160 THE NATIONAL MEDICAL JOURNAL OF INDIA VOL. 11, No.4, 1998

Original Articles

WHOQOL-Hindi: A questionnaire for assessing quality oflife in health care settings in India

S. SAXENA, K. CHANDIRAMANI, R. BHARGAVA

ABSTRACTBackground. Quality of life is becoming an important

component of overall assessment in health care settings. However,satisfactory instruments are not available for use in India.

Methods. Qualitative and quantitative work was conductedat the Delhi centre as a part of the WHOQOL (World HealthOrganization Quality of Life) project at 15 centres In developingand developed countries to construct a new quality of lifeinstrument (WHOQOL). The pilot field trial at Deihl was con-ducted on 304 adult subjects using the 236-ltem questionnaire.

Results. Based on the pilot field trial data, item reductioncould be done to develop a 1OO-Item version (WHOQOL-l 00,Hindi). The items are distributed into 4 domains (physical; psy-chological, social and environmental health) and 25 facets. Eachfacet has four items, rated on a five-point scale. The initial psycho-metric properties of this instrument are satisfactory. A 26-itemshort version has also been developed (WHOQOL-Bref, Hindi).

Conclusion. The WHOQOL-l 00, Hindi appears to be asuitable instrument for comprehensively assessingquality of life inhealth care settings. WHOQOL-Bref, Hindi can be used for inter-vention studies including drug trials.

Natl Med J India 1998; 11 : 160-6

INTRODUCTIONDiseases affect human life in aprofound way. They cause prematuredeath resulting in decreased 'quantity' oflife, but more often theycause structural and functional limitations that may seriouslyaffect the 'quality' of life. Death is easier to identify and record;hence mortality has been a standard method for quantifying theimpact of diseases. Quality of life (QOL) has been difficult tomeasure, hence itsuse inhealth care settings has been comparativelyrecent.

Assessment of QOL has several uses in heal th care. It providesa measurement of functioning and well-being rather than ofdiseases and disorders, hence is more comprehensive and com-patible with the WHO's (World Health Organization's) conceptof health. 1It can guide appropriate management strategies-! andalso act as one of the outcome measures for comparing them,"?including drug trials+" QOL assessment focuses attention onaspects of a patient's life beyond symptoms and signs. It thussensitizes health care personnel to look for and correct direct and

All India Institute of Medical Sciences, New Delhi 110029, IndiaS. SAXENA, K. CHANDIRAMANI, R. BHARGAVA

Department of Psychiatry

Correspondence to S. SAXENA

© The National Medical Journal of India 1998

indirect effects of disease and treatment on individuals. QOL alsohelps in policy research including programme evaluation andresource allocation. 14,15

Quality of life assessments have been used most widely in thearea of malignancies 16-18though now their use has become commonin a number of other diseases and conditions including diabetes, 19.20hypertension.Pt-" chronic diseases such as arthritis and bron-chitis,23-25mental illnesses, 26-29substance abuse, 30cerebrovasculardisease," renal disease.v-" head injury ,34and old age. 35

Most instruments used for assessing QOL were constructed inthe developed countries of North America and Europe" and theircross-cultural compatibility has not been demonstrated. Thismakes their direct application in developing countries questionable.Adaptation and translation of these instruments for use in diverse

-cultural and linguistic settings poses serious methodologicalproblems.":" Simultaneous development in different culturesand languages has been suggested as an appropriate method forensuring cross-cultural applicability. 39.41

QOL assessment has been extremely rare in India.42-44One ofthe important reasons for this is non-availability of a suitableinstrument. The subjective Well-being Inventory" has beendeveloped as a QOL instrument in India, but it appears to be moresuitable for the general population than for ill patients seen inhealth care settings.

This paper presents the process involved in developing a QOLinstrument (WHOQOL-Hindi) in collaboration with the WHOQOLGroup, Division of Mental Health and Prevention of SubstanceAbuse, World Health Organization, Geneva. The WHOQOL wasdeveloped simultaneously in 15centres across the world. Commonelements of methodology and results are described in detail in aseries of papers by the WH OQO L Group. 13.46-54The present paperfocuses specifically on the work done at the Delhi Centre, whichled to the development ofWHOQOL-Hindi.

The concept and definitionThere has been lack of clarity in the concept of QOL, with differ-ent groups of workers using the term in widely different ways. 14.55.56There is also some overlap between functional status, healthstatus, subjective well-being, health-related QOL and subjectiveQOL. WHO has defined QOL as 'individuals' perceptions of theirposition in life in the context of the culture and value systems inwhich they live and in relation to their goals, expectations,standards and concerns'. 49It is believed to be a broad conceptincorporating in a complex wayan individual's physical health,psychological state, level of independence, social relationships,personal beliefs and her/his relationships to salient features of theenvironment. The definition highlights QOL as a subjective self-

SAXENA et at. : WHOQOL-illNDI: QUALITY OF LIFE ASSESSMENT

report from the individual which is not based on reports or judg-ment from others (e.g. family members, clinicians). QOL is alsomultidimensional, incorporating positive (e.g. feeling happy,contented, energetic) as well as negative aspects (e.g. not havingpain, sadness, sexual difficulties).

