CYBERPSYCHOLOGY & BEHAVIOR

Volume 4, Number 4, 2001Mary Ann Liebert, Inc.

Immersive Virtual Environments in Cue Exposure

MARCUS F. KUNTZE, M.D.,1 ROBERT STOERMER, M.D., Ph.D.,1 RALPH MAGER, M.D.,1

ANDREAS ROESSLER, Ph.D.,2 FRANZ MUELLER-SPAHN, M.D.,1

and ALEX H. BULLINGER, M.D.1

ABSTRACT

Cue reactivity to drug-related stimuli is a frequently observed phenomenon in drug addic-tion. Cue reactivity refers to a classical conditioned response pattern that occurs when an ad-dicted subject is exposed to drug-related stimuli. This response consists of physiological andcognitive reactions. Craving, a subjective desire to use the drug of choice, is believed to playan important role in the occurrence of relapse in the natural setting. Besides craving, othersubjective cue-elicited reactions have been reported, including withdrawal symptoms, drug-agonistic effects, and mood swings. Physiological reactions that have been investigated in-clude skin conductance, heart rate, salivation, and body temperature. Conditioned reactivityto cues is an important factor in addiction to alcohol, nicotine, opiates, and cocaine. Cue ex-posure treatment (CET) refers to a manualized, repeated exposure to drug-related cues, aimedat the reduction of cue reactivity by extinction. In CET, different stimuli are presented, forexample, slides, video tapes, pictures, or paraphernalia in nonrealistic, experimental settings.Most often assessments consist in subjective ratings by craving scales. Our pilot study willshow that immersive virtual reality (IVR) is as good or even better in eliciting subjective andphysiological craving symptoms as classical devices.

497

INTRODUCTION

AT THE PSYCHIATRIC UNIVERSITY CLINIC inBasel, Switzerland, there is ongoing re-

search in immersive virtual reality (IVR).2,18

The Center of Applied Technologies in MentalHealth (COAT) is part of the clinic and pro-motes studies in IVR, especially with partnersin the European Union. The COAT-Basel divi-sion “Integrated Multimedia in Cognitive-Be-havioral Therapy” works on information tech-nology and cognitive-behavioral therapy ingeneral as well as on specific diagnostic and

therapeutic applications of IVR in mentalhealth. Besides neurocognitive disorders16 andspecific phobias,3 there is research with drugaddicted patients.12

Our heroin and methadone outpatient cliniccomprises two groups of opioid-dependentsubjects. Methadone maintenance is well es-tablished in many countries as an effectivetreatment. It reduces illicit drug use, infectiousdiseases, and criminal activity. The methadonemaintenance outpatient clinic is specialized intreating addicted patients with comorbid dis-orders.12

1Department of Clinical Psychiatry, University of Basel, Basel, Switzerland.2ICIDO Ltd., Stuttgart, Germany.

In Switzerland, multicity trials with heroinmaintenance were initiated in 1994.24 Basel isone of the biggest study centers, with 150 sub-jects in treatment. Heroin maintenance is sim-ilar to methadone maintenance insofar as it in-volves the medical prescription of a narcotic asone component of treatment for addiction to anopioid. This maintenance provides restrictedlegal access to the same pharmacological classof drugs that subjects previously obtained byother means.

In IVR, a standardized and reliable assess-ment center provides the opportunity to mea-sure subjective and physiological data while asubject is exposed to relevant stimuli. There-fore, reasonable effect sizes can be expected notonly for subjective measurements but also forobjective ones. Our pilot study has two objec-tives. Firstly, will IVR be able to elicit subjec-tive and physiological craving symptoms inopioid-dependent patients? Secondly, is IVReffective compared to standard devices in elic-iting craving? Craving can be defined as astrong subjective desire to use a drug and is be-lieved to play an important role in the occur-rence of relapses.

SUBJECTS AND METHODS

Technical devices

Programming was done with True Space,TCL (both software packages are commerciallyavailable) and Lightning (custom built virtualinteraction and animation software package,ICIDO Ltd., Stuttgart, Germany). The virtualenvironment (VE) was visualized by a SiliconGraphics (SGI) Onyx2 Deskside System for dataprocessing, a head mounted display (V8, Vir-tual Research), a tracking system (Flock ofBirds, Ascension), and a custom built interac-tion device (Space Mike, ICIDO Ltd.).

