Addiction (1996) 91(2), 199-220

RESEARCH REPORT

WHO cross-cultural applicability research ondiagnosis and assessment of substance use

~ disorders: an overview of methods andI selected results,t"t.

1 2 3R. ROOM, A. JANCA, L. A. BENNETT, L. SCHMIDT& N. SARTORIUS4

Addiction Research Foundation.. Toronto; 1 Division of Mental Health.. World Health

Organization.. Geneva... 2 Department of Anthropology.. Memphis State University.. Memphis...3 Alcohol Research Group.. Berkeley & 4 Department of Psychiatry.. University of Geneva..

Geneva.. Switzerland

With: J. Blaine (NIDA, Washington DC); C. Campillo (Mexico City, Mexico);C.R. Chandrashekar (Bangalore, India); D. Dawson (NIAAA, Washington DC); A. Gogiis(Ankara, Turkey); B. Grant (NIAAA, Washington DC); O. Gureje (lbadan, Nigeria);Ho Young Lee (Seoul, Republic of Korea); V. Mavreas (Athens, Greece); J. Rolf (Baltimore,Maryland); L. Towle (NIAAA, Washington DC); R. Trotter, III (Flagstaff, Arizona);T.B. OstUn (WHO, Geneva); J.L. Vazquez-Barquero (Santander, Spain) & R. Vrasti Gebel,

Romania)

AbstractThe cross-cultural applicability of criteria for the diagnosis of substance use disorders and of instruments used

.for their assessment were studied in nine cultures. The qualitative and quantitative methods used in the studyare described. Equivalents for English terms and concepts were found for all instrument items, diagnosticcriteria, diagnoses and concepts.. although often there was no single term equivalent to the English in thelanguages studied. Items assuming self-consciousness about feelings, and imputing causal relations, posed

~ difficulties in several cultures. Single equivalent terms were lacking for some diagnostic criteria, and criteria

were sometimes not readily differentiated from one another. Several criteria-narrowing of the drinkingrepertoire, time spent obtaining and using the drug, and tolerance for the drug-were less easy to use incultures other than the United States. Thresholds for diagnosis used by clinicians often differed. In mostcultures, clinicians were more likely to make a diagnosis of drug dependence than of alcohol dependencealthough behavioural signs were equivalent. The attitudes of societies to alcohol and drug use affects the useof criteria and the making of diagnoses.

Responsible project officer: Dr A. Janca, Mental Health Division, World Health Organization, 1211 Geneva 27,Switzerland.

Correspondence to: Robin Room, Addiction Research Foundation, 33 Russell Street, Toronto, Ontario,Canada M5S 2S1.

0965-2140/96/020199-22 $8.00 @ Society for the Study of Addiction to Alcohol and other Drugs

Carfax Publishing Company

,

200 R. Room et al.

Introduction The international diagnostic instrumentsThe WHO/NIH Joint Project on diagnosis and SCAN is a semi-structured diagnostic instru-classification ment primarily designed for use in diagnosingDevelopment of cross-culturally applicable diag- disorders and syndromes by clinicians, particu-nostic criteria and instruments for the assess- larly psychiatrists and clinical psychologists; itment of mental disorders in different cultures may be administered by other health profession-has been one of the major goals in the WHO/ als with appropriate training. It offers consider-NIH Joint Project on Diagnosis and able freedom for interviewers to frame questionsClassification of Mental Disorders, Alcohol- and about a particular symptom, and requests thatDrug-related Problems. This reflects the com- they make a judgement about the presence ormitment of the World Health Organization absence of a symptom defined in the SCAN(WHO) to the development of a "common lan- glossary. Stemming from the tradition of the .

guage" (Sartorius, 1989) which will allow mental Present State Examination (PSE) developed athealth professionals and others concerned with Maudsley Hospital in London, the SCAN hasthe management of mental health and psychoso- been developed as a clinical assessment tool for .cial problems to understand one another and the assessment of a broad range of psychiatricwork together. symptoms, syndromes and disorders listed in the

The Joint Project started in 1979 as a collabo- current classification systems (for details seerative endeavour between WHO and three US Wing et al., 1990).National Institutes, the National Institute of CIDI is a highly structured interview scheduleMental Health (NIMH), the National Institute which is designed to be administered by trainedon Alcohol Abuse and Alcoholism (NIAAA) and lay interviewers. The instrument consists of fullythe National Institute on Drug Abuse (NIDA)- spelt-out questions, fixed coding options and aformerly parts of the Alcohol, Drug Abuse and clearly specified probing system that allow theMental Health Administration (ADAMHA) and interviewer to determine the severity and likelynow research institutes within the National Insti- psychiatric significance of a positive symptom.tutes of Health (NIH). In a series of workshops The instrument is designed for adult respon-organized at the beginning of the project experts dents with varying educational and culturalfrom different countries, different cultures and a backgrounds. Stemming from the Diagnostic In-variety of clinical and social science traditions terview Schedule (DIS) which was developed atwere invited to define problems and recommend the Washington University, St Louis and appliedactivities which could lead to the development of in the US Epidemiological Catchment Areacrossculturally applicable instruments for the as- studies (Robins & Regier, 1991), the CIDI issessment of various aspects of mental and be- primarily intended for epidemiological studies ofhavioural disorders, to the formulation of criteria mental disorders in different cultures and set-for their diagnosis, and to the adoption of sci- tings (for details, see Robins et al., 1989).entifically and practically useful international The traditions from which the CIDI andclassifications (Sartorius, 1989). SCAN stem have been distinguished by a com- .

More than 100 centres from allover the world mitment to the operationalization of diagnostichave been participating in the various activities concepts and categories in a reliable form. Forof the WHO/NIH Joint Project. Major achieve- each diagnosis, a set of operational criteria werements of this fruitful international collaboration developed which were in principle objectively .have been related to the development of clinical and reliably measurable. The reliability of anand research diagnostic criteria for the ICD-I0 operational measure-the degree to which itClassification of Mental and Behavioural Disor- could be reproduced by an application of theders (WHO, 1992a; 1993) and instruments for same measure a second time or by a secondtheir assessment. The Composite International diagnostician-became a key criterion of its ac-Diagnostic Interview (CIDI)(Robins et al., 1989; ceptability. With the issuance of the Third Re-WHO, 1990), and the Schedules for Clinical vision of the Diagnostic and Statistical Manual ofAssessment in Neuropsychiatry (SCAN) (Wing the American Psychiatric Association (DSM-et al., 1990; WHO, 1992b), are members of the III), this approach gained general acceptance infamily of instruments developed within the the United States, and increasingly also in otherWHO/NIH Joint Project. countries. Its influence is strong in the mental

Cross-cultural applicability of diagnoses 201

disorders chapter of the new version of the Inter- increasing interest in applying clinical instru-national Classification of Diseases, ICD-I0. ments developed in one culture in another cul-With the advent of ICD-I0's Classification of tural situation. Applying the Munich AlcoholismMental and Behavioural Disorders (WHO, Test (MALT) developed in Germany to samples1993), it might truly be said that the emphasis in Spain and Ecuador, Gorenc et al. (1984)on operationalizability as the sine qua non of found that five of the 31 items were "relativelydiagnosis of mental disorders has become a free of cultural differences" by their criteria, butworld standard. the authors added that when used in Ecuador

none of the items passed all five of the filtersused to screen out items in the original German

The application of alcohol and drug concepts and study.°, diagnostic instruments cross-culturally The most ambitious effort in this direction was

Operationalization is necessary for major studies well under way before the CAR project startedin psychiatric epidemiology and in clinical trials. (Helzer & Canino, 1.992). The effort was a

.It may also be helpful to clinical practice; but it serendipitous by-product of the wide inter-.is not without hazards. This is particularly true national use of the illS, an instrument originally

when the resulting measures are to be used to developed for use in the United States. In gen-compare rates or trends across cultures and soci- eral, the illS was applied without adaptation.eties, while the operational criteria are based on Since the analysis is primarily at the level ofmaterial drawn from a narrow cultural range. diagnoses, findings about the cross-cultural ap-The application of culturally specific descriptions plicability of the instrument are mainly in the Iand symptomatologies to other cultures may lead form of side-comments. An example of this is theto inappropriate diagnoses and conclusions notation that whether the "period of heavy(Klausner & Foulks, 1982, Chapter 16; Room, drinking" required for a dependence diagnosis1984). had to last 2 weeks or 4 weeks made an import-

There has already been an object lesson in this ant difference in how many received the diag-danger in the history of the concept of alco- nosis among American Indians in the Unitedholism. In the late 1940s and early 1950s, as States, given a "well-defined cultural pattern of"alcoholism" became the main alcohol-related binge or episodic heavy drinking" (Helzer &psychiatric disorder, it was defined in rather Canino, 1992, p. 126).culturally specific terms-in terms, in fact, of the An earlier WHO study also analysed the cross-experience of an emerging US-centred mutual cultural applicability of alcohol dependencehelp group, Alcoholics Anonymous. Thus the symptomatology (Hall et al., 1993). Using data-classic description of alcoholism by the leading sets from six divergent countries which com-alcohol scholar of the time, E. M. Jellinek bined together drinkers among general(1952), was based on the results of question- health-service patients and clinical alcoholics,naires about symptomatology developed by and the study found a strong general factor for 13circulated among Alcoholics Anonymous mem- dependence-related items in factor analysing

.bers in the United States. Only when Jellinek each country's data. This was interpreted ashad acquired a wider experience in the field by supporting the cross-cultural generalizability ofworking as a consultant to the World Health the alcohol dependence syndrome, although

.Organization in Geneva did he develop the idea alternative interpretations are possible for athat there were a number of culturally influenced finding of a general factor for symptomatic items

."species" of alcoholism, with different symp- in factor analyses in different societies.tomatologies Gellinek, 1960a, b). Jellinek's dif-ferent species, distinguished by Greek letters,mapped onto the very different denotations he The genesis and material of the Cross-Cultural

.found that "alcoholism" had among health pro- Applicability (CAR) Studyfessionals in different cultures-"gamma" for the Over the past 10 years both CIDI and SCAN"Anglo-Saxon" variety he had earlier described, were field-tested in more than 20 centres world-"delta" for the French variety, "epsilon" for the wide, and were found to be generally appropriateFinnish variety. and reliable for use across cultures and settings

Despite the possible pitfalls, there has been an (Wittchen et al., 1991). However, the field tests

202 R. Room et al.

did not include large numbers of alcohol and applicability of concepts, criteria and symptomsdrug users, so that the modifications of the sub- of substance use disorders and their operational-stance use disorders sections of the instruments izations, and (d) producing recommendationshad not been field-tested in different cultural concerning cross-cultural research in the field ofsettings. alcohol, drug use and mental health.

Accordingly, in September 1990 an advisory In the present paper we give an outline of thegroup recommended a substantial programme of CAR project and its findings on the applicabilityresearch on the cross-cultural applicability of the in different cultural circumstances of items, cri-alcohol and drug sections of the international teria, diagnoses and concepts particularly rel-diagnostic instruments. The programme was evant to four major alcohol diagnoses in ICD-10:conceived as having two phases: a study of the acute intoxication, harmful use of alcohol, themeanings and interpretations of alcohol and drug alcohol withdrawal syndrome and the alcohol ,"use and problems in different cultures, and of dependence syndrome.their implications for creating uniform diagnostic Our analysis is primarily based on English-Ian-standards and international instruments appli- guage reports of findings from the collaborating .cable across cultures; and a cross-cultural study teams of investigators at each study site. As each .

of the reliability and validity of the instruments. component of the study was finished, the investi-The present report is concerned with the first of gators at each site prepared a report in Englishthese phases, known as the Cross-Cultural Ap- on the results. In addition, each site prepared anplicability Research (CAR) study. The second overall report on the findings from that site forphase of the programme of research is under inclusion as a chapter in a book on the study (L.way. Bennett et aI., Use and Abuse of Alcohol and Drugs

The CAR study was thus carried out in nine in Different Cultures: A Nine-Country Study, insites world-wide, selected for their cultural and preparation). The present paper is based onlinguistic diversity: Ankara, Turkey; Athens, these reports, and quotes from them as appropri-Greece; Bangalore, India; Flagstaff, Arizona ate.(Navajo); Ibadan, Nigeria; lebel, Romania; Mex-ico City, Mexico; Santander, Spain; and Seoul,South Korea. The proximal aim of the study was Premises of the studyto test and as necessary improve the cross- As we have described, the central purpose of thecultural applicability of two existing international CAR project was to study the cross-cultural ap-diagnostic instruments-the CIDI and the plicability of the alcohol and drug portions of theSCAN. Our task was simplified by the fact that CIDI and SCAN instruments. Early in the proj-for alcohol and drug conditions, the SCAN was ect it was concluded that this required a studyclearly derivative from the CIDI, so that only for with a broader reach than simply settling thea few special topics was it necessary to cover two issue of whether CIDI or SCAN items could bedifferent approaches. translated and understood in different languages

The study's design included five sub studies and cultures. Behind the items lay the diagnosticwith diverse data collection techniques and sam- criteria they were designed to measure, and be- .

pIing frames, including a translation and back- hind the diagnostic criteria lay the diagnosestranslation study, key informant interviews and themselves and the conceptualizations on whichfocus groups with cultural informants, self-ad- they were based. A full understanding of cross- .ministered questionnaires filled out by local clin- cultural applicability and comparability requiredicians, and trials of diagnostic schedules with investigating the cultural relevance of and vari- ."reference cases" in alcohol and drug treatment ation in the diagnoses and criteria as well as infacilities. The findings of the study were planned the instrument items.to be used in: (a) future work improving of the The CIDI and SCAN specifically measurediagnostic instruments and developing guidelines three ICD-10 diagnoses: the Dependence Syn- .and instructions for their use in different cul- drome, Withdrawal State and Harmful Use. Thetures, (b) making the final adjustments for the instruments also measure the DSM-III-R's ver-large-scale testing of these instruments for re- sion of the first two diagnoses, along with theliability and validity in population-based sam- DSM Alcohol/Drug Abuse diagnosis. While eachpIes, (c) analysing of the cross-cultural of these diagnoses is given a technical meaningI

