decreased. Thirty of 50 patients completed the treatment, and, at 24 months, 80% (24 of 30) were reported to be drug-free.

Since 2012 a new approach to aversion therapy for heroin has been developed at the Schick Shadel Hospital in Seattle, Washington, under the supervision of Ralph Elkins, PhD. This heroin treatment combines hypnotically guided covert sensitization with the program’s well-established medical (emetic) counterconditioning. Seven heroin countercondi-tioning sessions are provided within the traditional 10-day inpatient multimodal treatment format that is used for alcohol and other drug dependencies. The psychologist provides hypnotic treatments as follows. The first two treatment sessions are devoted to establishing the patient’s individualized heroin use rituals, to introducing the patient to hypnotic induction, and to teaching the hypnotized patient to reexpe-rience via imagination his/her habitual heroin procurement and use rituals. The treated patients typically have reported highly realistic imagined heroin use. Elkins concludes each imagined heroin use ritual with nausea induction suggestions. The objective of the nausea suggestions is to induce genuine nausea that stops short of actual emesis. The hypnotically induced nausea starts the formation of the coun-terconditioned heroin aversion before the patient begins his/her subsequent five counterconditioning treatments that feature emetic nausea induction. The hypnotic nausea suggestions of the initial two sessions also enhance the patient’s subsequent ability to focus on his/her imagined heroin use rituals while experiencing emetic nausea induction. The patient’s ability to focus on the imagined use rituals despite emetic nausea is important because there is, as yet, no placebo heroin product that can be smoked or injected.

The patient first views typical usage paraphernalia and his/her placebo heroin visual aids upon entering the treatment room for the first of five emetic counterconditioning sessions. The patient then uses a bipolar Likert type visual desire/aversion scale (DAS) to rate his/her heroin desire or heroin aversion intensity; the ratings range from intense desire (−5) to intense aversion (+5). The ratings are repeated at the beginning and end of each treatment session to record the patient’s acquisition of an anticraving conditioned heroin aversion. The Treatment Nurse records every DAS score report as part of the patient’s permanent medical record. Following the initial DAS rating, the Treatment Nurse administers physician-ordered oral ipecac, an emetic drug that typically induces nausea and emesis within 10 to 15 minutes. The psychologist then induces hypnosis and guides the patient through a typical heroin use ritual as nausea builds to produce emesis. The emesis episode signals the end of the first imagined use ritual. According to Dr. Elkins, nausea and not emesis is the known causal basis of aversion formation. The act of emesis relieves the nausea and provides the opportunity to begin another imagined use ritual. A second imagined heroin use ritual is initiated following a brief respite during which the patient drinks water to replace lost fluid. A treatment typically includes from four to six such rituals. The patient then reports the end-of-session DAS score and returns to his/her room with the placebo heroin and usage paraphernalia for a period of extended nausea with heroin use contemplation.

Upon discharge, patients are encouraged to return for up to three overnight recap (booster) treatments during the next 90 days at no additional charge. Each recap treatment includes one emetic counterconditioning treatment, and the patient also participates in all regularly scheduled hospital activities. The patients also are encouraged to take naltrex-one, an opiate receptor blocking drug for at least 6 months following treatment.

Elkins reported that 36 patients have completed treatments since the beginning of July of 2012 (53). These patients typically entered treatment with strong DAS-confirmed heroin cravings and left with strong aversions. Patient satisfaction has been very high as confirmed via exit ratings. Additionally, more than 60% of patients who have been out of treatment for 30 days or more have returned for at least one recap treatment. Abstinence from use of the treated substance(s) is a recap requirement, and patients must pass urine drug tests before any recap treatments are administered. The pre-recap treatment DAS scores of these returning patients have confirmed their typical maintenance of strong heroin aversions. The recap patients typically have reported a marked absence of any problematical heroin cravings. Elkins is submitting case reports of several recap patients for publication. These case reports include treatments for smoked and injected heroin as well as treatments for additional problem substances including smoked opiate pills.

Use of Reinforcement (Booster) Aversion Treatments

Smith and Frawley (54) followed up at 1 year on 437 patients of 600 patients treated with chemical and faradic aversion for alcohol, marijuana, or cocaine. One-year complete abstinence rates for alcohol for those who did not return for any reinforcements (n = 51) was 29.4%; for one booster aversion treatment (n = 93), the abstinence rate was 50.5%; the two booster aversions’ (n = 273) abstinence rate was 68.5%; and for more than two aversions (n = 10), the abstinence rate was 80%. Wiens and Menustik (40) reported that the use of reinforcement aversion treatments (or recaps) was associated with improvement in abstinence at 1 year after treatment: no recaps, 24%; 1 recap, 21%; 2 recaps, 40%; 3 recaps, 27%; 4 recaps, 64%; 5 recaps, 72%; and 6 recaps, 99%. Of note, 144 of 385 (38%) patients received the 6 recaps.