Though consequences of disease affect QOL in a major way,these are not themselves assessed while measuring QOL. Only theeffects of these symptoms on a person's life are assessed. Thismethod of measuring QOL by generic questionnaires is concep-tually more sound because a number of mediating factors determinehow much and what kind of effects a disease will have on aperson's QOL. These factors include personal and environmentalcontextual variables. For example, a knee injury may limit jointmovement. For a young man whose aim in life is to become aprofessional football player, this disability seriously affects hisQOL. But for another person whose profession involves mainlyreading and writing, the same disability affects the QOL to a lesserextent. Hence, a QOL questionnaire aims to assess the extent towhich significant aspects of a person's life have been affected,rather than what symptoms and disabilities are present. Thisconcept of measuring QOL also makes it easier to construct a gen-eric instrument that can be applied to individuals suffering fromillnesses of diverse nature and severity, than to devise an instrumentfor each condition separately.

METHODSDevelopment of WHOQOL

Preparatory work. Each step involved in the development ofWHOQOL was carried out simultaneously in many centres of theworld. These IS field centres (from developing and developedcountries) were selected to take into account the variations indifferent cultures (e.g. level of industrialization, available healthservices, etc.).

The first step consisted of establishing an agreed upon definitionofQOL. Important characteristics ofQOL construct were identified,with special emphasis on cultural relevance. A list of domains andfacets were then drafted by each centre. Six broad domains wereidentified which describe core aspects of QOL cross-culturally(physical, psychological, level of independence, social relation-ships, environment and personal beliefs/spirituality). Domainswere further classified into a number offacets, e.g. psychologicaldomain included facets that assess positive feelings, body image,self-esteem, thinking, memory and concentration, and negativefeelings. Facets describe behaviour (e.g. personal relationships),states of being (e.g. fatigue), capacities (e.g. ability to moveabout) or subjective perceptions of experiences (e.g. pain).

The next stage involved group interviews (focus groups) withpatients, well persons and health personnel in each of the fieldcentres, to evaluate the appropriateness of the facet definitionsdrafted by health professionals and QOL researchers. Based onthese reports, the facet definitions were revised. Questions weredrafted by focus groups. Thus ideas were generated, within eachcentre, as to how and in which form questions relating to QOLshould be asked. Following the focus group work, a questionwriting panel in each centre framed a maximum of six questionsin their own language. This process ensured that the questionswere raised by both health personnel and patients in each of thecentres. A conceptual distinction was made between two types ofquestions: 'perceived objective questions', i.e. global evaluationof functioning (e.g. How well do you sleep?) and 'self-reportsubjective questions' i.e. perceived satisfaction/dissatisfactionwith functioning (e.g. How satisfied are you with your sleep?).

161

The' global question pool' of some 1800 items was reduced toa set of 1000 items, after excluding duplicate and semanticequivalent questions. These questions were rank ordered for eachfacet according to 'how much it tells you about a respondent'sQOL in your culture' . Questions were further eliminated from thecombined ranking for all centres.

Similar care was taken to generate cross-culturally comparableresponse scales. Each item was rated on a 5-point response scaleconcerned with intensity (not at all-extremely), capacity (not atall-completely), frequency (never-always) and evaluation (verysatisfied-very dissatisfied, very good-very poor). Although end-points (e.g. never-always) are relatively universal, ambiguity andcultural variations exist for intermediate responses. Descriptorswere derived to find words/terms falling at 25 %, 50 % and 75 %positions between the two end-points. Details of these exerciseshave been described elsewhere. 53

The pilot field trialThe pilot instrument contained 236 questions covering 6 domainsand 29 facets of QOL. This was based on approximately 8 ques-tions per facet; 4 'perceived objective' questions and 4 'self-report subjective' questions. Some of these questions were genera-ted at the Delhi centre and were in Hindi. Others were translatedinto Hindi according to the WHOQOL translation methodology. 57

In addition, 21 national questions were incorporated in theinstrument (Table I). These questions were selected from a largerlist ofitems developed at the Delhi centre. The items were thoughtto be culturally relevant but could not be included in theinternational pilot version.