Cross-cultural applicability of diagnoses 203

Table 1. ICD-IO substance use diagnoses and their conceptuallocat:ion

Contrasting Related &(non-diagnostic) lay

Diagnoses Criteria states concepts

Acute Vs. intoxication,intoxication not medically

significant**Harmful Vs. normal use Abuse

use*Withdrawal Vs. hangover

state.Dependence Alcoholism,.syndrome: addiction

1. Strong desire, Craving,compulsion compulsion

2. Impaired capacity Loss of control" to control

*3. Withdrawal/ Vs. hangoveruse to relievewithdrawal

4. Tolerance5. Neglect of alternative

pleasures & activities**6. Harmful use Vs. normal use Abuse

(use despite physicalor psychological harm)

x. Narrowing of repertoire

*Withdrawal state is both a diagnosis and an element in a criterion for the dependencesyndrome

**Harmful use is both a diagnosis and a criterion for the dependence syndrome

and specific criteria in the nosologies, we may this terrain required the use of several method-expect their practical use to be influenced by ologies in a series of substudies.diagnostic concepts which are widely recognized Across the various methods and substudies,in lay as well as professional circles-such con- our general approach has been comparative andcepts as alcoholism, addiction, withdrawal and contrastive. The fundamental comparison of theabuse. Several of the component criteria for the study, of course, is between the nine culturesDependence Syndrome themselves also tap into and eight primary languages of the study. In eachwell-recognized diagnostic concepts. Besides site, data were collected systematically for al-withdrawal, these include increased tolerance, cohol and for one other drug class of interest,

.compulsion, impairment or loss of control and allowing contrasts of the application of conceptscraving. Associated with each diagnosis or diag- and diagnoses to alcohol and to the drug class.nostic criterion in the ICD-lO and DSM nosol- The data also allow for comparisons within each

.ogies is one or more characterizations or culture, such as comparisons of professional andsymptoms; in a more or less direct fashion, these lay terminology related to each diagnostic con-

.are translated into items or subitems in the CIDI cept. Thus respondents were asked for their ownand SCAN. ways of describing behaviour covered by such

The CAR study therefore set out to measure terms as intoxication, withdrawal, tolerance andthe cross-cultural applicability of terms and for- harmful use, as well as about their understanding

.mulations which fell at each of four distinguish- of the meanings of diagnostic terms-both thoseable conceptual levels: at the level of typifications in the nosologies and those in popular use, suchor characterizations of problems related to drink- as alcoholism and addiction (see Table 1).ing or drug use; at the level of diagnoses; at the In addition to our interest in differences in thelevel of diagnostic criteria; and at the level of meaning of terms and concepts cross-culturally,instrument items. As we shall describe, covering across drug classes and between experts, profes-

204 R. Room et al.

sionals and lay people, we were also particularly main national ethnicity and people living thereinterested in differences in the scope of application spoke the dominant national language. In otherof the terms and concepts. For example, while cases, the cultural and linguistic situation wasthe ideal-type definition of alcoholism given by more complex. In Bangalore the emphasis wasrespondents from different cultures might be on Kannada, the local language and ethnicity,quite similar, there could still be vast differences and similarly in Ibadan the emphasis was onin the threshold of problem severity at which Yoruba, but in both places some data were col-they would apply the term alcoholism to a par- lected in English, which is in widespread use asticular case. This question of the scope of appli- a lingua franca. In Flagstaff the emphasis was oncation is as significant as the meaning assigned to Navajo, an American Indian nation with its owna diagnosis in determining the diagnostic process language, but all data were collected in English,in a given culture. We thus asked respondents since English is known to nearly all and is the :themselves to compare and differentiate between usual language of therapeutic and official com-states which were and were not of diagnostic munication, while not all Navajos understandsignificance (see Table 1). One such set of con- spoken Navajo. The inclusion of both San- .trasts was between "normal" use of a particular tander, Spain and Mexico City allowed a com- ..

drug, abuse of the drug and harmful use. An- parison of two very diverse cultures sharing aother was between simple intoxication and in- common language.toxication which merited medical attention The sites included in the CAR study not only(assuming the latter to correspond to the Acute represented a diversity of language groups, butIntoxication diagnosis of ICD-10). Yet another also varied greatly in terms of the place of al-was between a hangover (or the equivalent for cohol in the culture. The position of alcohol in adrugs other than alcohol) and the withdrawal given culture is often discussed in terms of astate. In one of the sub studies (SARS), the rough dimension of greater or less "wetness"boundaries of application of dependence con- (Pittman, 1967; Room, 1989, 1992; Levine,cepts were also explored, with questions con- 1992). In the ideal type of a wet culture nearlycerning whether a hypothetical person with everybody drinks nearly every day; alcohol is aparticular symptoms or clusters of symptoms domesticated and indeed banalized part of dailyshould be considered addicted or alcoholic. life. Heavy drinking is thus an extension of social

drinking; the norm for the heavy drinker, indeed,is to keep drunken behaviour as much like sober

The study sites and the place of alcohol and behaviour as possible. As described by thedrugs in their cultures study's investigators and respondents, the studyNine centres from different cultures and repre- sites in Santander, Spain and Athens, Greecesenting different language groups participated in probably come as close as anywhere to embody-the CAR study: Ankara, Turkey; Athens, ing this "wet" type. Jebel, Romania would alsoGreece; Bangalore, India; Flagstaff, Arizona, approach this end of the continuum, but with theUSA; Ibadan, Nigeria; Jebel, Romania; Mexico harsher economic conditions enforcing less regu-City, Mexico; Santander, Spain; and Seoul, larity in drinking, and perhaps also with heavy .

South Korea. The sites were selected to assure a drinking seen somewhat more as "time out".wide range of diversity in language-groups of At the "dry" end of the continuum, as it isthe main language and in cultural patterns of commonly discussed, are cultures in which .drinking and drug use. The availability on site of drinking is set apart from daily life, on fiestas orexpert investigators with a command of English weekends, and in which there are many abstain- .and an ability to mount a substantial project ers. Drinking is "time out" behaviour, andwas also a practical consideration. As Table 2 drunkenness can serve as an explanation of badsummarizes, the nine sites include substantial behaviour (MacAndrew & Edgerton, 1969). In avariation in dominant religions. They are also at further extension of this, indeed, extreme drunk- .different levels of economic development and enness to the point of passing out can take on ageographically widely dispersed. positive value for some in the culture. Among

In each site, the main emphasis of the study our study sites extremely heavy drinking is well-was on the predominant local culture. In many established in the culture in Flagstaff, Arizonacases the predominant local culture was also the (see also Kunitz & Levy, 1994) and Seoul,~

Cross-cultural applicability of diagnoses

205

U

e ~

;.~

.'fi'

'8 ~

u :a'Q

U

; .~

~

0.=

'E

"S

~oS

~

"a ]."a u

~~

...;:§~o

0 t:

~..

'0 -=

.. d."

Po

0 000

~

0 0

O..u~

-

CI)

ZC

l)~

~U

U

~

>-;JQ

~

.,; '-"'"

=

0-t

?;-] ,.,

~-e

U

~

.~

=

tj '.

"'.. c.~

... .,...=

1;

-tOO

" "d:E

'C

'@ .~

e oS

.~

.~ u:e

~.';j

~to

°Po

0=

Po

" 0"'"

0 ~

CI)

ZC

l) ~

U

CI)

;JOC

l)~;J

-~.;:

."0'8 1

.0 -;

tJ" -; 0

; ~

~

;

~

.=

Q'R

t

'S

~.g

,,~'S

.g ..8 :E

...

u?;- 'O

.,d" =

"Q

~

~~

~

~

u.~

u ~

0 =

P

o ~

~

.,

0; Q

<1..1

~Q

U<

~U

C

I) <

""~

'" ~

."d

.~

; u

e.~

o.~

-;

£..'0

d -..;'

'i ~

=

5

0 =

.s 's.

~

ad.,e~~

e

d.,,c

.8 =

~

0 =

t:

0 ~

~

""

u,!!,

fo-~P

'~~

O

~

<...

-0 '0~

g §

~§

'iJ t'

'8 ~

~

c ~

=

.~

= .E

j 'e

1t""5 "0 .=

-~~

to

oS 'C

.,

,c ~

u _.~

.:E

\J

IJ ~

~

.fp] ~

"bb 'S

:§ ~

&

~

...: 8

,r 19

~

Z

U

>- ~

;J

U :I: Q

~

Z

:.-~

P

o.~~

...: >

-,g .C

I) bI)

U

;J O

.$?;- ~

g, It=

ti;

.d =

...; 8. .,

.,.~

-=~

.s

d .~

~

,g,

,~

~

d ~

0 .E

] t

'0

~

J'~

] "1

~.~

[

~,~

~

~

,2, §"S

"': £

~

~.(

U

~

Z :I: C

I) Z

.(

;J ~

~

:I: ;J Z

,8-

.a-N

.,

a-u

u--u,-.,,Q

=

dC

l)".";0

U~

.s ~

~

~

.s

; .~

~

-; ~

g ~

~

'8

~

= .E

j .a ]

.~":§

=:;:

~'"

3 g

.8 u oS

e d

'" .,.

~

"Q

.Q

'Q

-;;J ~

; g

g ,8

~

]] ~

~

'iJ

,8 ~

=

~

~

u

u 'Z

0

0~

~

CI)

:I:<U

~~

fo-fo-:I:~Z

""o:I:Zc

~

U

u .s

.~

~0

~

.'fi' ~ '8

-g-'0

~uo=

t:

5 ~

~

0

~

,~ ~

~

:§ 8.

~

U

~~

~

~

"OQ

l :

:.=

dU

?;- ,.,

0P

o,~

tj 'Q

j]

>-

e ~

u~

o=

Po

U

.~

." t: U

,'~

a

~

~

~

u:'-""U

S~

5~

~

~

~

'8~>

- IX

) .,

~

Ufo-

~

fo- :I:;JQ

~

0 g

~

U..-'"

"- oa-

.-"Cn'

g~

.d~=

,-.,

'-" bI) ~

.Q

'" 0

U

., .,

d ..,~

u":E

d

'E'

~

'Q .g

-Po §

~

-; "Cn'

0 =

=

0

U

dP

o'Q

.0 ~

P

o .8

"0 .d

~

0 0

~

=

6b bO

,~

.Q

0

.,~

'Q

~

U

d :=

d

d .U

.,

-U

d -=

00 ,.,u

'.,=

Ed

~_o.

~u

~

-~

..~

-~

,~

~

=

u:=

~

~

206 R. Room et al.

South Korea, even though only a substantial linguists, with the lead provided at most centresminority of the population engages in it. The by psychiatrists and their staff. All the partici-other four study sites all show a pattern where pants listed at the head of this article, and someabstention is common, and drinking is defined as others as well, were involved in the design of thepotentially disinhibiting. In Ankara, Turkey and study and in an intense phase of writing, testingBangalore, India, indeed, most adults are ab- and refining the new or revised instruments usedstainers, and in Bangalore any drinking at all in the study. Since the study's use of qualitativemay be problematized and seen by the drinker's and ethnographic methods in the context of afamily as causing bad behaviour. multi-national effort in psychiatric epidemiology

The per capita consumption figures shown in represented a new departure, considerable effortTable 1 confirm the places of Athens, Santander went into training in and demonstration of theand Jebel at the "wetter" end of the drinking methods. In addition to demonstrations as part .

spectrum, and the status of Ankara, Bangalore of the meetings of investigators two trainingand Mexico City as located in societies with courses were organized for investigators and in-much lower consumption. The relatively high terviewers, covering CIDI and SCAN adminis-per capita consumption in South Korea, how- tration as well as key informant interview andever, alerts us that a differentiation in terms of focus group techniques.the drama surrounding drinking is not only a The study consisted of five core components,matter of the level of consumption. The level of designed to complement each other with differ-drinking in South Korea has risen dramatically in ent methods, study populations and focal con-the last three decades, but the cultural patterning cerns. A strength of the study's design is itsof drinking, with an emphasis on ostensive drink:: diversity of data collection methods, which al-ing bouts, is far removed from the banalized lowed for some convergent validation of findingspattern of everyday drinking in a wine culture. from different substudies.The cultural variation to be found among the Due to the compressed timetable and limited"dryer" cultures in the CAR study material sug- resources of the study not all components weregests, in fact, that a single "wet/dry" continuum completed at each site, and there was also vari-does not adequately capture the dimensions of ation between sites, as we shall describe, in thecultural variation in the position of drinking. extent to which the full design of a component

Limits on resources meant that data collection was carried out. Overall, the completion of thewas limited to covering alcohol and one other greater part of the full design is a tribute to thedrug class at each site. The other drug or class of commitment and perseverance of the site investi-drugs covered at each site was chosen as having gators.the highest apparent prevalence of harmful usethere. In Ankara, Athens and Santander heroinwas chosen as the most significant drug, while Bilingual expert consultation and exploratory trans-cannabis was the choice in Bangalore, Flagstaff, lation/back-translationIbadan and Mexico City. In Jebel the choice was A bilingual expert group was formed at each site,sedative medications, and in Seoul am- consisting of individuals conversant both with .

phetamines. In most sites, use of the other drug the native language and with English who couldcovered by the study was seen as substantially be considered experts in the alcohol and drugmore culturally alien than drinking alcoholic field. The bilingual expert group conducted a -beverages. However, in Jebel the use of sedatives specific protocol of translation and back-transla-is somewhat normalized in the culture, while in tion of the CIDI and SCAN questionnaires, asBangalore drinking alcoholic beverages may be well as of the other questionnaires and materialsat least as marginalized as using cannabis. used in the CAR study.