Use of Support Programs and 12-Step Meetings after Receiving Aversion Therapy

Beaubrum (55), in a follow-up study of patients, some of whom were offered group emetine aversion and AA followup, found that those who went to AA after receiving aversion had a 71% 2-year abstinence as compared to 27% who did AA alone, 18% who did aversions alone, and 23% who did neither. However, these were not random assignments. Smith and Frawley (54) found that those who used some form of support groups after aversion treatment did better than those who did not use such support. The use of any support (n = 232) versus no support (n = 195) is associated with a total abstinence rate of 68.1% versus 48.7% (p < 0.001), respectively. Smith and Frawley (54) found an additive effect of the use of reinforcement treatments and support and/or 12-step meetings after completion of a hospital aversion program. Those who did not complete at least two booster aversions and did not go to any support after treatment (n = 79) had a 1-year complete abstinence rate from all drugs of 34.2%. For those who went to support groups but did not receive at least two booster aversions (n = 65), the rate was 52.3%. For those who had at least two booster aversions but did not go to support (n = 116), the abstinence rate was 58.6%. For those who had at least two booster treatments and attended support groups (n = 167), the abstinence rate was 74.3% (Table 62-2). Though total abstinence was associated with use of support groups after treatment, for those with urges to drink, increased support use was negatively associated with abstinence. For those using AA more than once a week with urges (n = 25), the abstinence rate was 20%. For those going once a week (n = 17), the abstinence rate was 35%. For those who were going less than once a week (n = 29), the abstinence rate was 66% (p < 0.01). A similar pattern was found for patients going to Schick Shadel Hospital–sponsored support groups. For those going more than once a week (n = 9), the abstinence rate was 38%. For those going once a week (n = 30), the abstinence rate was 50%. For those going less than once a week (n = 24), the abstinence rate was 63% (p = n.s.; Table 62-3).



Reprinted from Smith JW, Frawley PJ. Treatment outcome of 600 chemically dependent patients treated in a multimodal inpatient program including aversion therapy and pentothal interviews. J Subst Abuse Treat 1993;10:359–369, with permission.



Reprinted from Smith JW, Frawley PJ. Treatment outcome of 600 chemically dependent patients treated in a multimodal inpatient program including aversion therapy and pentothal interviews. J Subst Abuse Treat 1993;10: 359–369, with permission.


Baker and Cannon (56) studied individual patients and showed evidence for conditioned autonomic responses such as heart rate and skin conductance that increased during the course of aversion treatment and were related sequentially to the presentation of alcoholic beverages. They also demonstrated reduction of autonomic responses in taste tests to alcoholic beverages but not nonalcoholic beverages. They did note that attentional factors played some role in the autonomic responses. A detailed review of chemical aversion therapy, its rationale, and comparison with other treatments for alcoholism have been presented by Howard et al. (57). They emphasize that it has been shown that patients are able to develop an aversion to alcohol as demonstrated by autonomic markers and patient report and that there is an association between the strength of the aversion and abstinence from alcohol. Elkins (2) notes that a variety of researchers have reported that some patients do not seem to develop aversions. He indicates that animals can be bred to have different susceptibilities to the development of taste aversions (TA). In rats, TA-prone animals have lower plasma tryptophan levels than those who are TA resistant. This line of research may ultimately help identify patients who are more or less able to develop aversions to alcohol.

Howard (44) provided a detailed evaluation of the process of aversion development during treatment. He recruited 94 of 181 patients who were receiving aversion therapy for the first time. Eighty-two of these patients agreed to participate in the study. He found that there were significant differences between pre- and posttreatment responses to patients’ desire to drink alcohol (p < 0.01), confidence that you will not drink in the next 2 months (p < 0.01), confidence that you will be drinking less alcohol at 6 months (p < 0.01), difficulty with the ability to cut down drinking (p < 0.01), and strength of desire to quit drinking permanently (p < 0.01). An alcohol expectancy questionnaire revealed significantly reduced positive expectancies from drinking (p < 0.001). Confidence in not drinking in high-risk situations was also significantly improved (p < 0.001). Videotaped alcohol olfaction tests that included heart rate response compared the patient’s response to the smell of their favorite alcoholic beverage, a nonpreferred beverage, and a nonalcoholic beverage. After treatment, heart rates significantly increased to the alcoholic beverages and were reduced to nonalcoholic beverages (p < 0.05). The evaluation of the videotaped aversion response was blinded to condition. The difference between pre- and posttreatment responses was highly significant (p < 0.001). An adjective-rating scoring system also showed significant changes between pre- and posttreatment (p < 0.001). To document the production of nausea, both patients and nurses provided ratings. By the third treatment session, more than 90% of patients and nurses reported vomiting.

Howard (44) did regression evaluations of up to 17 variables to attempt to identify pretreatment factors that might predict differences in developing an aversion response. He used posttreatment heart rate responses as one measure of conditioned aversion. As the number of lifetime arrests (15% of variance), instances of drug use, negative life events, and family income increased (1% of variance), posttreatment heart rate response to the preferred alcoholic beverage decreased when compared to that of a nonalcoholic beverage (p < 0.001). Pre- to posttreatment change in heart rate response to the preferred alcoholic beverage was less in patients who rated themselves as more alcohol dependent, had driven after having more than three drinks, had more positive life events in the 6 months prior to treatment, and had lower family income. Of interest, those who had more lifetime nausea and vomiting with drinking had less of a change in heart rate over the treatment session. These associations are of interest, in particular, since chemical aversion induces nausea; however, they will need to be replicated.

Howard (44) used two blinded raters in evaluating videotaped aversion responses. One rater found a correlation between greater pretreatment perceived difficulty with cutting down drinking and a stronger videotaped conditioned aversion response (p < 0.06). A second rater found a correlation between both, a patient’s pretreatment feeling of nausea due to drinking and difficulty with cutting down and a greater conditioned aversion after treatment (p < 0.01).