A multimodal acquisition and stimulationsystem was used for on-line data acquisition,recording, and processing.1,21 Electrocardio-graphic (ECG) activity was recorded as singlederivation from RA (right anterior), LA (left an-terior), and G (ground) using an Escort 2 Mul-tifunction-monitor/amplifier (MDE Inc.), thesame signals were additionally utilized for im-

pedance pneumography. Oxygen saturationwas measured at the middle finger. A contin-uous noninvasive finger-hemodynamometrywas realized by a Finapress-system (Finapress2200, Datex Ohmeda).

Digital converting was done by an ADA3300Board (16*12bit, Analog Devices). The systemwas integrated by PC-based data-acquisitionsoftware (TurboLab, Bressner Technology,Germany). This integration results in real-timevisualization, data processing, and storage. Allevents are triggered online by the SGI work-station, according to the respective user–VE in-teraction.

The different physiological signals men-tioned above were transformed to clinically rel-evant and well-known dimensions. After reso-lution of the QRS complexes, the single-channelelectrocardiogram (ECG) derivation yieldedparameters such as the interbeat interval and acardiotachogram. The cardiotachogram wasstatistically evaluated, furthermore a spectralanalysis was performed using a Fast Hartley–transformation. The signal generated by the im-pedance pneumography reflects the air contentof the thoracic/pulmonary tissue as modulatedby respiration. The latter signal can be used tocalculate the curve of the respiratory rate afterseparating those parts of the signal that arecaused by aortic pulsation. The systolic and di-astolic blood pressure such as the arterial meanpressure and the amplitude of blood pressurewere derived from the pulse wave curve asgenerated by the Finapress system.

Virtual environment

Cue exposure is presented in a virtual bar(Fig. 1). Tables, other furniture, and decorationin a neutral style offer the background for pre-sentation of paraphernalia. We use, as stimuli,diacetylmorphine powder (DAM, heroin),DAM lyophylisat, swab, syringe, needle, andused material with/without blood.

We apply two different orders of cue pre-sentation. Order A is “naturalistic.” Firstly, theempty bar appears. Secondly, one swab fol-lowed by a lot of swabs show up. Then needleand syringe are presented. Afterwards, DAMas powder and as lyophylisat, and then usedmaterial without (and, last, with) blood arises

KUNTZE ET AL.498

on the counter. The exposure ends with anempty immersive virtual bar environment.

Order of appearance B is “by chance”: acomputer-based stochastic program selectseach cue individually.

Study design

We assess three different groups (Fig. 2).Group I is exposed in IVR, group II receivescue exposure with pictures, and group III is ex-posed to neutral-control stimuli (books, bread).Each group contains five subjects. Standard-ized diagnostic procedures such as SCID I and

II (Structured Clinical Interview for DSM-IV,axis I and axis II disorders) and ASI (AddictionSeverity Index) are applied at study intake.Physiological activation at baseline, during ex-posure in 5-min intervals, and after exposureare measured. Subjective ratings are a VAS (Vi-sual Analog Scale), the Y-BOCS (Yale-BrownObsessive Compulsive Scale), and differentcraving scales, which are taken at baseline, andduring and after exposure. The manual-guidedexposure itself lasts approximately 20 min.

After cue exposure, we apply manual-guided and supervized cognitive therapy11 andprogressive muscle relaxation techniques. We

IMMERSIVE VIRTUAL ENVIRONMENTS IN CUE EXPOSURE 499

FIG. 1. Immersive virtual environment “BAR.”

FIG. 2. Assessment.

try to modify cognitive distortions (all-or-noth-ing thinking, must/should/never statements,overgeneralization, catastrophizing, emotionalreasoning), dysfunctional beliefs and expecta-tions, and automatic thoughts. Relaxationtraining is used to reduce physiological tensioninduced by the stimuli.