The experts were selected on the basis of theirability to elicit information from monolingual '

informants and served as liaisons between inves-Study methods tigators, interviewers and other members of theThe study represented a multi-disciplinary en- study teams. For the translation study, a mono-deavour of psychiatrists, anthropologists, sociol- lingual group of people knowledgeable in theogists, epidemiologists, psychologists and alcohol and drug field was also formed, as a

Cross-cultural applicability of diagnoses 207

group which could register the comprehensibility study was that translating all the study instru-of the study materials (as translated and edited ments into Navajo would not be a sensibleby the bilingual group) for study subjects who choice, given that Navajo is primarily a spokendid not possess a knowledge of English. and not a written language, and given the bi-

The exploratory translation and back-transla- lingual patterning of daily life, with health andtion exercise was viewed as a substudy in its own social service interactions commonly conductedright, contributing to the understanding of the in English. However, a translation of the CIDIcross-cultural applicability of concepts and their was prepared in oral form on audiotape, andoperationalization. In addition to assessing the extensive lists of translations of terms intocomprehensibility of CIDI and SCAN questions Navajo were made and drawn on for the findings

.in the particular culture, the exercise explored of the study. The translation/back-translationcultural obstacles related to concepts of sub- study was completed in each of the other eightstance abuse which should be taken into account sites.in the instruments. The translation process was

.also seen as a source of data on cultural and.linguistic meanings of terms and concepts and Key informant interviews

was a necessary step in undertaking other com- The major objective of the key informant sub-ponents of the CAR study. study was to elicit information about how differ-

The steps followed in the translation and ent types of people living in each society thoughtback-translation of instruments and interview about the main concepts which are used in theschedules were as follows: (i) the translation diagnosis of alcohol and drug disorders. Infor-from English into the target language was pre- mants were encouraged to answer as spokespeo-pared by one or more translators who were not pIe for the culture as a whole, rather than as anpsychiatrists or other mental health profession- individual. Investigators were trained to useals; (ii) the translated version was discussed by semi-structured interviewing techniques, follow-the bilingual expert group to identify areas of ing an interview schedule which elicited largelylikely difficulty and to agree on issues to be open-ended answers. The interview was conduc-explored in the monolingual groups; (iii) the ted in a conversational style, starting with open-members of the bilingual group discussed the ended, general questions about what words ortranslated version with the monolingual group; phrases the informant might use to describe a(iv) after reviewing the translation in question, particular behaviour or state related to drinkingthe monolingual group leaders reported back to or drug use. Subsequently, more specific ques-the bilingual group regarding whether the ques- tions asked about the meaning of particular con-tions were comprehensible, whether there were ceptual or diagnostic terms, and how people incultural obstacles to responding to the questions, the culture might distinguish them from otherand whether there were other important prob- terms. The schedule included optional probes tolems related to specific symptoms which should follow up answers to the main questions. De-

.be added to questions in CIDI and/or SCAN; (v) tailed information was provided to the investiga-the bilingual expert group discussed these re- tors and interviewers regarding the intent of eachports and decided whether there were major question and guidelines to follow in asking ques-problems which would require a second revision tions.

.of the translation to be passed in the same way Each centre was asked to interview at least 20through the monolingual group; (vi) the transla- informants concerning concepts and terms used

.tion was back-translated by another (non-pro- for alcohol, and 20 informants concerning con-fessional) individual; and (vii) the bilingual cepts and diagnoses used for the other drug typegroup prepared a summary report specifying par- studied at that centre. The aim was to tap bothticular areas which needed in-depth exploration lay and professional constituencies with substan-

.and addressing in the final report of the study. tial knowledge concerning drinking or drug use,Typically, the bilingual expert groups discussed but reaching well beyond the narrow and oftenthe results of each component of the study as it cosmopolitan circles of specialist expertise. Itwas completed and made summary remarks for was suggested that each centre include withinthe final report. the 20 informants for each drug three health

In the case of Flagstaff, an early finding of the workers working in the area of alcohol or drug~

l

208 R. Room et al.

problems, seven other health or social service members and to allow free expression of theworkers in regular contact with alcohol or drug opinions of group members. In the course ofproblems, five heavy users of alcohol or the drug focus groups, the interviewers/researchers hadand five family members of heavy users. In most several tasks: to moderate, to listen, to observecases, the heavy users and families were drawn and to analyse discussions. There was no press-from clinical cases and their families. In using ure on the moderator to have the group reachthe relatively limited resources of the study the consensus. These group discussions were meantchoice was thus not to expend resources on to provide more information on how alcohol andinterviewing members of the general population drug problems are seen in each culture, how thewho might have only a distant and hazy knowl- questions about these problems can be formu-edge concerning alcohol or drug matters, but lated in a culturally appropriate way, and howrather to elicit information from knowledgeable well diagnostic criteria and concepts of substanceconstituencies, lay and professional. use disorders apply in each culture.

With minor variations, each centre collected Focus groups were organized as discussionsthe key informant data in accordance with the with seven to 10 participants. Study sites weresuggested methods and sampling. Besides the asked to convene separate focus groups for eachwritten record of the interview, most or all in- of five categories of participants: (a) psychiatristsformant interviews were tape-recorded in all cen- and other professionals with alcohol or drugtres other than Jebel. Participating centres were diagnostic experience; (b) other health profes-able to put much effort into carrying out the sionals such as family physicians, social workerssub study well, since it was the first major data or nurses; (c) other community workers whocollection effort of the study, and one which was dealt with alcohol and drug problems, such ascharacterized by the Jebel team, for instance, as policemen, judges or priests, (d) members of the"the real core of the research". culture who used alcohol or drugs heavily, and

(e) family members of heavy alcohol or drugusers.

Focus groups All sites convened at least two focus groupsAs the CAR study was designed, the focus group but the limited resources, the backbreaking CARsub study was intended to allow a more extended fieldwork schedule, and in some cases thediscussion among members of different social difficulty of finding members of a constituency,constituencies of concepts and terms identified limited the extent of the focus group substudy.as problematic in the translation and the key As for the key informant interviews, the study byinformant substudies. Group discussion of con- design concentrated data-collection efforts onceptual meanings and differences, it was felt, relatively knowledgeable constituencies, lay andwould help clarify ways of thinking in the culture professional. All sites conducted a focus groupand perhaps also areas where there was no clear with heavy alcohol users (usually clinical or re-cultural consensus. covering cases; including some drug users in

Each focus group addressed a relatively small Flagstaff); most sites conducted a separate focusnumber of general questions: group with heavy drug users (not Jebel or. h t . a1 d b 1 f th b Flagstaff); most sites conducted one or more

w a IS norm an a norma use 0 e su -..,t ( 1 h 1 dru ) groups WIth psychiatrists and other health pro-

s ance aco 0 or g; ... h t th ' f th . d .fesslonals (not Mexico CIty or Seoul); and Ban- .w a are e meanmgs 0 e vanous lagnos-tic terms related to the concept of alcohoVdrug galore, F~agstaff~ Ibadan and Jebel conductedd d groups WIth faInlly members of alcohol or drug

epen ence; .h th .' 1 .. d d '&f' b clients. .w at are e slml annes an 1J.1erences e-

tween alcohol and drug abuse and addiction; S I'l: d " d R ' S h d Ie i'S' ARS~\ and e~-a m~nlStere aung c e u \. n Y

, .., .The SARS substudy was primarily focused on.whIch preventIon and mtervennon strategies ,.. .

t l.k 1 t b ffi n.' t 1 h 1 obtammg VIews from treatment proVIders con-are mos ley 0 e e ec ve agams a co 0 -.' ..d 1 t d bl . th It ~ cernmg the cultural applicabIlity of particularor rug-re a e pro ems m e cu ure. CIDI d SCAN . d th .. an Items an elf appropnate-

The focus group interview technique used ness as indicators of diagnostic criteria. Theaimed to facilitate interaction among group questions asked in it were pitched at a lower level

Cross-cultural applicability of diagnoses 209

I of generality from the preceding substudies, study was designed to focus on the actual formu-namely that of the individual SCAN or CIDI lation and wording of alcohol- and drug-relateditem. The precoded SARS questionnaires aimed questions in the diagnostic instruments, as they(i) to elicit data in a systematic way on the extent were administered to individuals known to haveto which various symptom-items or question- alcohol and/or alcohol problems ("referenceitems represented diagnostic criteria in a particu- cases"). As the study was designed, subjects werelar culture and their cultural appropriateness; (ii) to be first interviewed with the alcohol and drugto assess which item or subset of items repre- sections of CIDI or SCAN, and then questionedsented the minimum set to represent or cover in a semi-structured exploratory interview aboutadequately and completely a diagnostic criterion; the meaning of, and concrete details behind,

.(iii) to explore the cultural appropriateness of their answers to CIDI and SCAN questions. Theindicators of harmful use, abuse and dependence draft schedule also included alternative formula-in different cultures; and (iv) to obtain sugges- tions of items and follow-up questions to asktions on alternative or additional symptoms or respondents about their views and understanding

.criteria that are needed to better represent con- of specific items (e.g. "You mentioned that you.cepts of alcohol and drug abuse and dependence found it difficult to stop drinking before youacross the cultures. became completely intoxicated. Please describe

Five separate questionnaires were used in the to me what it means to you to be completelySARS substudy. The SARS Cultural Appropri- intoxicated. What are the signs of complete intoxi-ateness (SARS-CA) questionnaires, one for al- cation?").cohol and one for drugs, were intended to be In the design of the reference cases componentfilled out by social and health workers dealing a minimum of 24 cases per centre were to bewith alcohol and drug-related problems, but not assessed, including substance users with mildnecessarily schooled in diagnostic concepts and problems and those with severe problems (buttechniques, while the SARS Diagnostic (SARS- without serious cognitive impairments). Half ofDX) questionnaires for alcohol and for drugs the cases were to be patients who had receivedwere intended to be filled out by health profes- treatment for alcohol- or drug-related problems;sionals familiar with psychiatric diagnostic sys- the other half had not received any treatment. Intems. A Slang/Street Name Drug Supplement the selection of subjects, both alcohol and drug(SARS-SDS) questionnaire was also developed users were to be included, drawn from varyingfor the identification of culturally appropriate socio-demographic groups.terminology for various types of drugs at each The timetable tied to the study's fundingsite. meant that work on this component had to be

Fieldwork on the SARS substudy was carried considerably curtailed; it was decided in theout on an expedited timetable, since the sub- course of the study that the component would bestudy was a late addition to the CAR study, and optional. Four of the sites collected data fromin several sites required extensive translation. In 21-24 cases (Bangalore, Ibadan, Jebel and Mex-

.the event, both the CA and DX versions were ico City), while Santander collected 10 cases,primarily administered to expert staff close at Athens 15 and Ankara 32. The semi-structuredhand to the investigators; overall, 66% of those follow-up was dropped and in almost all casesfilling out the DX and 46% of those filling out data collection was limited to asking the CIDI

.the CA were psychiatrists, and over half of those alcohol and drug schedules, and comparing thefilling out each form had jobs which included results in a case conference with expert diag-

.research work in the alcohol field. At least 15 noses.responses to each of the four forms were gath-ered in each of seven sites: Ankara (n = 106 CAor DX forms completed), Athens (n = 66), Ban- Study limitations

.galore (n = 154), Ibadan (n = 60), Jebel (n = 61), Among the limitations inherent to the CAR proj-Mexico City (n = 67) and Santander (n = 60). ect's research design, four are of particular rel-

evance to this report. First, as we have noted,there is some unevenness in the data collected,

Reference cases due to limited resources and the inherentThe reference cases component of the CAR difficulty of collecting data simultaneously in

210 R. Room et al.

nine geographically dispersed sites, and this lim- mostly in another language. To get these state-its the scope of the cross-cultural analysis. Sec- ments and thoughts into a cross-cultural com-ondly, cross-cultural analysis is also limited by parative frame, they have had to pass throughthe fact that the second drug class, besides al- translation-a methodological limitation in-cohol, considered in the project varied by study herent in comparative research across languages.site. This is in part simply a reflection of reality: In the second place, these summaries have beencommonly used drugs vary considerably by cul- prepared and organized by very special culturaltural setting. It also reflected that in-depth infor- representatives-in most sites, bilingual researchmation could only manageably be collected on psychiatrists. On the one hand, the special exper-one other drug class besides alcohol. Conse- tise and cultural position of the site investigatorsquently, while we draw on the drug findings in has been essential to the study; but on the otherthe present report, we focus on conceptions of hand, it is possible that their familiarity withalcohol-related problems, since it is for these that international concepts and terms has pulled thethe most comprehensive cross-cultural data are study material closer to those concepts and termsavailable. than would otherwise have been the case.