A greater change from pre- to postaversion response was correlated with pretreatment nausea with thinking about drinking, less confidence in being able to drink less in the next 2 months, high pretreatment alcohol-related positive expectancies, and the lowest level of physical dependence. Posttreatment adjective ratings were also correlated with pretreatment variables. The greater the client’s pretreatment level of confidence that he or she would be drinking less alcohol in 6 months, the more negatively the patient rated alcohol after treatment. Patients who more frequently drove after drinking, who had higher levels of pretreatment self-reported efficacy in high-risk situations, who had fathers who were heavier drinkers, and who did not accept the need for total abstinence had more positive adjectives for alcohol after treatment (p < 0.0001). The change in adjective rating of the preferred alcoholic beverage was correlated with being a man, having higher pretreatment positive alcohol-related expectancies, and clients with spouses who were lighter drinkers (p < 0.0001). If replicated, these findings along with those earlier may help to stratify patients in terms of difficulty with developing aversion and/or evaluating their aversion in a different way.

Relapse to alcohol and drug use is the result of a variety of factors. Patients who report that aversion therapy greatly reduced their urges have the best outcome (49). Smith and Frawley (54) found that at follow-up, patients who reported the loss of all urges to drink after aversion treatment (n = 165) had a total abstinence rate at 1 year from alcohol of 89.7%. Those who reported only the loss of uncontrollable urges (n = 183) had an abstinence rate of 56.8%. Those who reported that they still had urges had an abstinence rate of 6.3%.

Smith and Frawley found no association between the use of support groups during the period of being at risk for relapse and the status of urges to drink. There was an insignificant trend toward more urges with greater support group use. However, there was an association between use of reinforcements and urge to drink. Seventy-six and one-half percent of those who used no reinforcements (n = 51) reported urges to drink, while 54.8% of those who had one reinforcement (n = 93) and 49.8% of those who had at least two reinforcements (n = 283) reported urges (p < 0.01).


Ninety-seven participants within the relaxation training, covert sensitization, faradic treatment, and emetic therapy conditions of the Elkins et al. (51) study provided pre- and postsession subjective ratings of cocaine cravings. The ratings were on a scale of 0 to 10, with 0 representing a complete absence of craving and 10 representing the most extreme craving the participant had ever experienced. Each participant’s four craving scores from treatment sessions 1 and 2 were dichotomously classified as either reflecting positive cravings (e.g., at least one score in excess of 0) or no craving (all scores were 0). The same tabulation was applied to the craving scores from sessions 7 and 8. The emetic therapy and faradic therapy participants all reported positive cue-induced cravings for cocaine upon their initial exposure to the placebo cocaine materials. The high rates of craving of the faradic and emetic participants during sessions 1 and 2 exceeded those of the relaxation and covert sensitization subjects; these differences probably occurred because the faradic and emetic participants were exposed to the placebo cocaine products whereas relaxation and covert sensitization participants were not so exposed. The emetic therapy group was the only treatment condition in which every participant’s reported cravings were reduced to 0 during the final two treatment sessions. All emetic and faradic therapy participants had reported positive cravings for cocaine during the first two treatment sessions. A comparable pattern of maximum craving loss by emetic therapy participants also was reported in an analysis of a subset of the present data (58). That preliminary data analysis was included in a University of Georgia doctoral dissertation by Patrick S. Bordnick, who had participated in the research study while completing a predoctoral internship in Elkins’s VAMC laboratory.


In addition to craving, however, relapse may be related to reflex-conditioned responses to drink or use in response to either external cues such as being around others who drink or use drugs or going to parties, or may be in response to internal cues such as negative or positive emotional states (50,59). The development of new, more appropriate responses to these emotional states and a change in the recovering person’s associations remain important goals of treatment. Aversion therapy does not interfere with their development but instead enhances the readiness of the patient to avoid the alcohol or drug and thus prepares him or her for new approaches and patterns that do not involve chemicals. Frawley and Smith (50) found that the use of support groups was associated with improved abstinence in patients treated with aversion for cocaine and methamphetamine dependence. Smith and Frawley (54) report that 49.5% of patients who relapsed after aversion treatment did so for one of two reasons: (i) “intrapersonal determinants of relapse” associated with stress from work (9.4%) or marriage/family relationships (17.6%) or (ii) “interpersonal determinants of relapse” associated with being around others who were drinking (10.6%) or at a celebration or special event (11.8%). Relapse to the use of cocaine or marijuana showed this trend more strongly. Nearly three of four (72.9%) who relapsed did so in an association with these two factors. Interpersonal factors were the most powerful in inducing relapse. Being around others who were using or pressuring the patient to use was reported as the principal cause for relapse by 34.9% of cases and participation in a celebration or special event in 19% of cases. Intrapersonal factors were not as important in cocaine/marijuana relapse as in alcohol relapse. Marital/family relationship stress accounted for only 11.1% of relapses, and work-related stress accounted for 7.9% of relapses.


Faradic aversion has virtually no unwanted side effects and has been found to be safe for patients with pacemakers and pregnant women (because the current only travels between two electrodes on the arm). To be eligible for chemical aversion therapy, patients must be free of medical contraindications such as esophageal varices, serious coronary artery disease, or active gastrointestinal (GI) pathology. Emetine is given only orally, effectively eliminating the risk of cardiotoxic effects of intramuscular emetine, since very little of the orally administered emetine is absorbed (60,61). Emetine exerts its principal action by irritating the GI mucosa, hence stimulating the afferent vagus nerve to stimulate the vomiting center in the medulla. Oral emetine is effective in stimulating nausea in more than 95% of cases, whereas intramuscular emetine produces nausea only 30% to 40% of the time, probably through excretion in the bile (61,62). Pilocarpine and ephedrine are not used in patients with asthma or serious hypertension, and tartar emetic is held for patients with excessive diarrhea or prolonged nausea. Smith et al. (14,63) found that there was no increased incidence of medical utilization or hospitalization in the 6 months after treatment in a group treated with aversion therapy, as compared to matched controls treated without aversion.