Subjects

After getting informed consent, we includedotherwise healthy male subjects with an opioiddependence (according to DSM-IV) currentlytreated in our heroin maintenance outpatientclinic in Basel, Switzerland. The subjects werebetween 25 and 45 years old.

RESULTS AND DISCUSSION

On a learning-based model (Fig. 3),20 weevaluated how craving can be induced in opi-oid-dependent subjects using IVR. Condition-ing processes are an important factor in addic-tion according to, for example, Siegel19 andKouimtsidis.10 External and internal stimuliprovoke cognitive, physiological, emotional,and behavioral reactions. Therefore, differentkind of physiological and subjective data canbe recorded.26 We refer to Hamilton et al.8 andVerheul et al.25 in our physiological assess-ments. Craving, physiological activation, andcognitive and emotional coping strategies,9,13

and finally drug use are consequences of spe-cific stimuli inherent to drugs such as alco-hol,17,22 opioids,6 cocaine,5 and nicotine.14,15 Es-

pecially the strong desire to use a drug playsan imporant role in the natural setting.7,23

Usually, in cue exposure treatment (CET) dif-ferent stimuli are presented using slides, videotapes, or pictures. In 1999, Carter et al.4 re-ported on a meta-analysis study of 41 studiesusing both drug-related and neutral-controlstimuli. Studies assessing heroin-dependentsubjects included a total of 282 subjects (aver-age age, 30 years). They exclusivly used pic-tures or video presentation. The effect sizeswere much bigger for self-report scales than forphysiological measures. Immersive virtual en-vironments (IVR) can solve some of the prob-lems mentioned by Carter et al.4 Our technol-ogy provides a standardized and reliableassessment center for all physiological reac-tions. Even brain waves can be investigated inrespect to specific pattern or regional distribu-tion. In order to get reliable data, we have coun-terbalanced the order of cue presentation andwe compare to neutral stimuli.

Some ethical concerns have to be addressed.One can expect aggravation of the substanceuse disorder. IVR could have effects like tele-vision commercials for drug use. Because cueexposure is intented to induce craving, it canlead to subsequent drug use as a result of ourstudy design. We apply cognitive therapy andprogressive muscle relaxation technique afterexposure to minimize these risks.

Fortunately, we have not experienced any se-vere incidents. Our VE works well in elicitingphysiological activation and subjective cravingsymptoms. In a pilot study, five subjects re-

KUNTZE ET AL.500

FIG. 3. Learning-based model of cue exposure. Stimuli (S) provoke reactions (R; craving and physiological activa-tion) leading to coping, drug use, and its consequences (C).

ported unpleasant feelings and urges to usedrugs, and they were highly autonomicallystimulated. Right now, we are working on acontrolled study using our experiences in thepilot study.

REFERENCES

1. Bullinger, A., Kuntze, M., & Mueller-Spahn, F. (1997).Computer-generated (virtual reality) three-dimen-sional exposure as a tool in behavioral therapy ofclaustrophobia and acrophobia. In: Proceedings of theWorld Congress on Rehabilitation in Psychiatry. Belgrade,p. 50.

2. Bullinger, A., Roessler, A., & Mueller-Spahn, F. (1998).3D-virtual reality as a tool in behavioral therapy ofclaustrophobia. CyberPsychology & Behavior, 1:139–145.

3. Bullinger, A., Roessler, A., & Mueller-Spahn, F. (1998).From toy to tool: the development of immersive vir-tual reality environments for psychotherapy of spe-cific phobias. Studies in Health Technology Information,58:103–111.

4. Carter, B.L., & Tiffany, S.T. (1999). Meta-analysis ofcue reactivity in addiction research. Addiction, 94:327–340.

5. Darke, S., & Hall, W. (1995). Levels and correlates ofpolydrug use among heroin users and regular am-phetamine users. Drug and Alcohol Dependence, 39:231–235.

6. Franken, I.H.A., & De Haan, H.A. (1999). Cue reac-tivity and effects of cue exposure in abstinent post-treatment drug users. Journal of Substance Abuse Treat-ment, 16:81–85.