A cross-sectional design was used to field-test the pilotquestionnaire on at least 300 subjects in each field centre. At theDelhi centre, 304 adult subjects were drawn from the Hindi-

TABLEI. List of national questions

1. To what extent have you achieved your desired position in life?2. How much do you worry about your health?3. How satisfied are you with your ability to fulfil the sexual needs of

your life partner?4. How satisfied are you with the number of children you have?5. How satisfied are you with your children being sons or daughters?6. Do you feel yourself a burden on others?7. How satisfied are you with the complexion of your skin?8. To what extent are you fed up with life?9. To what extent do you feel guilt and remorse?10. Does expenditure on medicines cause you financial difficulty?11. Does expenditure on tobacco, alcohol or intoxicating drugs cause

you financial difficulty?12. To what extent are you able to express the feelings of your mind in

front of others?13. To what extent do you feel yourself an integral part of your family?14. How satisfied are you with the way responsibilities are shared within

your family?15. To what extent do you remain worried about your family?16. To what extent can you fulfil the needs of your family?17. To what extent do you have to ignore your own needs for the sake of

others?18. How satisfied are you with your capacity to bear sudden major

expenses?19. To what extent are you able to get essential medical services with

the money available with you?20. To what extent does lack of cleanliness in your physical

environment distress you?21. To what extent do your religious beliefs come in the way of your

desired activities?

162

speaking population; the size being based on the required numberof subjects per cell (generated by the sampling quota) needed foranalysis of the pilot data. The sampling quota was as follows:

1. Age (50% ~45years, 50% 2.45 years)2. Sex (50% male, 50% female)3. Health status (250 persons with disease or impairment; 50 well

persons)

Purposive sampling was followed in data collection and thesample was not representative of the patient and healthy populationatthe centre. Of 304 adult subjects, 50.7 % were males. The mean(SD) age of the sample was 40.7 (14.3) years. Two hundred andfifty-three persons comprised the 'unwell' group. Persons withillness or impairment were outpatients and inpatients from medicaland surgical departments of the All India Institute of MedicalSciences (AIIMS), with problems ranging from very mild (e.g.fever.headache, etc.) to severe (e.g. malignancies, renal failure,etc.). Patients with neurological and psychiatric illnesses were notincluded in the sample. Fifty-one' well' respondents were recruitedfrom amongst the relatives of the patients in the hospital and thegeneral community.

The instrument was largely self-administered. The subjectswho were illiterate or had a disability that interfered with self-administration of the questionnaire, were administered theWHOQOL as a structured interview. The subjects were assessedcross-sectionally. Forty-one questions concerning perceivedimportance of facets were also asked in a separate questionnaireto the same respondents.

RESULTSThe data obtained from the New Delhi centre as well as othercentres (n =4802) were analysed with the aim of striking a balancebetween a minimum number of facets/questions and adequatecoverage of key areas across cultures. The selection of items wasbased on the analyses done at the level of individual field centres,summaries across centres and at the level of the pooled globaldata.

The analyses carried out at item, facet and domain level forelimination/selection of items/facets is briefly summarized below:

1. Frequency problem. Items that had two or more adjacentscale points showing < 10%of the response were deleted. Suchproblems were present in 43 items.

2. Reliability problem. Problems of low item-total correlation(<0.40) with their own facets led to the deletion of that item.Eleven items had a reliability problem.

However, these items were eliminated only if they hadproblems in the global analyses or in at least 8 centres.

3. Overlapping. Fifteen items which overlapped conceptuallywith other items were deleted.

4. Multitrait analysis programme (MAP) problems. MAPanalysis" was carried out in order to identify items that hadhigher loadings on another facet rather than on its own predictedfacet. No item showed this pattern. However, 7 items that werehighly loaded (r > 0.40) on more than one facet were identifiedand 6 finally eliminated.

Facet and domain inter-correlations were examined. Sexualactivity of the physical domain was moved to the socialrelationship domain. Facets were found to be fairly independentof each other.

5. Re-analysis for reliability (item-total facet correlation)problem in each centre. Eleven items that showed non-

THE NATIONAL MEDICAL JOURNAL OF INDIA VOL. 11, NO.4, 1998

significant or negative correlations in any centre were excluded(with one or two exceptions). In the Delhi data, 24 items hadlowcorrelations ( <0.40).

6. Validity problem. Twenty-four items were eliminated as theyfailed to distinguish significantly between the 'sick' and 'well'respondents. In the Delhi data, only one item failed to discrim-inate between the two groups and was eliminated.

The analyses did not suggest that the national questions per-formed better than thegeneral questions and hence were eliminated.Five problematic facets (activities as provider/supporter, sensoryfunctions, dependence on non-medical substances, communicationcapacity and work satisfaction) were identified and dropped.

Some facets were excluded because responses to these itemswere highly skewed (activities as provider/supporter, dependenceon non-medical substances, and communication capacity). Further,in facet activities as provider/supporter, nearly half the itemsshowed reliability problems. Two items correlated well withanother facet (personal relationships) and hence were moved tothis facet. Communication capacity also had reliability problems(low item-total correlation and facet-domain correlation). Sensoryfunction was dropped primarily due to low facet-domain correla-tion. Items of work satisfaction facet discriminated poorly between'sick' and 'well' persons. Some of these items also had problemsof low correlations (facet-domain correlations and items-globalQOL correlations).