There are some contraindications to covert sensitization, similar to those for chemical aversion; however, with this therapy, emesis can be prevented in most cases. The drawback to covert aversion therapy is that the induction of nausea or other aversive state is not as predictable as with medication and requires more patient preparation (46).

Satiation therapy with nicotine has been studied by Hall et al. (22) in 18 patients (9 men and 9 women; average age, 45.8 years) with pulmonary and cardiac disease. Nine had definite or probable cardiac ischemia, but none showed ischemic changes during rapid smoking. Premature ventricular contractions were not increased above baseline by rapid smoking. Wiens et al. (64) have reviewed the outcome of treatment with chemical aversion therapy in patients who were 65 years and older. Seventy-eight of 87 patients (89%) who were admitted were able to complete an initial course of chemical aversion therapy. The report does not indicate that patients had any adverse effects from treatment, though six patients did not complete treatment owing to medical concerns such as pneumonia, chest pain, shortness of breath, other pulmonary problems, and ane-mia. Two patients had organic brain syndrome or Korsakoff syndrome, and one patient chose to detox only. Sixty-five percent of the patients who completed the initial treatment were able to stay sober for at least 12 months.


Selecting the appropriate treatment for a particular patient involves the patient’s having full informed consent. The practitioner needs to counsel the patient about the risks of continuing the addiction and the risks, benefits, and expected outcomes of various methods of treatment. Studies of patients who voluntarily received aversion therapy do not show higher rates of leaving against medical advice than is found in patients in Minnesota Model programs (6567). Wiens and Menustik (40) reported that of 835 patients admitted in 1978 and 1979 to the Raleigh Hills Program in Oregon, only 52 (6%) did not complete treatment. Hence, for patients seeking aversion therapy in clinical populations, they complete treatment at the same rate as patients seeking alternative established treatments.

Chemical aversion therapy has been reviewed and approved as a treatment for alcoholism by the National Center for Health Services Research and Health Care Technology Assessment (68), the body charged with determining which services should be reimbursable under Medicare. A 1987 Office of the Civilian Health and Medical Program of the Uniformed Services (OCHAMPUS) demonstration project recommended coverage for chemical aversion therapy for alcoholism (69). A California Medical Association Scientific Advisory Panel approved aversion therapy as appropriate treatment as part of a multimodality program (70). The American Society of Addiction Medicine includes aversion therapy as an appropriate part of treatment for alcoholism and other drug dependencies (71). In a 1987 review, the American Medical Association’s Council on Scientific Affairs (72) found that the available research for the efficacy of faradic aversion for alcohol was weak but that there was moderate support for the efficacy of chemical aversion. The Council found some support for both faradic and rapid smoking aversion techniques. However, the report emphasized that there have been few controlled trials. The Office of Technology Assessment of the U.S. Congress found that aversion therapy was effective for some patients under certain conditions (73). As a result of these reviews, aversion therapy has been recognized by both governmental and private agencies as appropriate treatment for patients with addictive disease.


Much of the criticism of aversion therapy is similar to criticisms of other forms of addiction treatment and stems from a lack of multiple well-designed controlled trials and the need to identify which treatment is appropriate for a given patient (73,74). Elkins (57) and Wilson (75) have taken up a debate in the literature about aversion therapy. Wilson summarized a variety of criticisms, including (i) greater medical expense than some other forms of treatment, (ii) intrusiveness of aversion, and (iii) the theoretical framework of aversive conditioning. Howard and Jenson (76) responded to these with a detailed review of the literature and concluded that chemical aversion therapy effectively produces a conditioned aversion and this contributes to the positive treatment outcome. Elkins (2) has also responded to these criticisms and initiated a debate in the literature with Wilson. With regard to intrusiveness, Elkins (2) quotes from Bandura that “The brief discomfort occasioned by a program of aversion therapy is minor compared to the repeated incarceration, social ostracism, serious disruption of family life, and self-condemnation resulting from uncontrollable injurious behavior.” He also thoroughly reviews the physiologic and psychologic bases of aversion therapy both from animal and human work. Wilson (77) subsequently responds by agreeing with Elkins on some of his points, but he then outlines the following areas of disagreement: (i) The methodologic problems associated with outcome studies of aversion therapy leave its results in question, (ii) he questions whether aversion itself is the cause of improved outcomes as opposed to other unmeasured patient or treatment characteristics, and (iii) he advocates that the least intrusive and most cost-effective methods should be used first. Elkins (78) responds that because there are many factors that lead to relapse, emetic therapy should continue to be available as part of multimodal treatment program for alcohol addiction treatment especially given the fact of high rates of relapse with the forms of treatment currently available. Elkins also challenges the concept of aversion’s being less cost-effective in that the costs of aversion therapy, which is provided in an inpatient setting, are within the midrange of costs of other commercially available programs. Moreover, he reports that though there are no commercially available outpatient aversion programs in the United States, there are reports of outpatient aversion therapy in research studies and from other countries.

Though aversion therapy is more expensive than some other forms of treatment, the OCHAMPUS demonstration project did compare costs of OCHAMPUS beneficiaries admitted to the hospital with other forms of treatment and found the average cost to be slightly less for aversion therapy (which may be partially attributable to the shorter length of stay for aversion). The principle that, when treatment outcomes are the same, the least intrusive treatment should be tried first has been raised by Wilson (75) and echoed by Carter (68). As further work is done on patient–treatment matching, patients for whom aversion clearly is more effective should be offered this treatment before less effective but less intrusive treatment. Smith et al. (14) found significantly better outcomes with aversion therapy in males and daily drinkers, and Boland et al. (34) found significantly better outcomes with aversion therapy in indigent male alcoholics. This work should be replicated. The theoretical framework of aversion conditioning as a treatment does not exclude the need to develop alternative modes of behavior as part of a multimodality treatment program. DeSilva and Rachman (7) demonstrated that aversions are common and can persist for a long time. Cannon et al. (6) demonstrated a relationship between the strength of an aversion and the duration of abstinence after treatment.