Thirdly, the scope of the study inevitably lim-ited the depth of data collected on particulartopics in particular cultures. A whole study, for Selected results from cross-cultural analysisexample, might well be focused on the meaning- A full discussion of the results of the study isfulness and boundaries between terms for hang- presented elsewhere (L. Bennett et al., Use andover and withdrawal in Korean, and an Abuse of Alcohol and Drugs in Different Cultures: Aethnographer or other researcher doing such a Nine-Country Study, in preparation). Thestudy would undoubtedly be able to collect findings described here serve as an illustration ofdeeper and richer data than could be done in the the wealth of data collected and of some of thepresent study (for an example of such a study, possible ways of using it.see Taipale, 1979). On the other hand, thecross-cultural design of the CAR study allowedan explicit comparative framing which is unusual At the level of instrument itemsin detailed studies, and indeed in the literature. The findings of the CAR study generally supportThe present study offers a rich lode of tentative the feasibility of translating instruments such asobservations and conclusions to be tested and CIDI and SCAN into a diversity of languages inrefined in more detailed studies. very different cultural circumstances and getting

Fourthly, the CAR project relied on the exten- back meaningful and usable answers. At thesive cooperation of treatment professionals, and most concrete level of analysis, then, the overallpatticularly of psychiatrists, who served both as findings of the study are encouraging of cross-investigators and as respondents in various facets cultural work on diagnostic instruments in theof the research. Many pf these professionals re- alcohol and drug arena.ceived training in the United States or British The findings also suggest substantial cautionpsychiatric and nosological traditions, and this both in the translation and interview proceduresmay have influenced their perception and de- of such studies and in the interpretation of theirscription of indigenous concepts and terminol- results. At the item level, some of the difficultiesogy. Furthermore the respondents to the study, have obvious solutions: for instance, to ask aboutlay and professional, were by design more knowl- amount of drinking in terms of customary bever-edgeable about alcohol or drugs than the average ages and container sizes, and to substitute localperson in their culture. Concepts and perspec- modes of transportation as appropriate fortives may thus be more sharply defined in the "bicycle, car or boat". However, not all of thestudy results than they would be in general dis- problems can be solved so simply. The meaningcourse in the culture. or underlying assumptions built into a CIDI

It should be recognized that the results re- item may be quite foreign to the language orported here are derived from interpretation, in culture.two different senses of the term. In the first An overarching finding, in fact, is that theplace, we are relying on summaries in English of diagnostic criteria and their operationalizationswhat informants and respondents had to say, assume a self-consciousness about feelings,

Cross-cultural applicability of diagnoses 211

knowledge and consciousness which is foreign to translation can drop the "find" and the "feel",the folk traditions of some cultures. Here are for instance, from such constructions. However,some examples of problems of this sort in trans- this leaves open the question of whether the itemlating the study materials for use in different really offers a comparable stimulus without thesites: softening effect of such phrases.

I Kann d " £ 1 . 11 " 1 d While discrepancies and difficulties in the cul-.n a a, ee emotlona y was trans ate "

1 "' th th 1 . th tural translation of indiVIdual Items can some-to menta state, WI e exp anatlon at , , ,th ' 1 d £ " K times be resolved through rewording, In some

ere was no simp e wor or emotion In an-, "d th 1 . 1 bl d cases our analysIs suggested that the problems In

na a; e on y aval a e wor was seen as, "" b ki h " th " " th translation were symptomatic of deeper prob-

00 s , at IS, not In e common ver-. "1 1 " F 1" ft drinkin ' lems WIth how well particular concepts travelled

.nacu ar anguage. ee a er g was "1 d "th d h " h 1 across cultural boundanes"

trans ate WI a wor w IC a so means ex-perience. Positive and negative feelings be-came good/pleasant and bad/unpleasant

.experiences".."Anxiety" is a new term in Romanian and not A h le 1 if ' "

'I d "" , h d fr th " f £ t t e ve 0 cntenaeasl y Isnnguls e om e sensation 0 ear. A fin 11 d d th ICD 10 d d"" say a opte e -epen ence syn-

.In Korea, "many people Carlnot distInguish .." "£ I " fr thi ki " Th d th drome has SIX cntena, three of which must haveee Ings om n ng. e wor at con- b xh"b " d . d k th d "

" " " "een elite or expenence to ma e e lag-notes feeling IS rarely used In conversation, . T f th '" "

thd 1d " £ bl d £ h " 1 nosls. wo 0 e cntena-WI rawa or use to

an IS more com orta y use or p yslca '" ."" (d £ 1 tho " " k"') th £ relIeve WIthdrawal, and using despite harmful

sensations 0 you ee IS pln-pnc r an or "" " -consequences-relate to phenomena, WIthdrawal

affective states (how do you feel about thiS.). d h ful h " h ICD 10 d " ,K " h d "' d ' an arm use, w IC are -lagnoses In

orean as many a Jecttves an expressions, th .. gh d 'II b d " d "th, " elr own n t an WI e Iscusse In e next

but they are not well differentiated In terms of" '" " ." th gh d "" section. Other cntena cover craVIng or compul-

emotions, ou ts an sensations. , '"I Y b th " "" d "ffi 1 slon; loss or lInpalrment of control; tolerance;

.n oru a e term emotion was I cu t to "', and neglect of pleasures or Interests or Increasedconvey. time spent seeking or using (see Table 1). In the

The CIDI items often also have built-in attrl- text of ICD-10 concerning the dependence syn-butional, causal and other relational assumptions drome in the "clinical descriptions and diagnos-which are not customary in some languages and tic guidelines" (WHO, 1992a), although not incultures. Such language as "trouble because of the diagnostic criteria for research (WHO,drinking", "after you had realized it had caused 1993), two other aspects which had originallyyou"..", "where it increased your chances of been proposed as criteria are mentioned: rapidgetting hurt" presume both self-consciousness reinstatement, and narrowing of the repertoire ofand a style of causal attribution which is un- use. Rapid reinstatement had already beenrecognizable in some cultures. Some items also dropped from the ICD-10 list of criteria by the

.build in assumptions about intentionality which inception of the CAR study, but narrowing ofdo not travel well; thus it was reported that in the repertoire was included in the study" TheAnkara respondents found vague and inscrutable findings on the responses of key informants at

.the idea of intention at the beginning of a drink- each site concerning the applicability of the al-ing occasion, as in drinking "for a longer period cohol dependence syndrome criteria are sumrna-

.of time than you intended to". rized in Table 3"Some of the language in the items which cause In a majority of cultural situations, narrowing

difficulty is not part of the main meaning of the of the repertoire was felt to be neither clearlyitem, but instead is derived from a well-estab- definable nor an appropriate indicator of alcohol

" lished English-language style of question-item dependence. In some cases this was because theconstruction which uses introductory phrases traditional cultural repertoire of patterns of usesuch as "did you ever find" or "did you ever feel" was narrow for all: in Santander, for instance,as a way of softening items which might sound "the consumption pattern is socially established,accusatory and making them more colloquial" and heavy consumers tend to adopt the sociallyThese difficulties are amenable to solution. The accepted pattern", and in Bangalore, the concept

212 R

. R

oom

et at.

..0z

u u

u =

-°;j

~

°;j °;j

0

'" ~

'"

'" ~

"3 041

0 0

41

0..~

"..~

..~

6

0=

0 0=

0=

0

c>J

z:a z

z:a z:a

5.du

u u

CI.

°;j °;j

°;j'" '"

'" ~

0 0

O";;J

t ~

~

og

~...

"Q

~=

=

'" =

0

B

5:a:a 0

:a co

c ~

..oS

Od

000 0=

o;a

(/) Z

Z

Z

z:a

z ~

.8

E~

0 e

e e

~.B

.

.~

~

~

~

§ "QO

J 't;o ~

~

~u

Z

Z

Z

]~s

co co

~~

..N""

C~

4IC

'-' '-'

'S

0 ""

] '"

'" ~

~

=

8~

~

~

~

~

~

41 41

CI.";;J

i0;:

0;: ~

~

6~

6~

"Q

.g41

41 41

~~

4I ~

~4I ~

O~

.D?

>

00 ~

~

=

~

0 0

Z

Z

~o~

. ~

'"C

I.=

u co;:

co co

eco

co .;j

41 41

-C

'" 41>

~

~

08 ~

o~

~

~

.a

~

.~

~

0 ~

0

a "Q

~

~~

~~

~

~=

~~

~~

..~

4Ii:i.

~

~

je:aje:az je:aje:a

z..a ~

~~

c~

O§co

co ~

oC

C

41

.~

0 .S

os

t3 8

6 11=

=

=

~

5'=

>.s

41 41

"O"Q

'" 6

~

6 ~

..C

~

co ~

~~

4I 541

~

~

j:aj~

§&...u

u U

U

"Q

"Q~

.;j

.;j °;j

°;j 0

"' '"

'" '"

coo,0

41 ~

g

~

g g

g o~

-§

rS

"e

'" co",

ee

co bO

bO

U

~

0 C

I. =

CI.

=

=

=

=-0;

~

-;; =

o~

=

'-"'

'-"':a ~

~

~

~o;:

"Qo;:

~

~

"0"0 ..~

~

~

041 041

0 0

0 0

0 ~

""

~

~

z 0

z 0

z z

z z

z jog

.". .~

'"

~

'" 0

..~

C

a co

.~

bO~

08

41:6

Q

~

og og

e

..; 5

§ §

§e~

i ~

~:?fi

.s.g 0

0 o~

~o=

O

="Q

'Q,Q

-<

z

z z

j :a

z :a

z :a

8 o~

to co~

Eo-

041

'" '"

.g c:

6

=

~

~

e ~

=

~!=

0;:

0;: I)

c .D

"Q~

41

41 bO

-41

~

>

? o

0=

v-<

0

c 0

~z

z:a o~

~

.

0; ~

~=

U

41

0"~

001

~.~

U

..

=

~

~

0 ~

o

~

°;j C

I. ~

c:

c:~

't;o

8.g ~

4I.d&

08

~

°u 41

o~

...~

-0; ~

5

c:

"0:18 =

41 ~

~

~

~.u...

641

.8 o~

8

~

~

;;] 41

:=

~

j ~

g

§5

~

001 "Q

] ~

~

U

0

'" co ,g

.c: ~

5

.5 co"3

.~

c: -a

"0:1 ~

~

~

.s ~

~

C

I..a,

~

§ S

' C

I. 8 :E

~

~

~

~

°e

~

~

8 .u

6 (/)8

.§8 ~

~E

o- Z

"E.

p.,oZ

~~

r,"-" "-""-"

0 C

""", N

'"

~

In \0

~Z

bO'-'

'-' '-'

'-' '-'

'-' '-'

=:a

Cross-cultural applicability of diagnoses 213

is "rather alien.. ,insofar as there is not a wide ing were recognized in most of the sites, but therepertoire to begin with", meaning was often not differentiated from loss or

All sites recognized phenomena related to tol- impairment of control or from dependence moreerance, though only in a minority of the sites was generally. In Bangalore, for instance, the conceptthere a specific term for the concept in ordinary of craving was recognized, but it was equateduse. The idea of tolerance for alcohol often elic- with loss of control. In some sites, the usualited responses indicating that this state was not terminology was closer to "having a desire orviewed in the culture as a symptom of pathology, urge", and it took some explanation to convey toor as associated with problematic drinking; of- respondents the connotation of an irresistibleten, indeed, it had positive associations, In a urge, In Cantabria, Spain, the best equivalentwetter cultural context such as Santander, the term was a local word unlisted in Spanish dic-

, connotation was of an ability to drink without tionaries; the term was unknown in Mexico City,behavioural change, In a context such as Seoul, Impaired control of drinking was a recognizablewhere behavioural change with drinking was ex- concept everywhere, though its meaning varied

.pected, the connotation was of endurance or between cultural settings, In Athens it referred toimmunity; here it was declining rather than in- incompetent drinking, the drinker's failure tocreasing tolerance which was seen as a marker keep behaviour appropriate to the context andfor alcoholism. The terms used in Ankara con- circumstances, while in Bangalore the emphasisnote being able to hold one's liquor as a signal of was on a generalized craving taking over themanhood, while in Bangalore and Jebel the term drinker's whole life. In some places (Athens,"tolerance" was frequently assigned such mean- Jebel and Mexico City) impaired control wasings as being indulgent, accustomed to or at- seen as the characteristic which distinguishedtracted to, alcoholics from others. In the Key Informant

In several sites tolerance was understood in its substudy respondents were asked whether theytechnical meaning for the drug type included in thought the amount people drink was controlledthe study but not for alcohol. This was an exam- by availability or by people's self-control. InpIe of a wider phenomenon encountered in the Ibadan and Bangalore most respondents saw thestudy, where the relative novelty of drug use in control as external-that drinkers had no in-the culture had brought with it technical and herent self-control over their consumption. Inloan-words, while the understanding of alcohol Seoul, losing control was not seen as problematicdisorders remained organized around older con- or unusual but rather as the purpose of drinking.cepts and traditional language. For instance, in Along with respondents in Flagstaff, where theJebel both professional and lay people use pri- questions tapped into the Navajo cultural valuemarily pejorative words describing drinking be- on personal autonomy, Seoul respondentshaviour and social reactions to drinking, words showed the strongest commitment to the ideasuch as etilic (heavy ethanol user), betiv (drunk- that control of drinking was always a matter ofard) and alcoolic (alcoholic), rather than pro- personal choice,fessional terms, But: Neglect of alternative pleasures or increased time

seeking or using proved to be a difficult criterion, th 1 d ' th 1 h 1 to understand in some cultural situations. Inm contrast to e anguage use mea co 0 '"fi Id R ' h d d th d several sItes, mcludmg Santander and Ibadan,

e, omanlans ave a opte e mo em ,',. 1 f fi ' 1 h d ' b ' th alcohol was so easIly available that tIme spentanguage 0 pro esslona s w en escn mg e, ,

" procunng made no sense to Informants as anuse of drugs such as benzodlazepmes or hyp- ' d ' f yth' In B 1 '

' D d 1 1 fr 1 m lcator 0 an mg. anga ore, tIme was notnotIcs. rug users an ay peop e equent y , ,

d h ' , , d d VIewed as a scarce or expendable commodIty, souse wor s suc as mtoXlCatIOn, epen ence, th ' d d d '

nkin d, at tIme evote to n g or rug use wasabuse, tolerance, etc., the very words whIch bl if th 'th ' fi th 1 h 1 fi Id not seen as a pro em, except e tIme was

ey reject or e a co 0 e , "" subtracted from work tIme, The notIon of alter-nate pleasures also caused trouble: in Romania,

In all sites except Bangalore, there was a ten- it was remarked: "almost all pleasures are relateddency to see drug use in more clinical terms than to alcohol consumption", Giving up social oralcohol use. recreational activities for drinking was not seen

Concepts and terms around compulsion or crav- as much of a problem in Ibadan, since drinking~

214 R. Room et al,

Table 4. Signs of intoxication, when help is needed, and when medical attention is needed

Signs of intoxication mentioned

Uncooordinated Aggressive Quiet, sad When medical attention seen as necessary

Santander (:f) + + Loss of consciousnessAthens (:f) (3:) + Unconscious, nonstop vomiting, very serious

aggressivenessJebel (3:) (3:) Coma, seizures, fitslbadan (:f) + Sustained injuries, endless vomiting,

unconsciousAnkara (3:) BlackoutMexico (3:) (:f) Disturbance of consciousness, violence,

physical illnessBangalore (:f) + Head injury, unconsciousSeoul + Coma, almost dead, seizures, severe illnessFlagstaff (:f) * + Bleeding, DTs, seizures

+ Mentioned as a sign(:f) Mentioned as a sign, and outside help is seen as necessary*"Outside help must be requested"

"could in itself be the social or recreational ac- pleasures given up are all quite widely seen astivity of the individual". The criterion was more questionable criteria, particularly when appliedcommonly recognized as a problem with respect to a culturally entrenched drug such as alcohol.to drug use.