Though some of the facets (e.g. sensory functions, communi-cation capacity) are likely to be important for assessing QOL, theitems of these facets could not compete with other items in theglobal pool. Hence, they were excluded from the generic versionof the core instrument and for having poor psychometric properties.Nevertheless, these items are being explored in add-on modulesthat are being developed for assessing people with specificconditions/situations/diseases (HIV, cancer, refugees, informal

.care-givers, individuals with communication difficulties, etc.) inwhich the core module does not provide sufficient detail. Forexample, the items of sensory functions were thought to be perti-nent in the elderly group. Items related to activities as provider/supporter may be explored in the 'carer group'.

Thus, the analyses of the WHOQOL pilot data eliminatedsome items and facets, resulting in the development of theWHOQOL field trial form comprising 100 questions grouped intosix domains.

Further, multivariate analyses were carried out to confirm thedomain and facet structure of the WHOQOL-100. It also examinedwhether this universal structure could describe most aspects ofpeople's QOL.

Earlier exploratory and confirmatory factor analyses carriedout on the global data (using structural equation modelling 59)hadshown the comparative fit index (CFI) for single-factor model andsix-domain model to beO.875 andO.888, respectively (CFI> 0.9is considered to be a good degree of fit). Analyses (on new globaldata sets) also provided some evidence that a four-domain solutionmay be the most appropriate (CFI=0.901).54 The four-domainstructure was obtained by merging physical health with the levelof independence; and psychological with spirituality domain.Similar analysis of the Delhi data for four-domain structureindicated the CFI to be 0.876 (Table II). By allowing errorvariances to covary for the environment and social relationshipsdomains, the CFI increased to 0.991, thereby improving the fitindex and indicating acceptability of the model in the Hindi-speaking population. In addition, when each of the domains of

SAXENAet al. : WHOQOL-HINDI: QUALITYOF LIFE ASSESSMENT

TABLEII. Multivariate analyses for confirmation of domainstructureofWHOQOL-lOO, Hindi (n=304)

Domain CFI Regression coefficients

Physical 0.957 0.253Psychological 0.982 0.267Social relationships 0.971 0.219Environment 0.922 0.157-------------------

0.876* 51.9t

* Comparative Fit Index (CFI) at domain levelt Adjusted R2; 'overall quality of life and general health' facet score asdependent variable and domain scores as independent variables.

WHOQOL-lOO, Hindi were analysed separately keeping onlyone factor upon which all facets loaded, the CFI on all the fourdomains ranged between 0.922 (environment domain) to 0.982(psychological domain) (Table II). Further analyses to assessother possible structures on new data sets using WHOQOL-lOO,Hindi are required.

Multiple regression analyses confirmed the four-domain struc-ture (R2=51. 9) ofWH OQO L-l 00, Hindi with each domain contri-buting significantly towards the regression equation (Table II).

The psychometric properties of the WHOQOL-lOO, Hindiwere examined following a series of subscale reliability analyses,and item, facet and domain correlation analyses. Cronbach alphawas found to be moderately high for most of the facets (Table III).All facets correlated significantly with their respective domains.All the facets and domains also correlated significantly with theoverall QO L score ranging from r = 0 .18 (environment) to r = 0.65(psychological). In addition, all items distinguished significantlybetween the 'diseased' and the 'healthy'.

Thus, the series of analyses at different levels refined thestructure of the pilot form into WHOQOL-lOO that had satisfac-tory psychometric properties.

Structure ofWHOQOL-IOOThe WHOQOL-l00, Hindi instrument encompasses 24 facets andone general facet that questions overall QOL and health. Eachfacet is represented by four questions. These facets are groupedinto 4 large domains: physical, psychological, social relationshipsand environment (Table IV).

The period of reporting for all questions is two weeks. This ismentioned in the scale at the beginning of each section (a total ofeight times) to ensure that the respondents kept the time-frame oftwo weeks in mind while responding to each item.

All items are rated on a five-point scale (1-5). The WHOQOL-100, Hindi produces individual facet scores (e.g. positive feelingscore, social support score, etc.), domain scores (e. g. psychologicaldomain score) and a score relating to overall QOL and generalhealth.

Structure of WHOQOL-Bref, HindiAlthough the WHOQOL-lOO provides a comprehensiveassessment of QOL, its length can limit its use. As a result, anabbreviated version (WHOQOL-Bref, Hindi) of 26 items wasdeveloped using data from the field-trial version of the WHOQOL-100. The WHOQOL-Bref contains two items from the overallQOL and general health, and one item from each of the remaining24 facets included in the WHOQOL-l00. (The WHOQOL-IOOand WHOQOL-Bref questionnaires along with details of instruc-tions, headers, question order, response scale and scoring methodare available from the corresponding author.)