We were pleased to note that the work by Elkins et al. (51) at the VAMC in Augusta, Georgia, was funded by the National Institute on Drug Abuse (NIDA). Though private hospitals can fund some research, they are not set up to do randomized trials. Drug companies will fund medication trials to help reduce craving and relapse but with the hope to get a patentable compound. Medical treatments are not generally patentable procedures; hence, the private funding for this research is limited. The National Institutes on Alcohol Abuse and Alcoholism conducted a trial to look at three different counseling interventions, which were found to be equivalent in outcome (79). We would like to see more research focused on aversion and its ability to reduce craving and hence relapse.

The emetic therapy recipients of the Elkins et al. (51) cocaine treatment experimental evaluation did more than simply lose their cravings for cocaine; they also developed strong active revulsions for the placebo cocaine materials as well as for cocaine-related cues as detailed in that report. However, the revulsions were not measurable by the 0 to 10 craving scale that was used in the study. Future studies should incorporate bidirectional scales that measure maximum craving at one extreme and maximum revulsion at the other extreme with a neutral zero-craving midscale region.

Future research should better characterize the physiologic changes that coincide with the transition from cue-induced cravings to cue-induced revulsion within a course of emetic aversion treatments. A traditional approach could build on the studies of psychophysiologic measurements such as those used during covert sensitization alcoholism treatment (45) and during emetic therapy alcoholism treatments (6,56). However, the dynamically expanding field of brain imaging research is likely to provide the greatest near-term advancements in our basic understanding and possible clinical applications of cue-induced brain changes that occur during the emetic therapy–induced transition from cocaine cravings to revulsions. Recent studies have reported activations of specific brain regions during cue-induced cravings for cocaine. Elkins et al. (51) have hypothesized that reliable change in brain activation patterns will be revealed by comparisons of the initial episodes of cue-induced cravings that typify the beginning of treatment with those that accompany the late-treatment cue-induced revulsions of successfully conditioned participants. The landmark positron emission tomography (PET) scan findings of Volkow et al. (80), obtained from cocaine-addicted human volunteers, have shown that dopamine in the dorsal striatum is involved in cocaine craving and addiction. The dorsal striatum is a region that has been implicated in habit learning and in action initiation; the dorsal striatum, therefore, is a high-interest area for studies of possible transitions from cravings to revul-sions. Based on the PET scan findings of Kilts et al. (81), cue-induced craving to revulsion changes also is likely to be found in brain regions that include the amygdala, the right nucleus accumbens, the dorsal anterior cingulate cortex, the ventral anterior cingulate cortex, and the frontal cortex. An obvious clinical application of such information would be to assess the strength of the attained aversion at the end of treatment. Additionally, the findings could support propitious individually tailored timings of booster treatments.

Functional magnetic resonance imaging (fMRI) also may be well suited to studies of emetic therapy–induced changes of cue-induced cravings to revulsions. The fMRI technology, unlike PET scan technology, does not involve the injection of radioactive compounds; it therefore can be safely used during repeated measures of the same participants across different time periods.

Elkins (2) notes that a variety of researchers have reported that some patients do not seem to develop aversions. Those reports plus his previously described individual differences in conditionability of alcoholic recipients of covert sensitization treatment prompted Elkins and his collaborators to develop and to study lines of selectively bred taste aversion–prone (TAP) and taste aversion–resistant (TAR) rats (8285). Selective breeding has produced a line of TAP rats that have robust TA conditioning propensities and a line of TAR rats that are extremely resistant to TA learning. The TAP and TAR lines have not been found to differ when tested within any other traditional non-TA learning paradigm. The differences between the two lines, therefore, may be largely confined to line-specific separations of the pool of genes that subserve efficient TA conditionability.

Studies of the TAP–TAR models give promise of leading to identifications of biologic indices to separate conditionable and nonconditionable potential emetic therapy recipients. Additionally, studies of the two lines may support the development of pharmacologic or nutritional interventions to increase the nausea-based conditionability or TAR substance abusers. Two preliminary studies in this area of research include reports of marked cocaine-induced differences in gene expression patterns in the amygdalae of TAP and TAR rats. The subjects received a cocaine injection, and brain tissue was harvested 6 hours later for analysis via an oligonucleotide probe arrays (Affymetrix Corporation Santa Clara, CA). The findings included between-lines differences in cocaine-induced gene expressions of glutamate receptor subunits and transporters (86), in Ca2+/calmodulin-dependent protein kinase subunits (87), and in 5-HT3 receptors and Na+/K + −ATPase pump subunits (88). Summaries and references for all published studies reporting the development and characterizations of the TAP and TAR rat lines will be found posted on the taste aversion learning Web site of Dr. Anthony Riley of American University (89,90).

A new treatment for oxycontin dependence has been developed in the Schick Shadel Hospital by Merchant (personal communication, 2008). This treatment capitalizes on the use of naltrexone to negate the psychotropic effects of oxycontin. Oxycontin-dependent recipients first are detoxified; they then are started on a daily naltrexone regimen that begins in the morning of the first treatment day. The recipients then use oxycontin in their customary manner during emetic therapy sessions. The treatment has been well received and is being requested by an increasing number of patients. If the clinical follow-up results are consistent with likely therapeutic efficacy, then this innovative treatment clearly will merit a NIDA-type experimental evaluation similar to that previously described for the emetic therapy treatment of cocaine dependence.


