Original Article
Translating Health Status Questionnaires and Evaluating Their Quality: The IQOLA Project Approach

https://doi.org/10.1016/S0895-4356(98)00082-1Get rights and content

Abstract

This article describes the methods adopted by the International Quality of Life Assessment (IQOLA) project to translate the SF-36 Health Survey. Translation methods included the production of forward and backward translations, use of difficulty and quality ratings, pilot testing, and cross-cultural comparison of the translation work. Experience to date suggests that the SF-36 can be adapted for use in other countries with relatively minor changes to the content of the form, providing support for the use of these translations in multinational clinical trials and other studies. The most difficult items to translate were physical functioning items, which used examples of activities and distances that are not common outside of the United States; items that used colloquial expressions such as pep or blue; and the social functioning items. Quality ratings were uniformly high across countries. While the IQOLA approach to translation and validation was developed for use with the SF-36, it is applicable to other translation efforts.

Introduction

With the growing international collaboration in clinical research, the need for international instruments for outcomes assessment has also increased 1, 2. Although patient-based measures of health status, or health-related quality of life, have been developed in different countries, the majority of the established measures stem from the Anglo-American literature 3, 4. Owing to the complexity of the constructs involved, these measures can not be assumed to be a priori invariant to cultural diversity 5, 6. Before these instruments can be used internationally, the source instruments must be culturally adapted. The issue of how to transfer measures from the national to the international domain has received longstanding attention in social and anthropological research, as well as in psychiatry and cross-cultural psychology 7, 8. As Hui and Triandis [9] have stated, the production of cross-culturally comparable versions of an instrument involves operational, scalar, functional, and metric equivalence. Necessary steps in attaining these goals pertain to: (1) the translation of the instrument, including assessment of the quality of the translation; (2) the validation of the instrument, including the psychometric criteria (i.e., testing of scaling assumptions, reliability, validity, and responsiveness); and (3) the norming of the instrument using representative national samples.

The International Quality of Life Assessment (IQOLA) project group determined that no one qualitative or quantitative step is sufficient to ensure a valid translation. Thus, it developed a three-stage process designed to produce cross-culturally comparable translations of the SF-36 Health Survey. In brief, the process included: (1) rigorous translation and evaluation procedures to ensure conceptual equivalence and respondent acceptance, (2) formal psychometric tests of the assumptions underlying item scoring and construction of multiitem scales, and (3) examination of the validity of the scales and the accumulation of normative data and other interpretation guidelines.

The focus of this article is the translation process; IQOLA methods for testing scaling assumptions, evaluating validity, and norming the SF-36 are discussed in detail elsewhere 10, 11. While literature is available on the conduct of the translation process itself, at the time the IQOLA project began in 1991, relatively little had been written in the health status literature about how to ensure the quality of translations of health status questionnaires and their cross-cultural comparability 12, 13. Likewise, guidelines on how to psychometrically evaluate health status instruments internationally were rare [14], although validation and norming procedures for health measures had been explicitly described 15, 16. Thus, at the onset of the IQOLA project, the research team developed a series of procedures, guidelines, and criteria to ensure cross-culturally comparable translations of the SF-36.

The translation protocol was developed in a series of international meetings in 1991 and 1992, involving National Principal Investigators from Australia, Belgium, Canada, France, Germany, Italy, Japan, the Netherlands, Spain, and Sweden; the IQOLA Project Principal Investigator and other researchers from the Health Assessment Lab; and representatives from Mapi Research Institute. Experience in several countries in which the SF-36 had already been translated, primarily in Sweden [17] and also in France and Italy, was used in developing the translation procedure. The protocol also benefited from the experience of the European Organization for Research and Treatment of Cancer (EORTC) Quality of Life Study Group [18] and researchers who had translated the Nottingham Health Profile [19].

After summarizing the translation protocol, this article presents results of the difficulty and quality ratings that were conducted as part of the translation process. Results from another part of the translation process, the Thurstone scaling exercise, are discussed by Keller et al. [20] elsewhere in this issue. Additional detail on the cross-cultural comparisons of the SF-36 translations is provided by Wagner et al. [21].

Section snippets

Translation Process

The first step of the translation process involved forward translations of the original U.S.-English questionnaire into the language of the target country by at least two translators who were native speakers of the language into which the SF-36 was to be translated (Figure 1). Translators had experience in questionnaire translation but were not familiar with the SF-36 (translators 1 and 2, native speakers of the target language). After explanation of the process by the National Principal

Interrater Agreement in Difficulty and Quality Ratings

The rate of agreement between the two raters in each country varied considerably (Table 2). The percentage of interrater agreement per country in item difficulty ranged from 42% (Spain) to 81% (Netherlands), with a median of 69%. Rater agreement generally was higher for the quality ratings than for the difficulty ratings, for countries in which both difficulty and quality ratings were available. For clarity ratings, the percentage of ratings in agreement between the two raters ranged from 78%

Discussion

Although different approaches to translation are possible (e.g., focusing more on focus or lay groups in developing and testing of a health instrument 19, 22, 23), the IQOLA approach incorporated a number of steps toward an extensive cross-cultural evaluation. Strengths of the approach included independent empirical tests of translation quality and comparison of translations across countries, as well as the international comparison of response scaling values. Changes that could be made to these

References (29)

  • S. Walker et al.

    Quality of LifeKey Issues in the 1990s

    (1992)
  • M. Bullinger et al.

    Developing and evaluating cross-cultural instruments from minimum requirements to optimal models

    Qual Life Res

    (1993)
  • S.M. Hunt

    Cross-cultural comparability of quality of life measures

    Drug Info J

    (1993)
  • Cited by (770)

    View all citing articles on Scopus
    View full text