163

TABLEIII. Mean scores and Cronbach alphas of facets of theWHOQOL-lOO, Hindi

No. Facets Mean (SO) Cronbach alpha

Overall quality of life and general health 13.3 (3.5) 0.811. Pain and discomfort 11.1 (3.7) 0.722. Energy and fatigue 12.6 (3.9) 0.733. Sleep and rest 14.5 (3.6) 0.764. Positive feelings 11.8 (3.3) 0.635. Thinking, learning, memory and 14.0 (3.1) 0.69

concentration6. Self-esteem 14.2 (3.3) 0.667. Bodily image and appearance 15.2 (3.4) 0.598. Negative feelings 10.7 (3.8) 0.859. Mobility 14.1 (4.5) 0.91

10. Activities of daily living 13.4 (3.6) 0.7411. Dependence on medication or 11.1 (4.0) 0.76

treatment12. Working capacity 13.1 (4.4) 0.9013. Personal relationships 15.1 (3.0) 0.6714. Social support 13.4 (3.5) 0.7715. Sexual activity 13.1 (3.5) 0.6016. Physical safety and security 13.7 (3.5) 0.7617. Home environment 14.4 (3.2) 0.7518. Financial resources 12.8 (3.9) 0.8219. Health and social care; availability 13.6 (2.7) 0.63

and quality20. Opportunities for acquiring new 13.2 (3.3) 0.77

information and skills21. Participation in and new 12.3 (3.3) 0.70

opportunities for recreation/leisure22. Physical environment 13.3 (2.9) 0.62

(pollution/noise/traffic/c1imate)23. Transport 12.8 (3.3) 0.6724. Spirituality/religion/personal beliefs 13.5 (3.5) 0.82

TABLEIV. The WHOQOL-IOO facets and domains

Domain Facets incorporated within domains

Physical health Activities of daily livingDependence on medicinal substances and

medical aidsEnergy and fatigueMobilityPain and discomfortSleep and restWork capacity

Psychological Bodily image and appearanceNegative feelingsPositive feelingsSelf-esteemSpirituality/religion/personal beliefsThinking, learning, memory and concentration

Social relationships Personal relationshipsSocial supportSexual activity

Environment Financial resourcesFreedom, physical safety and securityHealth and social care: Accessibility and qualityHome environmentOpportunities for acquiring new information

and skillsParticipation in and opportunities for reception/

leisure activitiesPhysical environment (pollution/noise/traffic/

climate)Transport

164 THE NATIONAL MEDICAL JOURNAL OF INDIA VOL. 11, NO.4, 1998

WHOQOL-Bref, Hindi produces an aggregate score and fourdomain scores but does not provide individual facet scores.Domain scores produced by the WHOQOL-Brefhave been shownto correlate at around 0.9 with the WHOQOL-1 00 domain scores,and hence provide an excellent alternative to the assessment ofdomain profile using WHOQOL-100. The brief scale is also likelyto be useful in busy clinics and wards since it takes only 5-8minutes to complete. It can also be conveniently used in studieswhich assess QOL longitudinally at several time intervals.

The scale has been shown to have good discriminant validity,sound content validity and good test-retest reliability at severalinternational WHOQOL centres. Despite the heterogeneity offacets included within domains, all domains display excellentinternal consistency. Acceptable comparative fit indices wereachieved when the data from original pilot, field trial and newcentres (CFI =0.906, 0.903 and 0.87 , respectively) were applied tothe four domain structures using confirmatory factor analysis.When three pairs of error variances were allowed to covary (i.e.Pain and dependence on medication, Pain and negative feelings,Home and physical environment) and two items were allowed tocross-load on other domains (i.e. safety on the global domain andmedication negatively on the environment domain), the CFI ofnew data also increased to 0.901.54

DISCUSSIONQuality of life is often regarded as a concept that is too nebulousto be measured reliably with a structured questionnaire and issubject to too much variability across cultures and individuals tohave any useful validity. The WHOQOL project has shown thatboth these assumptions are incorrect. This international projecthas demonstrated that QOL can be conceptualized and defined ina uniform way across cultures. Its constituent core domains andfacets can be assessed using structured questionnaire methodology,and cross-cultural as well as intra-cultural comparisons can bemade. These developments are of major significance to healthcare professionals, who aim not only to prevent and treat diseasesbut also to promote health and QOL.

WHOQOL-Hindi is available in two versions, the long 100-item version (WHOQOL-100, Hindi) and the brief26-item version(WHOQOL-Bref, Hindi). The long version assesses QOL compre-hensively and is a suitable instrument for use in studies' whereQOL is the only or the main determinant. Since it gives a profileof domain scores, effects of a disease or intervention on separatedomains (e.g. physical v. psychological) can be studied. Forexample, WHOQOL-100 may be suitable in a study which com-pares the survival time and QOL of cancer patients on a new, moreaggressive chemotherapy regimen compared to the standardtreatment. On the other hand, WHOQOL-Brefmay be useful instudies which incorporate QOL as one of several variables orwhere multiple assessments over a period oftime are envisaged.For example, a new antihypertensive drug may be as effective inlowering blood pressure as an older drug, but ifQOL is used as oneof the longitudinal outcome variables, patients on the new drugmay score higher/lower, indicating its superiority/inferiority.