The authors acknowledge the assistance of James W. Smith, MD, FASAM, who has reviewed the manuscript, and thank him for his insights.

Dr. Smith utilized aversion therapy in the treatment of addiction over his 42-year career in the field, improved the safety and efficacy of the treatment, and conducted much of the research on aversion therapy. In gratitude for his dedication to the field, the authors dedicate this chapter to his memory. They also thank Dr. Ralph Elkins for his assistance in preparation of the chapter.


1.Bandura A. Principles of behavior modification. New York, NY: Holt, Rinehart and Winston, 1969:509.

2.Elkins RL. An appraisal of chemical aversion (emetic therapy) approaches to alcoholism treatment. Behav Res Ther 1991;29(3):387–413.

3.Smith JW. Treatment of alcoholism in aversion conditioning hospitals. In: Pattison EM, Kaufman E, eds. Encyclopedic handbook of alcoholism. New York, NY: Gardner Press, 1982:874–884.

4.Smith C, Bookner S, Dreher F. Effects of alcohol intoxication and hangovers on subsequent drinking. NIDA Research Monograph 90. Rockville, MD: National Institute on Drug Abuse, 1988:366.

5.Ritchie JM. The aliphatic alcohols. In: Gilman AG, Goodman LS, Gilman A, eds. The pharmacological basis of therapeutics, 6th ed. New York, NY: Macmillan, 1980.

6.Cannon DS, Baker TB, Gino A, et al. Alcohol-aversion therapy: relation between strength of aversion and abstinence. J Consult Clin Psychol 1986;54(6):825–830.

7.deSilva P, Rachman S. Human food aversions: nature and acquisition. J Consult Clin Psychol 1987;25(6):457–468.

8.Hoeschen LE. The pharmacokinetics and pharmacodynamics of alcohol and drugs of addiction. In: Miller NS, ed. Comprehensive handbook of drug and alcohol addiction. New York, NY: Marcel Dekker, 1991:745–746.

9.Lemere F, Voegtlin WL. Conditioned reflex therapy of alcoholic addiction: specificity of conditioning against chronic alcoholism. Cal West Med 1940;53(6):1–4.

10.Quinn JT, Henbest R. Partial failure of generalization in alcoholics following aversion therapy. Q J Stud Alcohol 1967;28:70–75.

11.Lemere F, Voegtlin WL, Broz WR, et al. Conditioned reflex treatment of chronic alcoholism: VII. Dis Nerv Syst 1942;3(8):59–62.

12.Schwartz B. Psychology of learning and Co., 1978:55.

13.Voegtlin WL, Lemere F, Broz WR, et al. Conditioned reflex therapy of chronic alcoholism: IV. A preliminary report on the value of reinforcement. Q J Stud Alcohol 1941;2(3):505–511.

14.Smith JW, Frawley PJ, Polissar L. Six- and twelve-month abstinence rates in inpatient alcoholics treated with aversion therapy compared with matched inpatients from a treatment registry. Alcohol Clin Exp Res 1991;15(5):862–870.

15.Pomerleau OF, Fertig J, Baker L. et al. Reactivity to alcohol cue in alcoholics and non-alcoholics: implications for a stimulus control analysis of drinking. Addict Behav 1983;8:1–10.

16.Niaura R, Abrams D, Demuth B, et al. Responses to smoking-related stimuli and early relapse to smoking. Addict Behav 1989;14:419–428.

17.Sherman JE, Zinser MC, Sideroff SI, et al. Subjective dimensions of heroin urges: influence of heroin-related and affectively related stimuli. Addict Behav 1989;14:611–623.

18.Weddington WW, Brown BS, Haertzen CA, et al. Changes in mood, craving, and sleep during short-term abstinence reported by male cocaine addicts. Arch Gen Psychiatry1990;47:861–868.

19.Gawin FH, Kleber HD. Abstinence symptomatology and psychiatric diagnosis in cocaine abusers. Arch Gen Psychiatry 1986;43:107–113.

20.Sachs DPL. Advances in smoking cessation treatment. Curr Pulmonol 1991;12:139–198.

21.Sachs DPL. Cigarette smoking: health effects and cessation strategies. Clin Geriatr Med 1986;2(2):337–363.

22.Hall RG, Sachs DPL, Hall SM, et al. Two-year efficacy and safety of rapid smoking therapy in patients with cardiac and pulmonary disease. J Consult Clin Psychol 1984;52(4):574–581.

23.Glasgow RE, Lichtenstein E, Beaver C, et al. Subjects reactions to rapid and normal paced aversive smoking. Addict Behav 1981;6: 53–59.

24.Hall S, Tunstall C, Rugg D, et al. Nicotine gum and behavioral treatment in smoking cessation. J Consult Clin Psychol 1985;53(2):256–258.

25.Lando HA. Successful treatment of smokers with a broad-spectrum behavioral approach. J Consult Clin Psychol 1977;45(3):361–366.

26.Hall SM, Rugg D, Tunstall C, et al. Preventing relapse to cigarette smoking by behavioral skill training. J Consult Clin Psychol 1984;52(3):372–382.

27.Erickson LM, Tiffany ST, Martin EM, et al. Aversive smoking therapies: a conditioning analysis of therapeutic effectiveness. Behav Res Ther 1983;21(60):595–611.

28.Hajek P, Stead LF. Aversive smoking for smoking cessation. Cochrane Database Syst Rev 1997;4:CD000546. doi:10.1002/14651858. CD000546.pub2.

29.Smith JW. Long-term outcome of clients treated in a commercial stop smoking program. J Subst Abuse 1988;5:33–36.