7. Gossop, M., & Powell, J. (1990). What do opiate ad-dicts and cigarette smokers mean by “craving”? A pi-lot study. Drug and Alcohol Dependence, 26:85–87.

8. Hamilton, M.E., & Voris, J.C. (1998). Money as a toolto extinguish conditioned responses to cocaine in ad-dicts. Journal of Clinical Psychology, 54:211–218.

9. Hasenfratz, M., & Bättig, K. (1991). Psychophysiolog-ical reactions during active and passive stress copingfollowing smoking cessation. Psychopharmacology,104:356–362.

10. Kouimtsidis, C. (2000). Role of craving in substancemisuse. Current Opinion in Psychiatry, 13:299–303.

11. Kuntze, M., Bullinger, A., & Mueller-Spahn, F. (1997).Cognitive restructuring and prolonged exposure as atreatment program for posttraumatic stress disorder(PTSD): a case report [in German]. Verhaltenstherapie,7:161–167.

12. Kuntze, M., Stohler, R., & Ladewig, D. (1998). Preva-lence of comorbid disorders in a methadone mainte-nance outpatient clinic [in German]. Sucht—Zeitschriftfür Wissenschaft und Praxis, 44:96–103.

13. Lubman, D.I., & Peters, L.A. (2000). Attentional bias

for drug cues in opiate dependence. Psychological Med-icine, 30:169–175.

14. Mucha, R.F., & Geier, A. (1999). Modulation of crav-ing by cues having differential overlap with pharma-cological effect: evidence for cue approach in smok-ers and social drinkers. Psychopharmacology, 147:306–313.

15. Mucha, R.F., & Pauli, P. (1998). Conditioned re-sponses elicited by experimentally produced cues forsmoking. Canadian Journal of Physiology and Pharma-cology, 76:259–268.

16. Mueller-Spahn, F., Hofmann, M., Roessler, A., &Bullinger, A. (1998). 3D virtual reality (VR) systems,cognition and creativity. European Psychiatry, 13:147s.

17. Myers, R.J., & Smith, J.E. (1995). Clinical guide to alco-hol treatment: the community reinforcement approach.New York: Guilford Press.

18. Roessler, A., Mueller-Spahn, F., Baehrer, S., &Bullinger, A. (2000). A rapid prototyping frameworkfor the development of virtual environments in men-tal health. CyberPsychology & Behavior, 3:359–367.

19. Siegel, S. (1999). Drug anticipation and drug addic-tion. The 1998 H. David Archibald lecture. Addiction,94:1113–1124.

20. Siegel, S., & Hinson, R.E. (1982). Heroin overdosedeath: contribution of drug-associated environmentalcues. Science, 216:436–437.

21. Stoermer, R., Mager, R., Roessler, A., Mueller-Spahn,F., & Bullinger, A. (2000). Monitoring human-virtualreality-interaction: a time series analysis approach.CyberPsychology & Behavior, 3:401–406.

22. Stormark, K.M., & Field, N.P. (1997). Selective pro-cessing of visual alcohol cues in abstinent alcoholics:an approach-avoidance conflict? Addictive Behaviors,22:509–519.

23. Tiffany, S.T., & Carter, B.L. (1998). Is craving thesource of compulsive drug use? Journal of Psycho-pharmacology, 12:23–30.

24. Uchtenhagen, A., Dobler-Mikola, A., Steffen, T.,Gutzwiller, F., Blättler, R., & Pfeifer, S. (1999). Pre-scription of narcotics for heroin addicts. Basel, Switzer-land: Karger.

25. Verheul, R., & Van den Brink, W. (1999). A three-path-way psychobiological model of craving for alcohol.Alcohol & Alcoholism, 34:197–222.

26. Weinstein, A., & Feldtkeller, B. (1998). Integrating thecognitive and physiological aspects of craving. Jour-nal of Psychopharmacology, 12:31–38.

Address reprint requests to:Dr. Marcus F. Kuntze

PUK BaselWilhelm Klein—Strasse 27CH-4025 Basel, Switzerland

E-mail: marcus.kuntze@pukbasel.ch

IMMERSIVE VIRTUAL ENVIRONMENTS IN CUE EXPOSURE 501