QOL assessment clearly does not replace the existing outcomevariables. Mortality and morbidity measurements (e.g. survivalrates, symptom-rating scales, disability questionnaires) are alluseful, but QOL can be an additional outcome variable givinginformation about the individual's life that other variables cannot.Since QOL is a relatively stable state (the period of reporting inWHOQOL is 2 weeks), it is not a suitable assessment in studieswhere short term effects over hours or days are being studied.

Besides use in research studies, WHOQOL can be used inclinical practice to determine the impact of disease, disability andtreatment modalities on patients. Specific modules that are beingdeveloped as a supplement to the core instrument are likely to leadto a greater understanding of the diverse effects of diseases experi-enced by the patients on their lives. They may also guide towardsbetter and more comprehensive management strategies, includingpsychosocial interventions.

Incontrast to many other QOL instruments, WH OQOL includesa domain on environment. This is considered necessary as environ-ment plays a major role in determining health states, mediatingdisease pathogenesis and limiting or facilitating access to healthcare. Like all other domains in WHOQOL, environment domainis also assessed by a subjective self-report with the underlyingbelief that even if subjective reports are at variance with objectivereality, it is the former that determines QOL.

Since WHOQOL has been developed simultaneously in manycentres across the world and the different language and nationalversions are compatible and comparable with each other, multi-centre international studies can be conducted using this instrument.In India, besides Hindi, a version in Tamil is available and a ver-sion in Kannada is being developed. WHOQOL cannot be used inother languages after simple translation, but fresh versions can bedeveloped (e.g. for other regions of India) using the WHOQOLmethodology. 54 This includes preparatory qualitative researchfollowed by a pilot field trial to ensure cultural and linguistic equi-valence of different versions.

Although on initial assessment, the psychometric properties ofWHOQOL-100 and WHOQOL-Bref appear satisfactory, morestudies are needed to determine the reliability, validity and res-ponsiveness to change of these questionnaires. These studiesshould preferably be from diverse clinical populations to ascertainthe suitability ofWHOQOL-100, Hindi and its goodness of fit inthese settings. Large general population studies are also needed toestablish norms for comparison purposes.

WHOQOL-Brefprovides a shorter alternative to the assessmentof QOL based on the WHOQOL domain profiles. However,analyses on varied data sets are required for replication of selectionof items and their' goodness of fit' at the Delhi centre.

ACKNOWLEDGEMENTSThis work was done as a part of the WHOQOL Project of the Division ofMental Health and Prevention of Substance Abuse, World Health Organiza-tion, Geneva. The authors acknowledge constant help extended by theWHOQOL Group and particularly by Dr John Orley and Professor MickPower. Drs Sanjay Agrawal, Pratap Sharan, Anju Dhawan, Hemraj Pal andDhanesh Kumar assisted in the work at New Delhi.

REFERENCESI World Health Organization, World Health Organization Constitution. In: Basic

Documents. Geneva: World Health Organization. 1948.2 CroogSH. LevineS, Testa MA, Brown B, BulpiuCJ. JenkinsCD, et al, The effects

of antihypertensive therapy on the quality of life. N EngLJ Med 1986;314: 1657-64.3 Gelber RD, Goldhirsch A, Cavalli F. Quality-of-life adjusted evaluation of adjuvant

therapies for operable breast cancer: The International Breast Cancer Study Group.Ann Intern Med 1991 ;114:621-8.

4 Ganz PA, Lee JJ, Siau J. Quality of life assessment: An independent prognosticvariable for survival in lung cancer. Cancer 1991 ;67:3131-5.

5 McClellan WM, Anson C. Birkeli K, Tuttle E. Functional status and quality oflife:Predictors of early mortality among patients entering treatment for end-stage renaldisease. J Clin Epidemioll991 ;44:83-9.

6 Coates A. Thomson D. McLeod G RM, Hersey P, Gill PG, Olver IN. et al. Prognosticvalue of quality of life scores in a trial of chemotherapy with or without interferonin patients with metastatic malignant melanoma. Eur J Cancer I 993;29A: 1731-4.

7 Awad AG, Hogan TP. Subjective response to neuroleptics and the quality of life:Implications for treatment outcome. Acta Psychiatr Scand 1994;89 (Suppl 380):27-32.

SAXENA et at. : WHOQOL-HINDI: QUALITY OF LIFE ASSESSMENT

8 Johnson JR, Temple R. Food and drug administration requirements for approval ofnew anti-cancer drugs. Cancer Treat Rep 1985;69: 1155-9.