30.Berecz JM. Reduction of cigarette smoking through self-administered aversion conditioning: a new treatment model with implications for public health. Soc Sci Med 1972;6:57–66.

31.Vaillant GE. The natural history of alcoholism revisited. Cambridge, MA: Harvard University Press, 1995:231–277.

32.Vaillant GE. The natural history of alcoholism: causes, patterns and paths to recovery. Cambridge, MA: Harvard University Press, 1983:128.

33.Voegtlin WL, Broz WR. The conditioned reflex treatment of chronic alcoholism. X. An analysis of 3,125 admissions over a period of ten and a half years. Ann Intern Med 1949;30:580–597.

34.Boland FJ, Mellor CS, Revusky S. Chemical aversion treatment of alcoholism: lithium as the aversive agent. Behav Res Ther 1978;16:401–409.

35.Cannon DS, Baker TB, Wehl CK. Emetic and electric shock alcohol aversion therapy: six- and twelve-month follow-up. J Consult Clin Psychol 1981;49(3):360–368.

36.Jackson TR, Smith JW. A comparison of two aversion treatment methods for alcoholism. J Stud Alcohol 1978;39(1):187–191.

37.Smith JW, Frawley PJ, Polissar L. Six- and twelve-month abstinence rates in inpatient alcoholics treated with either faradic aversion or chemical aversion compared with matched inpatients from a treatment registry. J Addict Dis 1997;16(1):5–24.

38.Blake BG. The application of behaviour therapy to the treatment of alcoholism. Behav Res Ther 1965;3:77–85.

39.Neueberger OW, Hasha H, Matarazzo JD, et al. Behavioral-chemical treatment of alcoholism: an outcome replication. J Stud Alcohol 1981;42(9):806–810.

40.Wiens AN, Menustik CA. Treatment outcome and patient characteristics in an aversion therapy program for alcoholism. Am Psychol 1983;38(10):1089–1096.

41.Thurber S. Effect size estimates in chemical aversion treatments of alcoholism. J Clin Psychol 1985;41(2):285–287.

42.Caddy GR, Block T. Behavioral treatment methods for alcoholism. Recent Dev Alcohol 1983;1:139–165.

43.Garcia J, Koelling RA. Relation of cue consequence in avoidance learning. Psychon Sci 1966;4:123–131.

44.Howard MO. Pharmacological aversion treatment of alcohol dependence:1. Production and prediction of conditioned alcohol aversion. Am J Drug Alcohol Abuse 2001;27(3):561–585.

45.Elkins RL. Covert sensitization treatment of alcoholism: contributions of successful conditioning to subsequent abstinence maintenance. Addict Behav 1980;5:67–89.

46.Elkins RL. Aversion therapy for alcoholism: Chemical, electrical, or verbal imagery? Int J Addict 1975;10(2):157–209.

47.Smith JW, Schmeling G, Knowles PL. A marijuana smoking cessation clinical trial utilizing THC-free marijuana, aversion therapy, and self-management counseling. J Subst Abuse Treat1988;5(2):89–98.

48.Rawson RA, Obert JL, McCann MJ, et al. In: Harris LS, ed. Cocaine treatment outcome: cocaine use following inpatient, outpatient and no treatment (NIDA Research Monograph 67). Rockville, MD: National Institute on Drug Abuse, 1986:271–277.

49.Frawley PJ, Smith JW. Chemical aversion therapy in the treatment of cocaine dependence as part of a multimodal treatment program: treatment outcome. J Subst Abuse Treat1990;7:21–29.

50.Frawley PJ, Smith JW. One-year follow-up after multimodal inpatient treatment for cocaine and methamphetamine dependence. J Subst Abuse Treat 1992;9(4):271–286.

51.Elkins RL, Stoddard JL, Walters PA, et al. Aversion therapy treatment of cocaine dependent persons: an experimental evaluation. Addict Behav 2008 (in review).

52.Copemann CD. Drug addiction: II. An aversive counterconditioning technique for treatment. Psychol Rep 1976;38:1271–1281.

53.Elkins, R. “Hypnotically-Guided Counter Conditioning for Heroin Dependence: Case Reports of Anti-Craving Benefits and Abstinence Outcomes” at the Society for Clinical & Experimental Hypnosis’s 64th Annual Scientific Session October 4–6, 2013, Berkeley.

54.Smith JW, Frawley PJ. Treatment outcome of 600 chemically dependent patients treated in a multimodal inpatient program including aversion therapy and pentothal interviews. J Subst Abuse Treat1993;10:359–369.

55.Beaubrum MH. Treatment of alcoholism in Trinidad and Tobago. Br J Psychiatry 1967;113:643–658.

56.Baker TB, Cannon DS. Taste aversion therapy with alcoholics: techniques and evidence of a conditioned response. Behav Res Ther 1979;17:229–242.

57.Howard MO, Elkins RL, Rimmele C, et al. Chemical aversion treatment for alcohol dependence. Drug Alcohol Depend 1991;29:107–143.

58.Bordnick PS, Elkins RL, Orr TE, et al. Evaluating the relative effectiveness of three aversion therapies designed to reduce craving among cocaine abusers. Behav Intern 2004;19:1–24.

59.Marlatt GA, Gordon JR. Determinants of relapse: implications for the maintenance of behavior change. In: Davidson PO, Davidson SM, eds. Behavioral medicine: changing health lifestyles. New York, NY: Bruner/Mazel, 1980.

60.Kattwinkel EE. Death due to cardiac disease following the use of emetine hydrochloride in conditioned-reflex treatment of chronic alcoholism. JAMA 1949;240(25):995–997.