In responses to the SARS self-rating substudy,elements of three criteria stood out in all sites asamong the lowest-rated in terms of their At the level of concepts and diagnoses"cultural appropriateness" (how understandable As noted above, we took the distinction betweenand how acceptable they were) as "alcohol-re- the acute intoxication diagnosis of ICD-I0 andlated symptom items": narrowing of the reper- other intoxication as being that the former re-toire, time spent obtaining and tolerance when quired medical attention, In line with this, re-defined as the ability to function at doses which spondents in the Key Informant study werewould impair the casual user (Dawson, Grant & asked about signs of drunkenness, about whatTowle, 1993). While withdrawal, drinking to signs would indicate that the drinker is in need ofrelieve withdrawal and continued use despite help from others, and about what signs wouldsocial problems ranked highest overall across the indicate a need for medical attention, In severalseven participating sites in their "cultural appro- sites there was no exact equivalent of intoxi-priateness" as items, there was substantial vari- cation, as a semi-technical term in English. Theation between sites in their ranking. equivalent term in Spanish would indicate poi-

These results partially converge with the re- soning, while the most common term in Greek,sults from the qualitative substudies, but the methi, has positive connotations, being also usedmessage we are getting from the key informant as a metaphor to indicate extreme happiness, .interviews is probably mixed. In part we are Table 4 shows the signs of alcohol intoxicationbeing informed about the recognizability of the mentioned in each of the sites, with the sitesconcept and terminology in a particular cultural arranged in rough order of descendingframe. Also included in the message, we may "wetness". Respondents at most sites mentionsuspect, is a judgement about how much the lack of coordination, and this is the primarybehaviour or state referred to is or should be indication that help from others will be needed.considered a problem in a particular cultural Aggression is also commonly mentioned as a signsituation-both as a problem in itself, and as a of intoxication; only in Santander and Athens ispotential indicator of dependence/alcoholism. being quiet or sad mentioned as a sign. DespiteFrom these mixed perspectives, narrowing of the the major differences between the sites in drink-drinking repertoire, tolerance and time spent or ing culture there is strong agreement on when

Cross-cultural applicability of diagnoses 215

medical attention is seen as necessary: basically, family problems as well as physical and psychicthis is when the drinker has lost consciousness or health problems; the intention in ICD-I0 tois suffering seizures. "Ordinary" drunkenness, confine the terms to harm to health will clearlyshon of serious central nervous system (CNS) be hard to realize. For illicit drugs many sitesdisturbance, is thus clearly defined as not requir- reponed that lay respondents made no distinc-ing medical attention. tion between use and abuse or harmful use.

The existence of withdrawal symptoms was Many Navajo respondents also denied that thererecognized everywhere, although in most sites was a normal drinking pattern for Navajo people,there was no single commonly understood term distinguishable from harmful use. In Bangalore,for the phenomena. In Romanian, for instance, too, some respondents maintained that there was

.professionals used the term sevraj (weaning), but no such thing as normal use of alcohol-that all, this was not understood by lay people, who used use is harmful and will lead to addiction (see also

words meaning to renounce, abstain or stop Bennett et ai., 1993).oneself. In Ankara, withdrawal was sometimes Terms for alcoholism and addiction were well

.confused with craving: "these two states cannot recognized at all sites, but dependence or its, be differentiated easily and clearly. Alcoholic pa- equivalent was a new term to many respondents.

tients and some family members often used the Respondents in Seoul equated the term withterm 'crisis' instead of either of these terms". intoxication. In Romanian, dependence carried a

In several of the study sites there was no clear main meaning of subjugation or subordination.distinction for alcohol between the withdrawal In Athens and elsewhere no clear distinction wassyndrome and a hangover (see Table 5). In made between dependence, on the one hand,Santander, for instance, one-third of the key and addiction or alcoholism, on the other. Theinformants mentioned the word for hangover main features described for alcoholism variedwhen trying to describe the concept of with- quite widely by site. Respondents at thedrawal. In Athens, "the transition from hangover "wettest" sites did not mention amount or pat-to withdrawal was defined as the point when the tern of drinking as a sign of alcoholism, whileperson stans drinking in order to cover up the characterizations in terms of loss of control orunpleasant experiences of hangover". illness, as well as drinking behaviour, were com-

Mexican respondents distinguished between a mon at sites where drinking is more problema-moral and a physical hangover, and in several tized.other sites, too, regret and guilt were described There are clearly substantial variations fromas pan of a hangover. Feelings of guilt were one site to another in the threshold for identify-prominent among the descriptions of a hangover ing and defining dependence or addiction,in Athens, but were not included in the descrip- whether the definition is in global terms or intions of withdrawal. In Ibadan, "while most re- terms of qualifying with a cenain number ofspondents emphasized regret as an imponant dependence criteria. For alcohol dependence aneffect of hangover,...only one key informant illuminating extreme can be found in Bangalore,

.mentioned regret as a hallmark of withdrawal". a cultural situation in which only a minority ofThe Ibadan investigators suggest that the re- men drink at all and almost no women drink.spondents may have felt that those who had Table 6 summarizes the responses to CIDI ques-progressed to withdrawal are beyond feeling re- tions of four Bangalore non-clinical "reference

.morse, but it is possible too that withdrawal, as cases", none of whom met a local clinician'sa concept more within the medical and technical standards for a dependence diagnosis, but each

, sphere, attracts fewer moralized connotations. of which met three or more dependence criteria

Study sites with marijuana as the other drug from their answers to CIDI questions. A Banga-studied reponed either no withdrawal syndrome lore drinker who consumed the equivalent of twofrom cannabis or a much less clearly defined European bottles of beer (a total of 700 ml) oncecluster of symptoms than for alcohol. There was, every 2 months, and had never drunk more thanif anything, even less distinction between hang- this, nevenheless qualified for three criteria ofover and withdrawal. dependence: he reponed that drinking has less

In lay usage there was little distinction at any effect on him than it used to; his family andsite between harmful use and abuse. Harmful use friends objected to his drinking, but he contin-was understood to include social, economic and ued to drink, and when he had had tuberculosis

216 R

. Room

et al.

~ "

~~

+

~~

bQ

"~-e

++

++

+

; u

~oS

..,"~

0

't;-~

0

"~

8 "~

+

+ +

+ +

+ +

+E

~

~

§ .

oS"-

~~

m

OJ

~

,,0;~

,,\)

~

oS "~

+

+ +

+ +

+

+ +

"il O

.d"

Po

$:"t:"~~oS

~

] 5

++

++

++

+~

~

1~~

l1~§

oS "e

+ +

+

+]

o~~

=

t;i' "

~"

..r$:

>1

~

~

"~

d

8'1=

,. =

..r~

++

+

+

++

~.d

,,~~

613'"

~

0 "

t'5 '"

~.~

on "-m

.! 0;

,Q

"~~

~

+

++

++

++

++

f

,,"" ~

,," "

> ~

.:r;.~

.i~

~"~

11+

11++

+1+

:x= ..g "U

.§~

:a

" "

~"

" =

]'" d=

8~

.s;1 5 "O

J ~

12"\< bQ

"3 ~~

oS.c=

~~

~~

.!m

<~

;9<c.,.~

m~

Cross-cultural applicability of diagnoses 217

Table 6. Drinking and dependence in Bangalore, India: Jour nonclinica! cases meeting Jonna! criteria Jor dependence

Drinkingdespitephysicalharm

(bicycle DifficultyDrinking injury, TB, stopping,

Occupation age, and drinking despite social stomach cutting Irresistiblepattern of respondent Tolerance consequences disease) down desire

Police constable, age 40 years + + + +700 ml beer (two bottles)

.once in 2 months, Attendant at hospital, age 36 years + + + +

1000 ml beer (three bottles)1-3 days a month

Staff nurse, age 33 years + + +" 700-1000 ml beer (2-3 bottles)

1-3 times a monthFarm worker, age 55 years + + + +

90 ml arrack (30 gm alcohol= three drinks)once or tWice a week

his doctor had advised him to stop (he had concepts and instruments developed in a par-indeed stopped for a few months but then ticular cultural frame for use in other cultures.started again); and he had wanted to stop or cut Most centrally, they underline that the fact thatdown drinking but could not. Other respondents apparently meaningful answers to items can bewho drank only a little more than this qualified elicited with instruments that are a simple trans-for a dependence diagnosis. For instance, a re- lation-and that respondents will answer thespondent who accompanied his wife when she same way the next time-does not necessarilycame to the hospital for treatment reported the mean that the instruments are yielding valid orfollowing pattern of positive items: drinking the useful diagnoses.equivalent of three bottles of beer 1-3 days a The CAR study started from an already exist-month, and never drinking more; objections ing structure of diagnostic interview items, diag-from friends, doctor, or clergy; trouble driving nostic criteria and diagnoses, and studied theirbecause of drinking; the same amount of alcohol applicability in nine disparate societies. Thehaving less effect than before; and having had study's methodologies and time-frame allowedsuch a strong desire to urge to drink that he us to acquire a broad picture in each society andcould not resist it. Clearly, in a cultural situation to set it in a cross-cultural comparative frame,where there is much disapproval of drinking the but it did not allow us to match the level of detail

.threshold for positive responses to precoded that might be found in an in-depth study in aquestions has been set very low, so that a mech- particular culture. Also, given the study's pur-anical application of scoring algorithms for de- poses, it did not give us a full picture of any

.pendence would result in inappropriate alternative, non-diagnostic conceptualizations ofdiagnosis. alcohol and drug problems which might be

..prevalent in a given society.The study's findings suggest that, at the level

Conclusion of diagnostic interview items, cross-culturallyThe goal and promise of a valid and useful comparable formulations can usually be found.cross-cultural epidemiology of alcohol and drug However, the task is not straightforward. Theconditions remains before us, in no way invali- difficulties start at the level of cultural differencesdated by the results of the CAR study. But the in ways of thinking about subjective states. Manyresults do highlight the challenges inherent in of the items in instruments like CIDI assumethis endeavour. The CAR study's results suggest that the respondent can report on individuatedthat more than translation is involved in adapting feelings and states of mind in a detached way,

218 R. Room et al.

which is not a way of thinking used in all cul- items in a culture, but if the culture does nottures. The Cilll items assume what has been accept the diagnostic connections built into ancalled a "modem self' standing outside and sep- international scoring algorithm, the algorithm isarate from feelings and sensations, a self that can likely to yield a diagnosis which would be seen asevaluate, describe, and sometimes even control inappropriate and even wrong within that so-subjective states (Toulmin, 1990). The items ciety. For most of the ICD-I0 criteria for thethus reflect a post-Enlightenment Eurocentric alcohol dependence syndrome, indeed, the keytradition in which individuals have feelings informants from one or another of the study siteswhich they can express in words-not only sen- reported that, while the criterion was recogniz-sations, but also emotional states. In this world- able to them, it was seen as a commonplace orview, individuals may focus on and talk about even admirable quality in their culture, andthese feelings; and it is conceivable that relieving would not be recognized as diagnostic of depen- ,those feelings might be a reason for taking a dence or addiction.drink or a pill. The CAR study findings thus pose for nosolo-

There are also difficulties at the level of direct gists the question of whether the six criteriatranslation. Often there is no equivalent in com- designated in ICD-I0 are the most appropriatemon use for English terms, for example for with- way to characterize dependence as a disorder indrawal and tolerance. The English term may be different cultures. There are three main alterna-borrowed or a direct translation may be used, tive solutions to this question. One choice wouldbut this may have little meaning, or a very differ- be to decide that the validity of a criterion ofent meaning, for a lay respondent. Even where dependence in a given society does not dependthe denotations of a term or concept seem simi- on whether it is accepted as symptomatic of alar, two cultures may differ in where they draw disorder in the society. A second choice forthe threshold for applying it. The Korean term future nosologists would appear to be to returnfor "hangover", for instance, has a high to a culturally differentiated definition of depen-threshold, connoting that the drinker still has dence, as in Jellinek's Greek-letter typology or insubstantial alcohol in the blood the next morn- the earliest World Health Organization definitioning. On the other hand, Bangalore respondents of alcoholism as use going beyond "the socialreport experiencing tolerance and difficulty cut- drinking customs of the whole community con-ting down on the basis of a drinking pattern that cerned" (WHO, 1951). A third choice would bewould seem to a respondent from a Mediter- to undertake a search for new criteria or forranean wine culture to be homeopathic doses of reformulations of criteria which would be univer-alcohol. sally validated as diagnostic of dependence.