9 Guyatt GH, Veldhuyzen Van Zanten SJ, Feeny DK, Patrick DL. Measuring qualityof life in clinical trials: A taxonomy and review. CMAJ 1989;140:1441-8.

10 Gerin P, Dazord A, Boissel JP, Hanauer MT. Assessment of quality of life intherapeutic trials. In: Strauch G, HussonJM (eds). Recent trendsin clinicalpharma-cology. Paris:INSERM 1990;186:143-63.

II Osoba D. The Quality of Life committee of the clinical trials group of the NationalCancer Institute of Canada: Organization and functions. Qual Life Res 1992;1:211-18. .

12 Anonymous (Editorial). Quality of life and clinical trials. Lancet 1995;346: 1-2.13 Saxena S, Orley J. Quality of life assessment: The World Health Organization

perspective. Eur Psychiatry 1997;12 (SuppI3):263S-266S.14 Patrick DL, Erickson P. Health status and health policy: Quality of life in health

care evaluation and resource allocation. New York:Oxford University Press, 1993.15 Fox-Rushby JA. The relationship between health economics and health-related

quality oflife. In: Orley J, Kuyken W(eds). Quality of life assessment: Internationalperspectives. Berlin, Heidelberg, New York:Springer- Verlag, 1994:61-74.

16 Fallowfield L. Measuring the quality oflife in Breast Cancer. In: Orley J, KuykenW (eds). Quality of life assessment: International perspectives. Berlin, Heidelberg,New York:Springer-Verlag, 1994:109-27.

17 Maguire P, Selby P. Assessing quality oflife in cancer patients. Br J Cancer 1989;60:437-40.

18 Gotay CC, Korn EL, McCabe MS, Moore TD, Cheson BD. Quality-of-life assess-ment incancer treatment protocols: Research issues inprotocol development. J NailCancer Inst 1992;84:575-9.

19 Kaplan RM, Hartwell SL, Wilson DK, Wallace JP. Effects of diet and exerciseinterventions on control and quality of life in non-insulin-dependent diabetesmellitus. J Gen Intern Med 1987;2:220-8.

20 de Weerdt I, Visser AP, Kok GJ, de Weerdt 0, van der Veen EA. Randomizedcontrolled multicentre evaluation of an education programme for insulin-treateddiabetic patients: Effects on metabolic control, quality oflife, and costs of therapy .Diabet Med 1991;8:338-45.

21 SiegristJ, Matschinger H, Motz W. Untreated hypertensives and their quality of life.J Hypertension 1987;5:Suppl SI5-S20.

22 Testa MA, Anderson RB, Nackley JF, Hollenberg NK. Quality of life andantihypertensive therapy in men: A comparison of captopril with enalapril. Thequality of life. Hypertensive Study Group. N Eng/J Med 1993;328:907-13.

23 Deyo RA. Measuring the quality of life of patients with rheumatoid arthritis. In:Walker SR, Rosser RM (eds). Quality of life: Assessmens and application. Lancaster:MTPPress.1988:205-22.

24 Skevington SM. Quality oflife assessment in arthritis. In: Orley J, KuykenW (eds),Quality of life assessment: International perspectives. Berlin, Heidelberg, NewYork:Springer-Verlag, 1994:151-60.

25 Moody L, McCormick K, Williams A. Disease and symptom severity, functionalstatus, and quality oflife in chronic bronchitis and emphysema (CBE). J Behav Med1990;13:297-306.

26 Lehman AF, Ward NC, Linn LS. Chronic mental patients: The quality oflife issue.Am J Psychiatry 1982;139: 1271-6.

27 World Health Organization. Report of the meeting on quality of life in schizophrenia.(MNH/PSF/95.7) Geneva: World Health Organization, 1995.

28 Lehman AF. Measures of quality of life among persons with severe and persistentmental disorders. Soc Psychiatry Psychiatr Epidemioll996;31:78-88.

29 Orley J, SaxenaS, HerrmanH. Quality of life and mental illness: Reflections fromthe perspective of the WHOQOL. Br J Psychiatry 1998;172:291-3.

30 Chandiramani K. Substance abuse and quality of life. In: Rajkumar S, Kumar S(eds). Quality of life in health: A modern concern. Madras:Madras Medical College,1994:79-85.

31 Baird AD, Adams KM, Ausman JI, Diaz FG. Medical, neuropsychological, andquality of life correlates of cerebrovascular disease. Rehabil Psychol 1985;30:145-55.

32 Oldenburg B, Macdonald GJ, Perkins RJ. Prediction of quality oflife in a cohort ofend-stage renal disease patients. J Clin EpidemioI1988;41:555-64.

33 DenistonOL, Carpentier-Alting P, KneisleyJ, HawthorneVM, Port FK. Assessmentof quality of life in end-stage renal disease. Health Serv Res 1989;24:555-78.