61.Loomis TA. Emetine risk analysis for the Shadel Hospital’s aversion therapy program. Arlington, VA: Drill, Freiss, Hays, Loomis & Shaffer, 1986.

62.Klatskin G, Friedman H. Emetine toxicity in man: studies on the nature of early toxic manifestations, their relation to the dose level, and their significance in determining safe dosage. Ann Intern Med1948;28:892–915.

63.Smith, et al. Unpublished data from CATORIII Chemical Abuse/ Addiction Treatment Outcome Registry-modification III. St. Paul, MN.

64.Wiens AN, Menustik CE, Miller SI, et al. Medical-behavioral treatment of the older alcoholic. Am J Drug Alcohol Abuse 1982;9(4);461–475.

65.Smith JW, Frawley PJ. Long-term abstinence from alcohol in patients receiving aversion therapy as part of a multimodal inpatient program. J Subst Abuse Treat 1990;7:77–82.

66.Gilmore K. Hazelden primary residential treatment program: 1983 profile and patient outcome. Center City, MN: Hazelden Foundation, 1985.

67.Patton M. The outcomes of treatment: a study of patients admitted to Hazelden in 1976. Center City, MN: Hazelden Foundation, 1979.

68.Carter E. Chemical aversion therapy for the treatment of alcoholism. Health technology assessment reports, 4. Washington, DC: National Center for Health Care Services Research and Health Care Technology Assessment, 1987.

69.Mendes E. Letter to Honorable Jamie L. Whitten, Chairman, Committee on Appropriations, House of Representatives. Washington, DC, 1992.

70.California Medical Association. Scientific Advisory Panels on General and Family Practice and Internal Medicine and the Committee on Alcoholism and Other Drug Dependence. Medical practice question: is aversion therapy for the treatment of alcoholism considered accepted medical practice or is it investigational? San Francisco, CA: California Medical Association, 1984.

71.American Society of Addiction Medicine. Statement on treatment for alcoholism and other drug dependencies. Chevy Chase, MD: The Society, 1986.

72.American Medical Association. Aversion therapy report of the Council on Scientific Affairs. JAMA 1987;258(18):2562–2566.

73.Saxe L, Dovaterty D, Esty K, et al. Research on the effectiveness of alcoholism treatment. Health technology case study 22: the effectiveness and costs of alcoholism treatment. Washington, DC: Office of Technology Assessment, U.S. Congress, 1983:43–53.

74.Institute of Medicine. Broadening the base of treatment for alcohol problems. Washington, DC: National Academy Press, 1990.

75.Wilson GT. Chemical aversion conditioning as a treatment for alcoholism: a re-analysis. Behav Res Ther 1987;25(6):503–516.

76.Howard MO, Jenson JM. Chemical aversion treatment of alcohol dependence: 1. Validity of current criticisms. Int J Addict 1990;25(10):1227–1262.

77.Wilson GT. Chemical aversion conditioning in the treatment of alcoholism: further comments. Behav Res Ther 1991;5:415–419.

78.Elkins RL. Chemical aversion emetic therapy treatment of alcoholism: further comments. Behav Res Ther 1991;5:421–428.

79.Project MATCH Research Group. Matching alcoholism treatments to client heterogeneity: treatment main effects and matching effects on drinking during treatment. J Stud Alcohol1998;59(6):631–639.

80.Volkow ND, Wang GJ, Telang F, et al. Cocaine cues and dopamine in dorsal striatum: mechanism of craving in cocaine addiction. J Neurosci 2006;26(24):6583–6588.

81.Kilts CD, Gross RE, Ely TD, et al. The neural correlates of cue-induced craving in cocaine-dependent women. Am J Psychiatry 2004;161:223–241.

82.Elkins RL, Walters PA, Orr TE. Continued development and unconditioned stimulus characterization of selectively bred lines of taste aversion prone and resistant rats. Alcohol Clin Exp Res 1992;16:928–934.

83.Elkins RL. Separation of taste-aversion-prone and taste- aversion-resistant rats through selective breeding: implications for individual differences in conditionability and aversion-therapy alcoholism treatment. Behav Neurosci 1986;100(1):121–124.

84.Orr TE, Whitford-Stoddard JL, Elkins RL. Taste-aversion-prone TAP rats and taste-aversion-resistant TAR rats differ in ethanol self-administration, but not in ethanol clearance or general consumption. Alcohol 2004;33:1–7.

85.Orr TE, Walters PA, Elkins RL. Differences in free-choice ethanol acceptance between taste-aversion-prone and taste-aversion-resistant rats. Alcohol Clin Exp Res 1997;21:1491–1496.

86.Elkins RL, Orr TE, Rausch JL, et al. Cocaine-induced expression differences in glutamate receptor subunits and transporters in amygdalae of taste-aversion-prone and taste-aversion-resistant rats. Ann N Y Acad Sci 2003;1003:381–385.

87.Elkins RL, Orr TE, Rausch JL, et al. Cocaine-induced expression differences in PSD-95/SAP-90-associated protein 4 and in Ca2+/ Calmodulin-dependent protein kinase subunits in amygdalae of taste-aversion-prone and taste-aversion-resistant rats. Ann N Y Acad Sci 2003;1003:386–390.

88.Elkins RL, Orr TE, Edwards GL, et al. Cocaine-induced expression differences in of 5-HT3 receptors and Na+/K + APTase pump subunits in amygdalae of taste-aversion-prone and taste-aversion-resistant rats. Ann N Y Acad Sci 2003;985: 519–521.

89.Riley AL, Tuck DL. Conditioned food aversions: a bibliography. Ann N Y Acad Sci 1985;443:381–437.

90.Riley AL, Freeman KB. Conditioned taste aversion: an annotated bibliography.