The study's findings suggest that comparabil- Along with the issue of cultural differences inity is also an issue at the levels of criteria and conceptual linkages, there is also an issue ofconcepts. Some criteria used in diagnostic sys- cultural differences in the threshold for positivetems appear to be culture-bound, depending for responses to items or criteria. The Bangaloretheir force on specific cultural ideas about the reference cases described above alert us to theuse of time, about "alternative pleasures", about possibility of overdiagnosis, particularly with aself-control, or about customary patterns of sub- rigidly structured schedule and scoring pro-stance use (e.g. narrowing of the repertoire). cedure, as is commonly used in epidemiologicalAlso, while cultures readily borrow concepts and studies, in cultural situations where there isterms for behaviours that are seen as alien-as much social disapproval of any use. In the localdrug use is in many places-they may be quite circumstances of Bangalore, formulations such ,resistant to applying new medical terminology to as the Cilll question about whether there werewhat is seen as mundane and familiar, e.g. wine ever any objections about the respondent'sdrinking in Mediterranean societies. drinking, as the Bangalore investigators noted,

Perhaps the most problematic aspect of the are likely to be answered positively by anyfindings for cross-cultural comparisons concerns drinker. In an environment where drinking iscultural differences in the perceived linkage be- relatively infrequent and limited by finances andtween items or criteria, on the one hand, and social disapproval, respondents may also give adiagnoses on the other. An epidemiological study level of attention to their desires and to themay get usable and reliable answers to interview effects and possible consequences of drinking

Cross-cultural applicability of diagnoses 219

which those from a "wetter" cultural environ- pathological drinking: a cross-cultural comparison,t . ght find t d Alcohol Health and Research World, 17, 190-195.

men ml exaggera e . ...DAWSON, D. A., GRANT, B. F. & TOWLE, L. H. (1993)These data from a culture where drInking IS Cross-cultural issues in the diagnosis of alcohol-use

disapproved of remind us that the same kinds of disorders: evaluation of diagnostic criteria and symp-problems of overdiagnosis could well arise with tom-item indicators based on the CARS projectrespect to dependence on other drugs in cultures SARS substudy results, Paper presented at the 6th

h f th dru . 1 . 11 d .lberoamerican Congress on Alcohol, Tobacco and Drugs,were use 0 e gs IS strong y socIa y Isap- Mexico City, 25-29 May.

proved of. The disapproval results not only in GORENC, K.-D., BRUNER, C. A., NADEl$TICHER, A.,high rates of social and interactional problems PACURUCU, S. & FEUERLElN, W. (1984) A cross-cul-from use but also in increasing the perceived tural study: a comparison of German, Spanish andf'

th dru t t k ' 1' £ d Ecuadorian alcoholics using the Munich Alcoholism.power 0 ego a e over ones Ie an T t( "AT' T) A .'¥ al ifD ndAlcholes iV , , mencan Journ 0 rug a 0

actions, and m encouragmg users to mtrospect Abuse, 10,429-446.

carefully concerning variations in the psychoac- HAlL, W., SAUNDERS, J. B., BABOR, T. F. et al. (1993)tive effects of the drug (tolerance) and in feelings The structure and correlates of alcohol dependence:

.and urges concerning use. WHO coll~borative project on the early det~ction of...persons WIth harmful alcohol consumption-III,

One other aspect of the findmgs deserves a Addiction, 88, 1627-1636.comment. The field of alcohol and drug prob- HELZER, J. E. & CANlNO, G. J. (Eds) (1992) Alcoholismlems has many terms-such as alcoholism or in North America, Europe, and Asia (New York andaddiction-which originate from medical usage Oxford, Oxford University Press).

...JEU.1NEK, E. M. (1952) Phases of alcohol addiction,but which have passed mto common usage.m Quarterly Journal of Studies on Alcohol, 13,673-684.many languages. Often such terms were ong- JEU.1NEK, E. M. (1960a) Alcoholism, a genus andinally put forward with the aim of removing the some of its species, Canadian Medical Associationmoral judgements imputed by the terms they Journal, 83, 1341-1345.

1 d Th CAR d find ' . d .JEU.1NEK, E. M. (1960b) The Disease Concept of Alco-rep ace. e stu y mgs remm us m h I . (H 'ghl d P k NJ Hillh P )..0 ISm I an ar" ouse ress.vivid terms, however, how much the dIagnosIs of KlAUSNER, S. Z. & FOULKS, E. F. (1982) Eskimoalcohol and drug disorders depends on the Capitalists: Oil, Politics, and Alcohol (Totowa, NJ,mores in a society, and how quickly and easily Allanheld, Osmun).new medical terminology can take on the same KUNITZ, S. J. & LEVY, J. E. (1994) Drinking Careers: A

..Twenty-Five-Year Study of Three Navajo PopulationsmorallZmg tones as the old. (New Haven and London Yale University Press).

The findings of this study provide a fertile LEVINE, H. G. (1992) Tem~erance cultures: concernground for future comparative and collaborative. about alcohol problems in Nordic and Engiish-research on these disorders. The network of cen- speaking cultures, in: LADER, M., EDWARDS, G. &

...DRUMMOND, C. (Eds) The Nature of Alcohol and~es and experts. created m thIS s~dy IS ~ po~en- Drug Related Problems, pp. 15-36 (Oxford, Oxfordtlal asset for thIS research, and ItS commg mto University Press).existence is undoubtedly one of the study's most MAcANDREW, C. & EDGERTON, R. B. (1969) Drunkenimportant results. Comportment (Chicago, Aldine).

PrrrMAN, D. J. (1967) International overview: socialand cultural factors in drinking patterns, pathologi-

.Acknowledgements cal and nonpathological, in: PnTMAN, DAVID J.This project was supported by funds from the (Ed.) Alcoholism, pp. 3-20 (New York, Harper and

US federal government under a cooper.ative Ro~:;,'L. N. & REGIER, D. A. (Eds) (1991) Psychiat-, agreement (5UOI MH35883) from the National ric Disorders in America: The Epidemiologic Catchment

Institutes of Health (formerly the Alcohol, Drug Area Study (New York, Free Press).and Mental Health Administration). Support for ROBINS, L. N., WING, J. K., W~CHEN,.H.-U. .et al.

.articipation was also drawn from collaborating (19~9) The ~omp?slte. ~ternatlonal D~agnostlc In-~ ..., .tervlew: an epIdemIologic Instrument suItable for usemstltutlons at each s1.te. The VIews expressed m in conjunction with different diagnostic systems andthis paper are the collective views of an inter- terns and in different cultures, Archives of General

..national group of researchers and do not necess- Psychiatry, 45, 1069-1077.arily express the views or policy of the World ROOM, R. (1984) Alcohol and ethnography: a case of

problem deflation?, with comments and a rejoinder,Health OrganIZation or supportmg agencIes. Current Anthropology, 25, 169-191.

ROOM, R. (1989) Responses to alcohol-related prob-References lems: characterizing and explaining cultural wetnessBENNETT, L. A., JANCA, A., GRANT, B. F. & SARTO- and dryness, Paper presented at an international confer-

RlUS, N. (1993) Boundaries between normal and ence, La ricerca ltaliana sulle bevande alcoliche nel

220 R. Room et al.

confronto internazionale, Santo Stefano Belbo, Italy, Diagnostic InteIView (Cilll)-results of the multi-22-23 September. centre WHO/ADAMHA field trials (Wave I), British

ROOM, R. (1992) The impossible dream?-routes to Journal of Psychiatry, 159, 645-653.reducing alcohol problems in a temperance culture, WORLD HEALTH ORGANIZAnON (1951) Expert Com-Journal of Substance Abus~, 4, 91-106. mittee on Mental Health, Report of the First Session

SARTORIUS, N. (1989) Making of a common language of the Alcoholism Subcommittee WHO Technicalfor ~sychiatry: development of ~e classi~cation, be- Report Series, 42, September. '

ha~~ural and developmental dlsord~rs In the 10th WORLD HEALTH ORGANIZAnON (1990) Composite In-reVISion of the ICD-10, WPA Bullenn, 1, 1. ternational Diagnostic Interview (CIDI) core version

TAIPALE, I. (1979) Terms Applied to Deviant Drinkers in 1 0 (WHO G) ,Finland, Reports from the Social Research Institute W . H ' enOeva.

(1992 ) Th 'CD 10...ORLD EALTH RGANIZAnON a e., -of Alcohol Studies, No. 134 (Helsinki, SRIAS). .. if M al nd B h . I D .

deTOULMIN, S. (1990) Cosmopolis: The Hidden Aspects of CI~~ficatwn .0. ent ~ e. a~ra. lSor rs:Modernity (New York, Free Press). ClinICal Descnpnons and DiagnostIC Guidelines (WHO"

WING, ]. K., BABOR, T., BRUGHA, T. et al. (1990) Geneva). J-fSCAN: schedules for clinical assessment in neu- Wo~ HEALTH ORG~nON (1992b? Schedules for

ropsychiatry, Archives of General Psychiatry, 47, 589- ClinICal Assessment In Neuropsychiatry (WHO, :'593. Geneva). ~.

WrrrcHEN, H.-V., ROBINS, L. N., COTn.ER, L. et al. WORLD HEALTH ORGANIZAnON (1993) The ICD-10 ¥I(1991) Cross-cultural feasibility, reliability and Classification of Mental and Behavioural Disorders: Di-sources of variance of the Composite International agnostic Criteria for Research (WHO, Geneva).

0'l"'~

Addiction (1996) 91(2),221-230

COMMENTARIES

Comments on Room et al.'s "WHOcross-cultural applicability research ondiagnosis and assessment of substance usedisorders: an overview of methods andselected results"

Published below are six commentaries on the research report by Room et al. (this issue of Addiction) followedby their response

The difficulties of building a lift in the Consider now the dissimilar groups of peopleBabel tower taking an interest in the treatment of alcoholPhilippe Batel abuse. They are as different as physicians,Let .. 1 B b 1 t Th .psychiatrists, psychoanalysts, psychologists,

us lffiagme a comp ex a e ower. IS .t h . fl din t .nurses; but also peers, socIal workers, M.

ower as nme oors correspon g 0 rune.difii t ltu Id .d T f th ThInk that, on each floor, all these people

eren cu res wor -WI e. wo 0 em ...se th . ffi . 1 1 b t could gather to dISCUSS this concept of mIsuse,

u e same pnmary 0 cIa anguage, u ..different languages and even different dialects and more precIsely ~e n~non of a!coholof a given language can be spoken on anyone dependence, among the inhabItants of their ownfloor. floor. Each expert would have, according to their

On each floor, people of different ethnic educati~n, experience.. or habit~, a differentorigins or various religious confessions coexist. percepnon and. defininon of thIS concept ofConsider now one of the main psychoactive dependence. Think now of a group of peopledrugs used allover the world: alcohol. Consider considered by the entire range of expertsits use and misuse within the nine floors, issues described above as alcohol-dependent after awhich are of course related to the availability of long discussion to seek a consensus. The

, alcohol in the country, the way it is consumed, significance of different. ae~iological factors, thethe habits of the population, the threshold of type, the pattern of dnnking and the form ofsocial tolerance, the public health policies alcohol dependence can vary considerably fromregarding alcohol, the social and economic levels one person to another.

.and numerous anthropological and historic Consider now a team of researchers comingfactors. from every floor under the auspices of an

Furthermore, the misuse of alcohol is a international organization (WHO) to work inmulti-factorial process linked to genetic, this tower with its complex stratification ofpsychological, cultural and social factors as well multi-dimensional culture.as a high individual inclination to dependence. This team, prompted by the purpose ofEach floor, according to each culture, has a finding a common language, has decided todifferent interpretation of this idea of misuse. explore the criteria for alcohol dependence.The anthropological dimensions of madness, Thanks to a complex but irreproachableguilt and tolerance thresholds for what counts as methodology, this team sets up a long anda problem in a given culture is highly different serious work to prove the comparability,from one floor to another. reliability and acceptability of the criteria

0965-2140/96/020221-10 $8.00 @ Society for the Study of Addiction to Alcohol and other Drugs

Carfax Publishing Company

222 Commentaries

(ICD-IO) as used on each floor. This ambitious Towards studying the real cross-culturaland courageous project comes to build a lift to similarities and differencespass from one floor to another, but also to break Shekhar Saxenathe intracultural stratifications. What is more, Sometimes it is more important to state thethe aim of standardization of this concept of obvious than to discover something new. Thedependence implies the need to consider each WHO cross-cultural applicability research onfloor crosswise through differences between diagnosis and assessment of substance abuseexperts, differences between the instruments disorders draws the conclusion that "theused (CIDI or SCAN), and number of necessary attitudes of societies to alcohol and drug usecriteria to obtain a diagnosis. affect the use of criteria and the making of

Can we really believe we achieve the stated diagnoses". This should have been obviousintentions despite the fact that some data were to any discerning clinician or researcher in themissing? field of substance use. However, with the added

The absence of distinction between hangover force of a multi-centred, multi-component .and withdrawal in almost half the countries is international study, this simple statement shouldextremely worrying. Physical withdrawal is attain the importance that it deserves. I wouldunanimously admitted because of its link consider the massive effort that has gone intowith physiological effects of alcohol, but the this study well rewarded if it helps in removingpsychological and affective symptomatology the exaggerated and unrealistic sense of(regret, guilt, shame linked to alcohol abuse) is complacency that has come to surround the usevery differently assessed from country to of "cross-culturally valid" diagnostic categories,country. "objective" research criteria and "international"

The two strong agreements of loss of control interview schedules. Essentially, they all consistand the point when medical attention is seen as of words and like all other words, they mean atnecessary could be reassuring confirmation of least different things to different people.the concept of alcohol dependence. Unfor- The applicability of highly structured interviewtunately, however, the notion of loss of control schedules including CIDI in less-developedand craving could not be clearly distinguished in countries among less-educated respondents indifferent countries. Was the standardization of languages other than the original remains, at thethe concept of dependence fulfilled through a face of it, highly suspect. This is partly becausework with too many compromises on too many structured interview technique itself may not belevels? Does the international scientific com- suitable for these respondents. In addition, themunity really need a common language for a conceptual and linguistic difficulties remainworld-wide phenomenon such as alcohol insurmountable, in spite of sustained efforts todrinking? Shall w~ really try to reduce inter- overcome them. In this context, systematiccultural measurement differences with the risk of assessment of applicability of interview schedulesimpoverishing the particularities of each country? attempted by this study is more than welcome.