34 Klonoff PS, Costa LD, Snow WG. Predictors and indicators of quality of life inpatients with closed-head injury. J c/in up NeuropsychoI1986;8:465-9.

165

35 Samuel R, Rajkumar S, Prabhu R. Quality of life of the elderly. In: Rajkumar S,KumarS (eds). Quality of life in health: A modern concern. Madras: Madras MedicalCollege,I994:96-105.

36 Bowling A. Measuring health: A review of quality of life measuremen: scales.Philadelphia: Open University Press, 1991.

37 Sinha D. The family scenario in a developing country and its implications for mentalhealth: The case of India. In: Dasen PR, Berry JW, Sartorius N (eds). Health andcross-cultural psychology: Toward applications. Newbury Park, California:Sage,1988:48-70.

38 Hunt SM. Cross-cultural comparability of quality of life measures. Drug Info J1993;27:395-400.

39 Kuyken W, Orley J, Hudelson P, Sartorius N. Quality of life assessment acrosscultures. Int J Menial Health 1994;23:5-23.

40 Patrick DL, Wild OJ, Johnson ES, Wagner TH, Martin MA. Cross-cultural validationof quality oflife measures. In: Orley J, Kuyken W (eds). Quality oflife assessment:lniernational perspectives. Berlin, Heidelberg, New York:Springer- Verlag, 1994:19-32.

41 Bullinger M. Ensuring international equivalence of quality of life measures:Problems and approaches to solutions. In: Orley J, Kuyken W (eds). Quality of lifeassessment: lnsernational perspectives. Berlin, Heidelberg, New York:Springer-Verlag, 1994:33-40.

42 Sundar S, Rajan AG, Somani PN, Kumar K. The effects of antihypertensive agentson the quality of life in Indian hypertensives. Acta Cardioll991 ;46:227-35.

43 Saxena S. Quality oflife assessment in India: Some conceptual and practical issues.In: Raj Kumar S, Kumar S (eds), Quality of life in health: A modern concern.Madras:Madras Medical College, 1994:106-14.

44 Rajkumar S, Kumar S, Quality of life in health: A modern concern. Madras: MadrasMedical College, 1994,

45 Nagpal R, Sell H. Subjective well-being, SEARO-Regional Health Papers No, 7,NewDelhi:WHO,SEARO, 1985.

46 The WHOQOL Group. Studyprotocol for the World Health Organization project todevelop a quality of life assessment instrument (WHOQOL), Qual Life Res 1993;2:153-9,

47 The WHOQOLGroup. The development of the World Health Organization qualityoflife assessment instrument (the WHOQOL). In: Orley J, Kuyken W (eds), Qualityof life assessment: lmemaiionalperspeaives. Berlin, Heidelberg, NewYork:SpringerVerlag,I994:41-57.

48 WHOQOLGroup. Development of the WHOQOL: Rationale and currentstatus.lnIJ Mental Health 1994;23:24-56.

49 WHOQOL Group. The World Health Organization quality of life assessment(WHOQOL): Position paper from the World Health Organization, Sac Sci Med1995;41(10): 1403-9.

50 Saxena S, Orley J, (on behalf of the WHOQOL Group), Quality of life and itsmeasurement: The WHOQOL approach (in Italian). Psicoterapia Cognitiva eComponamentale. 1996;1:5-12.

51 Szabo S, (on behalf of the WHOQOL Group). The World Health OrganizationQuality of Life (WHOQOL) assessment instrument. In: Spilker B (ed). Quality oflife and pharmacoeconomics in clinical trials. Philadelphia: Lippincott-Raven,1996:355-62.

52 World Health Organization. Programme on menial Health: WHOQOL measuringquality of life. (MNH/PSF/97.4). Geneva: WHO,I997.

53 Szabo S, Orley J, Saxena S, (on behalf of the WHOQOL Group). An approach toresponse scale development for cross-cultural questionnaires. Europ Psychologist1997;2:270-6.

54 WHOQOL GROUP. WHOQOL user manual (draft). Geneva: World HealthOrganization, 1998.

55 Bergner M, Quality of life, health status, and clinical research, Med Care 1989;27(SuppI3):SI48-S156.

56 Gill TM, Feinstein AR. A critical appraisal of the quality of life measurements.JAMA 1994;272:619-26.

57 Sartorius N, Kuyken W, Translation of health status instruments. In: Orley J,Kuyken W (eds).Quality of life assessment: Intemational perspectives. Berlin,Heidelberg, New York: Springer-Verlag,1994:3-18.

58 Hays RD, Hayashi T, Carson S, Ware JE. User's guide for the Multitrait AnalysisProgram (MAP). Santa Monica, California: The Rand Corporation, N-2786-RC,1988.

59 Bentler PM, Wu EJC. EQSfor Windows: Users Guide, California: Encino, Multi-variate Software, Inc. 1995.