The establishment of an international network The basic plan of the study including thefor research on alcohol and dependence seems to selection of centres and the five-componentbe the foremost benefit of this work. It is a very methodology appears appropriate towardspraiseworthy accomplishment, but what sort of a fulfilling the goal of this study. However, anetwork could we imagine for the future? Would number of compromises seem to have been .

it be possible to extend it to the most active made during the actual execution of the study.countries (in terms of research), or to those Particularly unfortunate is the considerablewhere the alcohol consumption per capita is at curtailment of the reference cases componentthe highest rate? and dropping of the semi-structured exploratory

Thus could we optimize the efficiency of interview about the answers to CIDI questions.building of a lift in the Babel tower. Thus an opportunity has been lost to find out

what the respondents understood of anyPHIliPPE BATEL particular CIDI question and whether there wasHOpital Beaujon, a good match between what the question pur-

100 Bd dn General Leclerc, ported to ask and what was actually reported. No92118 Clichy Cedex, France. other method seems as powerful to clarify these

Commentaries 223

questions than to ask the respondent directly A wealth of data and experience on which itabout what made him respond in the particular will be essential to capitalizeway to each question. The results, although John E. Helzerqualitative, would have ~een ~lluminating. .There are few examples of cross-cultural

The results present~d m thIS paper are highly comparisons in which there is sufficientselected. Even the ba~lc back~ound d~ta ?n the consistency across sites that the resulting datarespondents are not given. This makes It dIfficult are informative. Typically, illness concepts,to comment on whatever findings are given and interview instruments, study design and datathe conclusions drawn thereof. The literacy and analysis are so variable that any attempt ateducation levels of the respondents in the key comparison is hopelessly frustrated. It is

, informant, focus group and refe~ence cases impossible to say if differences in illness rates arecomponents of the study are not given. These methodological, definitional or real. Anwould be crucial for interpretation of the results informative cross-cultural dataset requiresfrom at least some centres. For example, immense good luck as was the case in our own

.Bangalore centre may have included educated recent effort (Helzer & Canino, 1992), or anrespondents who would find it easy to immense commitment of resources as in the caseunderstand and reply to most items, or illiterate of the CAR project, detailed in this paper byrespondents who would find it extremely difficult Room et al.to do so. It is hoped that the more detailed In our own case, insurmountable cross-publication being prepared will provide all the national differences were minimized largely bynecessary information. serendipity. The Diagnostic Interview Schedule

Behind every item of an interview schedule are (illS) used in our set of studies was firsta series of questions which are implicit but often developed for the Epidemiologic Catchmentnot recognized. Is the phenomenon known to Area (ECA) survey (Robins & Regier, 1991).occur in the culture? Is it recognized to be The ECA gained international attention andoccurring? How is it conceptualized and to what investigators in several other countries set out tois it attributed? How much importance is replicate the survey in their own countries.assigned to it? Is it reported? If yes, in what form Interested investigators came to St Louis, whereand using what words? The study by Room et al. the illS was developed, to obtain interviewhas attempted to answer some of these questions. training. This also afforded an opportunity toAnswering all these questions is a much larger discuss and help clarify the diagnostic conceptstask and a single study, however ambitious, can on which DSM-III, and hence the illS, wasnot be expected to undertake it. This study has based and to review and discuss the studypaved the way for many more efforts in this methodology used in the ECA.direction. Although the goals of the CAR and our

While developing common language is an cross-national consortium were similar, the twoeminently worthwhile, perhaps a basic task, it efforts differed conceptually and practically. Forshould not be an end in itself. Even if people example, the methodology described here byspeak the same language they will still report Room et al. for the adaptation of the CIDI tovarying experiences and behaviour spanning each culture is good deal broader and moreacross different cultures. True cross-cultural dif- deliberate. First, the translation/back-translation

, ferences of qualitative and quantitative nature are process was uniform at each site. Secondly, keypresent very widely in the area of substance use, informants, focus groups, self-administeredabuse and dependence. Speaking a common lan- rating scales and reference cases were all used toguage will at best be the means to study the real better understand the subtlties of intraculturalcross-cultural similarities and differences. This meaning of terms and concepts used in the CIDIwill eventually help in arriving at a deeper under- and the diagnostic criteria on which it is based.standing of the field, leading to better patient In our study, translation and adaptation was leftcare as well as more appropriate policy initiatives. entirely to local investigators. All sites recognized

SHEKHAR SAXENA the critical importance of this step andDepartment of Psychiatry, approached it conscientiously, but the methods

All-India Institute of Medical Sciences, and expertise differed across sites. Similarly withNew Delhi 110029, India. study design, that used in the ECA became the

224 Commentaries

model, but was adapted as necessary in specific The CAR represents a wealth of experiencesites. We did introduce more consistency into and data in cross-cultural alcohol research andthe data analysis. For example, the illS for psychological medicine generally.diagnostic scoring programs were used in their Capitalizing on this resource is critical. Oneoriginal form at each site; but apart from the core useful effort would be to determine the ways indiagnostic data, sites had complete freedom to which the methodological refinements in CARadd to the interview whatever additional enhanced the resulting dataset over what wouldquestions they desired. Core analyses were have been achieved if less elaborate methods hadreplicated at each site. Again, investigators were been used. It may be that some of these effortsfree to pursue additional analyses. had a major impact on cross-cultural consistency

Even the focus of investigative attention or validity and others had little effect. This coulddiffered in the two projects. Obviously this initial greatly help future investigators in making their"overview" by Room et al. is merely the first of own cost/benefit decisions regarding translation,many publications. However, even this initial study methods and data analysis.publication focuses on a group of issues that our Seldom are the resources available to permit a .group, perhaps naively, devoted much less massive effort such as the CAR. Thus, anotherattention to. One of these is the concern that way of capitalizing on this resource is to considerruns throughout Room et al. about the how what was learned here could be applied byappropriateness of imposing norms from one future individual investigators whose primaryculture onto other, very different cultures. This interest may be in their own culture but who arewas less of an issue for us since investigators also interested in contributing to and benefitingfrom each participating site had decided from a cumulative cross-national body of data,independently that a replication of the ECA perhaps by using standardized instruments suchwould be desirable and that using the DSM-III as the CIDI that have been applied in otherand the illS would be appropriate. The almost cultural contexts. Shapiro & Stromgren (1979)immediate, nearly world-wide popularity of the note with regret that the cumulative impact ofDSM-III (Spitzer et al., 1983) and the perceived research done throughout the world isutility of its diagnostic concepts in very diverse "considerably reduced by the fact that results ofcultures (Helzer & Canino, 1989) provide the various studies are not comparable becauseevidence in favor of the broad applicability of the of differences in diagnosis and assessmentDSM. Another issue is the conundrum this methods..." As useful as large and coordinatedpaper highlights: that adapting instrument items, efforts such as the CAR or our cross-culturalfor example deleting "softening" phrases such as consortium are, the goal of greater cross-cultural"did you ever feel' so as to make the translation understanding would be forwarded even more ifmore culturally appropriate, may detract from the kind of grass-roots comparability thatthe comparability of the item being translated. It Shapiro & Stromgren call for could be achieved.seemed to us that if the investigators and If the wealth of experience gained in CAR couldtranslators understood the original intent of the be utilized to assist such individual initiative,criterion on which an instrument item was perhaps through the World Health Organization,based, they would be in the best position to the payoff could be useful far beyond the CARadapt the item(s) to their own culture and the study itself.consequent variation in interpretation across JOHN E. HELZERcultures might not be much greater than the Professor, The University of Vermont,obvious variation in interpretation that occurs College of Medicine,between respondents within a culture. Evidence Department of Psychiatry,that our approach generated valid cross-cultural Medical Alumni Building,data is the great consistency we found in such Burlington, Vermont 05405-0068, USA.things as relative symptom frequency, illness riskfactors and co-morbidities. These and othersimilarities would probably have been greatly HELZ J E & C G (1989)Th . I ' ,

d .f th .. d th " HR, ., ANINO,. e Imp Icat1onsattenuate 1 e cntena an e mtemew were of cross-national research for diagnostic validity, in:inappropriate or had been poorly adapted to the ROBINS, L, N. & BARRF:rT, J. E. (Eds) The Validitynew cultural contexts. of Psychiatric Diagnosis (New York, Raven Press).

Commentaries 225

HELZER, J. E. & CANINO, G. J. (1992) Alcoholism in practice. If the alternatives are considered fromNorth Am~rica~, Europe, and Asia (New York, the viewpoint of policies and programmes, it isOxford UnIversity Press). ..readily seen that the first alternative has a strong

ROBINS, L. N. & REGIER, D. A. (1991) PsychiatrIC. . Th d . b. d. 1 . d.. d 1Disorders in America. The Epidemiologic Catchment pOSItion. e ommant 10me lca, m IVl ua-Area Study (New York, The Free Press). centred thinking informs western approaches to

SHAPIRO, R. W. & STROMGREN, E. (1979) The the problems of substance dependence. Itsrelevance of epidemiological methods, techniques, fundamental axiom is that the "disorders" areand findings for biological psychiatry, in: Handbook ..fi B . lo . al Ps h ',..., D rt I (N Y k M I essentially the same everywhere, only theIror 10 glC cyc 1a.'J' .a ew or, arceDekker, Inc.). prevalence varies, The line of action is to

SPITZER, R. L., WILLIAMS, J. B. W. & SKODOL, A. E. continue to spread the message to the rest of the(1983) International Perspectives on DSM-III world, introduce and enlarge treatment systems

, (Washington, DC, American Psychiatric Press). and try to make sure that local practitioners get

the proper schooling. All this can easily becombined with the third alternative; tinkering

. Wh at h 1 d d h .with criteria and measurement instruments keepsy co 0 an rug researc remams a ., .

. at ..researchers busy. Such actIVItIes have powerfulmargIn actiVIty d .. h d hJuh P agents, an at present It IS ar to see ow

a artanen tho d . d ld b dIS ommant tren cou ecome reverse.

One cannot help having mixed feelings about the Perhaps the authors were too bashful to pointresearch project reported by Robin Room and this out?his many co-workers. Its novelty of approach, the However, such policies and programmesdiversity of methods, theoretical sophistication, necessarily encounter obstacles and problems.the complex logistics of a study carried out at Money runs out constantly, indigenous notionsnine different sites and the sheer amount of of substance use tend to persist and, as theeffort cannot but win admiration. authors note, newly introduced "scientific"

Yet the results leave the reader, and perhaps notions easily become "contaminated" with localthe authors as well, disappointed. While the interpretations. Therefore the need toauthors bravely state that "The goal and promise understand the nature of cultural variation, asof a valid and useful cross-cultural epidemiology well as specific cultural idiosynracies ofof alcohol and drug conditions remains before substance use, remains. It is a good thing to beus, in no way invalidated by the results of the reminded ofJellinek in this connection; he in factCAR study", it would be hard not to conclude knew a great deal of these matters.that so far the efforts to develop universally valid As regards cultural variation in alcohol use theinstruments for the diagnosis and measurement authors seem to have finally realized how utterlyof substance use disorders have failed, and that inadequate the "wet/dry" dimension is as itsthe observed cultural variety in the meanings broad overall characterization. At the very leastand interpretations of alcohol and drug use one has to recognize that two basic dimensionspromises little success in the present circum- are involved; there is variation in how extensively

.stances. a society is engaged with alcohol (frequency ofWhat, then, are the implications of the cultural drinking, presence of alcohol in ritual contexts

applicability study? The authors formulate, and in everyday life), and in how intensively.concerning the notion of dependence, three drunkenness is sought among significantly large

alternative ways to look at the issue of cultural sections of population (consumption perdifferences: (1) to continue the application of drinking occasion, the role behaviour of drunkenexisting diagnostic instruments not caring about people). There are indications that, even thoughwhat the local people think of their validity; drinking is inherently risky, some societiesor (2) to assume a culturally differentiated extensively engaged with alcohol can alsodefinition of dependence Ii la Jellinek; or (3) accommodate and contain, or at least effectivelyto undertake a search for new criteria for a hide from public view problematic consequencesuniversally applicable definition of dependence. of even quite heavy drinking.

These formulations are rather unsatisfactory Such general dimensions by no meanssince they remain at the level of concepts and do suffice to characterize the configuration ofnot relate to the actual medical and social alcohol-related problems in a particular society.~

226 Commentaries.

Here the essential difficulty is that this calls for Present difficulties, future possibilities

an analysis of the socio-economic and cultural, George E. Woody

historically determined context of drinking and its consequences. The report on the This IS an mterestm~ pape~ th~t pomts out ~e

cultural applicability study contains interesting many unresol:ved diagnostic Iss~es that anse

observations on how such quite general matters w~en. attempting to apply ~ehaVlorally fo~us.ed

as cultural differences in ways of thinking about cntena that were developed m Western societies

subjective states or ideas about the use of time to non-Western cultures. Indications of these

affect the definition and perception of diagnostic problems have previously emerged

alcohol-related problems. The forthcoming book over .differences between DSM and ICD

will undoubtedly present such findings in much regardmg the concept of "substance abuse".

more detail. There is, however, a real danger that Although American psychiatrists feel

these results will remain a heap of separate ideas, comfortab~e with this ~erm, the framer.s of ICD

connected at most to current pet theories that have consistently declmed to accept It due to

most often have been psychological. The record problems that arise when attempting to define it .

of alcohol and drug research as a totality, with across a range of cultures.

few exceptions, displays an incapability of One of the first things that crossed my mind

grasping the total societal context of alcohol and on reading this paper was an immediate feeling

drug phenomena. of skepticism about epidemiological data on the

The irony of the situation is that these matters prevalence of substance use disorders in

are primarily dealt with in a health non-western cultures. The data, especially the

perspective-internationally WHO is the main anecdotes that are provided, clearly demonstrate

actor. Yet in many a developing country alcohol- how criterion items and language can mean very

and drug-related problems are essentially different things across cultures; in some cases,

problems caused by poverty and coping with one culture may not even have language to

them is hampered by corruption. In the most describe a term that is used in another. These

extreme cases-and the history of colonialism varied interpretations of criterion items and

knows many-these problems are a consequence language differences can translate into

of the total breakdown of society. In yet other prevalence rates of "disorders" that are quite

countries such as Japan, and presumably also disparate, thus leaving the reader wondering

Korea, heavy male drinking and its what people with substance use disorders are

consequences are essentially related to the really like in non-western societies.

patterns of work life and the relationships This paper indicates that one of the major

between genders. Such examples suggest that the contributors to this variability arises from the

very logic of substance use disorders in particular fact "the diagnostic criteria. ..assume a

cultures derives from the socio-economic and self-consciousness about feelings, knowledge and

cultural matrix of the whole society. consciousness which is foreign to the folk

There is, of course, a reason why alcohol and traditions of some cultures". This finding is

drug researchers, irrespective of whether they are interesting since it teaches us about the way

medical doctors, sociologists or anthropologists, people think in other cultures, while also

shirk adequate analysis. Taking cognizance of pointing to concepts we might wish to avoid in

the essential determinations of their object of future iterations of criterion items. .

study would reveal that in practice there is very Another interesting finding, and one which

little they can do to help to deal with substance suggests that consistent interpretation of

use disorders. Staying within the narrowly criterion items can be achieved, is that few

circumscribed sphere of their specialized problems were noted for "intoxication" and

research spares them from such unflattering "withdrawal", areas where objective findings are

revelations, but it also means that alcohol and more easily observed. The greatest difficulty

drug research remains a marginal field of study. occurred in items describing less objective

JUlIA PARTANEN concepts, particularly "narrowing of the

Social Research Institute of Alcohol Studies, repertoire", a finding that supports the ICD-IO

PO Box 350, decision to delete this item from the criterion set.

00101 Helsinki, Finland. Using these two extremes to define the ends of

Commentaries 227

an "agreement/disagreement" continuum, one it would be a mistake to point to the cross-might conclude that the consistency with which cultural difficulties that this paper demonstratesthe items are interpreted is a function of the and conclude that the clinical syndrome ofdegree to which they reflect objective findings. "substance dependence" signifies little moreAn analogy might be drawn to diagnosing than cultural values. However, it is also clear thatpneumonia or cancer. It is likely that these additional work is needed to develop betterillnesses are diagnosed in a reasonably consistent cross-cultural applicability of our currentmanner across cultures, primarily because the definitions.underlying conditions are defined by X-rays or One step that might be taken along these linesmicroscopic examinations of cells. is to extend the focus group and reference case

.Unfortunately, psychiatric disorders in general methods that are mentioned here for use withinand substance use disorders in particular are cultures, and apply them across cultures. Thelargely behaviorally defined. There are no strategy could be something like the following:objective, biological tests for schizophrenia, give experienced clinicians from different

« generalized anxiety disorder, manic depressive cultures case reports and/or videotapes of people

illness or for substance dependence. Thus we are from a wide range of cultures who demonstrateleft with doing the best we can with behaviorally different consumptive patterns; make the casebased criteria, while awaiting future reports as objective as possible; ask the cliniciansdevelopments that provide opportunities for to rate each individual and determine if he/shemore precision and objectivity. meets criteria for a diagnosis and if so, what;

There are signs that such developments are if there is no diagnosis, ask why; if there is aoccurring. For example, our ability to examine problem but it does not fit a ICD or DSMfunctional alterations in the brain has been category, encourage the clinician to formulateimproved by neuroimaging. One exciting finding a new category. After completing this exerciseis the emerging data that people who are hold workshops, conference calls or somedependent on cocaine have changes in the other means of interactive communication indensity of D2 receptors, as well as other which the clinicians collectively review eachindications of functional "down-regulation" in case with an eye toward developing consensusparts of the central nervous system.! The recent about the meaning of criterion items anddata that naltrexone reduces relapse to alcohol classifications.dependence among recovering alcoholics implies Such a process would be complex and difficultthat biological alterations in the central nervous to operationalize in a scientifically meaningfulsystem (CNS) are associated with alcohol way, but it could be useful in moving the fielddependence! It is entirely possible that work toward better cross-cultural agreement. Inof this type will progress and lead to addition to improving criteria items foridentification of alterations in brain function dependence, such a process might also helpthat are diagnostic of substance dependence, settle the long-standing disagreement over the

.and that will improve our current behavioral meaning of "abuse" and "harmful use". Bothmethods. ICD and DSM agree reasonably well on the

However, when thinking about the problems meaning of dependence, at least in westernassociated with cross-cultural applicability of cultures, but unfortunately there is not good

.diagnoses, as this paper encourages us to do, consensus on abuse and harmful use.3 Betterit is important to be mindful of the clinical agreement between ICD and DSM on this point"roots" of the dependence syndrome. Those might be an additional and welcome outcome ofworking in western cultures who practice a cross-cultural study of diagnostic criteria forsubstance abuse medicine have a reasonably dependence.good idea of the syndrome that is being GEORGE E. WOODYdescribed when someone is diagnosed as Substance Abuse Treatment & Research Center,substance-dependent.3 All of the criterion items Philadelphia Veterans Affairs Medical Centerthat we use to establish dependence originally and Department of Psychiatry,came from careful observations by skilled University of Pennsylvania,clinicians who had interacted with thousands of 39th & Woodland Avenues,individuals over extended periods of time. Thus, Philadelphia, PA. 19104 USA.

228 <;ommentaries

1. VOLKOW, N. D., FOWLER, J. S., WANG, G. J. et al. Pentecost involves a translation that is always(1993) Decreased dopamine D2 receptor a betrayal. Indeed in Italian a traduttoreavailabil~ty i~ associ~ted with reduced frontal (translator) is a traditore (traitor).'metabolism m cocame abusers, Synapse, 14,169-177. However, as the paper underlines, when using

2. VOLPICEUl, J. R., ALTERMAN, A. I., HAYASHIDA, a word both at the scientific and lay levels oneM. & O'BRIEN, C. P. (1992) Naltrexone in the should expect that widely used concepts have thetreatment of. alcohol dependence, Archives of same meaning. In Italy, for example, this is notGeneral Psychiatry, 49, 876-880. th fi th d " 1 h 1." h3. CO'ITLER, L. B., SCHUCKrr, M. A., HELZER, J. E. e ~ase. or e wor .a co. 0 Ism, w oseet al. (1995) The DSM-IV field trial for substance meaning m health professional circles has swunguse disorders: major results, Drug and Alcohol from a narrow definition-what we could callDependence, 38, 59-69. alcohol dependence-to a broader definition ,

close to the more recent idea of alcohol-relatedproblems, or of hazardous drinking (Allamani et

T 1 t d tr "t ai., 1994). The narrower approach has moralrans a ors an at ors AllAll .connections Wlth the hard Idea of viciousaman amarn b h . d .. d d b th . th d. 1e aVlour, an IS Wl esprea 0 m e me lca

Many years ago, at the beginning of my personal and public contexts. The few authors whointerest in alcohol issues, when I asked a quite proposed a broader approach used softer terms,famous alcohologist in the United States his such as "little" or "minor" alcoholism, todefinition of a person affected by alcohol underline the neutrality of the wording whenproblems, I received the following answer: "any extended to a large part of the Italian populationperson who applies to an alcohol service for which one would never dare to call "alcoholic".advice or treatment". Certainly nowadays, the flourishing movements

This psychiatrist meant, I think, that medical of Alcoholic Anonymous and of the Clubs forpractice surpasses theoretical arguments. Treating Alcoholics tend to conform to theHowever, the appeal of a medical treatment broad opinion by which alcoholism, even when itservice for alcoholics relies on the one hand on is not yet overt, is a more diffused problem thanthe service being highly visible and up to date, is commonly talked about.according to present scientific principles, and on The confusion between terms shows up in thethe other on how lay people perceive alcohol mass media and even with Italian legislators.problems and the service as able to help solve Newspapers are prone to publish news whichthose problems. Each community has its own highlights the dramatic, such as the alcoholicsalcohol problems and alcohol treatment services, among youngsters, or to reframe informationand its own experiences of how health about the high-risk drinking of a communityprofessionals and their clients interact. That is to population as a problem of alcoholism spreadsay, we are facing more a problem of local and throughout that population. Also a number ofcultural definitions than an universal one. legislative proposals have been recently

This excellent paper by Robin Room and his presented at both regional and national level .colleagues is built around three ideas. The first which refer more frequently to alcoholism thanis the real relevance of cultural differences in to alcohol-related problems.understanding such phenomena as substance In Italy people in different regions behaveuse disorders across countries. The second differently when facing the diagnosis of '

idea is the usefulness, or even necessity, of alcoholism, since the regions differ culturallysome common language both for the medical from each other, often significantly. Forand the lay "global village", relevant to such example, denegration and shame linked to theissues as prevention policies, project funding, diagnosis seem to be stronger in Tuscanyand so on. The third is the problem inherent in (central Italy) then in Veneto (north east Italy).translation of the concepts present in different Generally speaking, currently the concept ofnational settings into English language terms and "hazardous or harmful drinking" is morethe difficulties in handling inherent cultural accepted among people than in the past, and thisvalues. shift might explain the growing number of

Room also warns us that moving from the patients entering the alcohol services or alcoholictimes of the tower of Babel to the times of self-help groups..I

Commentaries 229

On the other hand, physicians, and other efforts such as the cross-cultural use of the illS.health professionals seem unable to identify Woody sees the results somewhat differently: theappropriately high-risk drinkers or even paper prompted "an immediate feeling ofalcohol-related problems. Such problems are skepticism" about the project of cross-culturalgenerally overlooked, which may be explained by epidemiology. For him, the potential way out isthe Italian "wet" culture. a shift from "behaviorally based criteria" to

Moreover, while the concept of addiction is "objective findings" from biological markers. Atnever translated into Italian, the tenD alcohol the other end of the spectrum both Partanen anddependence is not common, even in medical Saxena see the implications in more radicaltenninology (perhaps because it is close to tenDs: that the effort to develop "universally

r ~ the widely accepted tenD "drug dependence", valid instruments...has failed" (Partanen), thatwhich refers to patients who are socially the study "helps in removing the exaggeratedunaccepted). and unrealistic sense of complacence that has

AI.l.AMAN Au.AMANI come to surround the use of 'cross-culturally.Centro Alcologico Integrato, valid' diagnostic categories, 'objective' research

Ospedale di Careggi, schedules and 'international' interviewViale Morgagni 85, schedules" (Saxena).

50100 Firenze, Italy. Opinions also vary somewhat among thoseinvolved in the CAR study, although probablynone are at either end of this spectrum of

ALLAMANI, A., BARBERA, G., CALVIANI, L. & TANINI, opinion. We agree with Saxena and otherS. (1994) The treatment system for alcohol-relatedproblems in Italy Alcologia 6 247-252. commentators that the study undercuts any

, , , complacency about our existing tools and study

paradigms for cross-cultural epidemiology inB k t th dr . b d -the alcohol and drug field. Even the relatively

ac 0 e aWIng oar. .gh.& d find. f ul 1 ...R b . R Al k d J Lind A stral uorwar Ing 0 C tura vanauons In

0 In oom, e san ar anca, a. .B tt La S h . d & N S .the threshold of attenUon to a symptom or

enne, ura c ml t onnan artonus ...cntenon greatly complicates the task of a truly

We appreciate very much the generous and comparative epidemiology. The study's resultscollegial spirit in which all of the commentators do suggest that a global perspective will requirehave approached the work of the Cross-Cultural us to go back to the drawing board, not only withApplicability Research (CAR) study. We who respect to epidemiological instruments but alsohave been involved in the study are aware on the with respect to our concepts and diagnoses.one hand of the wealth of data the study has On the other hand, we question Woody'sgenerated, which our paper only begins to suggestion that biological markers offer theaddress, and on the other hand of its flaws and solution to problems in the cross-culturalproblems, some of them intrinsic to any such comparability of diagnoses such as addiction or

~ cross-cultural work. We are also very much dependence. The problem is that, by theiraware of the intensive work the study involved nature, such markers do not measure whatfor the teams of investigators in the nine sites. societies wish to describe by these diagnoses; ifHelzer comments on the "immense commitment we commit our operationalizations to such

, of resources" involved in the study, but it should markers, we tend to leave behind the social and

be recognized that by far the largest such behavioral realities with which Allamani'scommitment in the project was not in tenDS of comment starts-those which bring a person to acentralized funding but rather in tenDS of the clinic's door. Also we do not agree fully withwork, at a hectic pace and without special Partanen's implication that an effort to build acompensation, of the nine site teams. cross-cultural nosology and epidemiology must

There is a broad spectrum of opinion in the start from an analysis of the "socio-economiccommentaries about the implications of the and cultural matrix" of whole societies. Thestudy for cross-cultural epidemiological research. influence of such factors on drinking and drugHelzer might be seen as taking the anchor use behaviour should indeed be taken intoposition at one end: he does not see the results as account, but the effort to characterize andfundamentally threatening the validity of existing measure patterns or pathologies of use remains,

230 Commentaries

in our view, an enterprise with its own logic and building a lift in the tower of Babel vivid andutility. evocative. We would agree with his implication

Drawing on the Italian experience, Allamani that the CAR study did not achieve all the Jremi?ds us that e~en in what seems to an intentions evoked by his metaphor. However, we ~outsIder to be a smgle culture there can be are pleased that the responses of thesubstantially different framings of alcohol or commentators in themselves offer evidence thatdrug problems, affecting classifications and the study has been able to open a few shaftsdiagnoses. The nosological disagreement on between floors and to bring in some gleams of"abuse" versus "harmful use", mentioned by light.Woody, can be seen in a similar light: the ROBIN ROOM,! ALEKSANDAR JANCA,disagreement has basically been between LINDA A. BENNE'IT, fAmerican-based and British-influenced LAURA SCHMIDT & NORMAN SARTORIUS "diagnostic systems-that is, between cultures 1 Research and Development Division, i '

which are, in a global perspective, quite close to Addiction Research Foundation, J.each other. 33 Russell Street, Toronto,

We found Batel's image of the study as Ontario, Canada M5S 2S1.

'~If

~