Health research is an essential tool for advancing health and development, both for applying available solutions and for creating new knowledge for unsolved problems. Moreover, it is an important link to equity in development, as it promotes acceleration of health improvements, reducing health disparities worldwide.1,2

Health research is particularly imperative in low- and middle-income countries (LMICs) because these are where the great majority of the world's population, especially children,3 live and where health improvements must be achieved with limited resources.1,2 In nations with emerging economies, improvements in public health result in diminishing importance of infection and malnutrition as causes of death in childhood. For instance, as in high income countries, in Brazil cancer currently represents the most prevalent cause of death by disease in children and adolescents aged 1 to 19 years old, with 12,500 new diagnoses every year.4 Delivering cost-effective care to this sub-population is an obvious high priority and attaining this goal in low- and middle-income countries (LMICs) has been demonstrated, although the issue of affordability remains a challenge.5

To evaluate the quality and quantity of health care services, it is essential to measure health-related quality of life (HRQL), both in general and clinical populations.6 The HRQL is a general, multi-dimensional and dynamic construct influenced by physical, psychological and social functioning.6 There are few methods available for the HRQL measurement in preschool children, considered to be the most important phase of human development,7 and they are all health profiles.8 By definition, these do not provide utility scores. Instruments which use a preference-based approach have the advantages of providing utility scores for individual subjects. Preference/utility scores are measured, unlike values, under conditions of uncertainty, as is the case of health. Utilities are especially useful to integrate effects of both morbidity and mortality, or, in summary, measures, such as quality-adjusted life years (QALYs). Furthermore, utilities and QALYs are important metrics in cost-effectiveness/utility analyses, which inform decision-making professionals about the distribution of scarce health care resources.9 The measurement of the HRQL can be very challenging in clinical pediatrics, especially in preschool-aged children, given the rapidity of developmental changes and the need for proxy assessment.10 Generic multi-attribute utility measures are suitable for use in the general population and in groups with specific clinical conditions. Those developed for older children are the 16D 11 and 17D,12 the Health Utilities Index Mark 2 (HUI2),13 the Health Utilities Index Mark 3 (HUI3),14 the European Quality of Life 5-Dimension Youth version (EQ-5D-Y),15 the Assessment of Quality of Life 6-Dimension (AQoL-6D) 16 and the Child Health Utility 9D (CHU-9D).17 Only HUI2, HUI3 and EQ-5D-Y can be used by proxy respondents who are commonly parents, in the case of young children.

The Comprehensive Health Status Classification System-Preschool (CHSCS-PS) is a multi-dimensional tool for measuring comprehensive health status in young children. The CHSCS-PS evolved from the well-established Health Utilities Index (HUI) systems for subjects 5 years of age and older.18 The CHSCS-PS has been described previously 19 and its measurement properties assessed in studies of both community and clinical populations, with reliability, content and construct validity, and inter-observer agreement established in a large community sample,20 as well as in children with cancer.10 The CHSCS-PS forms the basis of the generic preference-based instrument Health Utilities-Preschool (HuPS), applicable to children 2 to 5 years of age.21

Our group of investigators is focused on cancer in children and has recognized the need for a utility-based HRQL instrument for use in this population in Brazil for over a decade.22 The translation of some HUI instruments into Brazilian Portuguese has been undertaken previously by members of our group.22,23 The remaining need is for a HUI-based instrument suitable for preschoolers, which will be enabled by the earlier work. It is important to recognize that the incidence of cancer in childhood is highest in the first 5 years of life.24 Thus, this study performed a formal linguistic and cultural translation of the HuPS classification system from its original form (English) to Brazilian Portuguese.

The HuPS classification system has 8 domains/attributes of health, each with 4 to 6 levels of function. One level for each domain/attribute constitutes an 8-element vector which describes a comprehensive health state. A compound “disability score” can be calculated as the sum of the levels.23 The classification system was translated and adapted by a team of 6 qualified professionals, being 2 females and 4 males. Forward and backward translations were performed by native Brazilian Portuguese speakers with at least a bachelor's degree, who were also fluent in the English language. The translation followed recommended protocols from published guidelines 25,26 and the corresponding author (KV) served as the project manager. The translation and cross-cultural adaptation comprised the following steps:

1) two independent forward translations of the original document, from English to Brazilian Portuguese (T1 and T2: authors MNV and BFB);

2) a third person (author KV) combined the two translated versions into a Brazilian Portuguese consensus version;

3) two independent back translations of the Brazilian Portuguese consensus version to the original language (T3 and T4: authors RCV and TMB) and the translators did not have access to the original documents or any related information other than the translated texts;

4) review and compiling of the English back translations by a native English speaker and experienced quality-of-life researcher (author RDB, who is a co-developer of the HUI), with a later comparison to the original document;

5) final revision and formatting of the Brazilian Portuguese version;

6) linguistic validation, as was undertaken in the translation of the HuPS into Canadian French, and; 27

7) the whole process was supervised by another of the developers of the HUI family of instruments (William Furlong), as was done previously.23

The linguistic validation was carried out through face-to-face cognitive debriefing interviews with a convenience sample of employees at a university hospital in São Paulo, Brazil who were parents of healthy children aged 2 to 5 years old. If the parents had more than one child in this age group, they were asked to think about their youngest child when providing feedback on their understanding of the instrument. All participants were asked in Brazilian Portuguese about their level of education and to read a copy of the Brazilian Portuguese HuPS. Then, they were asked to answer a series of open-ended questions about the instructions, questions and response options. The questions were: ‘Do you understand the instructions of this instrument? Do you understand the response options? Did you have any difficulties with the instrument? If so, how would you rephrase the instructions/responses? Any other comments or suggestions about the instrument?’. The answers provided orally by the participants were summarized by the interviewer (author KV). Suggestions and comments related to the translated instrument were reported and reviewed by the project managers (authors KV and RDB). Proposed changes to the instrument will be adopted when a consensus is established with the original HuPS developers, as was undertaken with the Canadian French linguistic validation.27 Information on the use of the HuPS instrumentation, including scoring instructions, may be obtained from the Health Utilities Inc. at http://healthutilities.com.

Table 1A shows the key word discrepancies between translators 1 and 2 and finalized words from the forward translation process. Table 1B shows the equivalent results from the back translation process. These exercises resulted in initial disagreements for 12 of 845 words (1.4%) for the forward translation and 6 of 812 words (0.7%) for the back translation, the denominator being the sum of nouns, verbs, adjectives and adverbs in the classification system. The discrepant terms were reconciled by the translators and the project manager, who reached a consensus on which were the most appropriate Brazilian Portuguese words, semantically and culturally, within the specific contexts. The final Brazilian Portuguese version of the HuPS is shown in Table 2.

The linguistic validation was conducted with 21 participants, 2 men and 19 women, of whom 6 (29%) had completed secondary education, 11 (52%), higher education, and 4 (19%), a graduate degree. All participants (100%) reported that they clearly understood the instructions and responses, did not have difficulties understanding the Brazilian Portuguese that was used and some made minor comments and suggestions to further improve the understanding of the instrument.

Our paper describes the process used to translate and adapt important HRQL tools, based on the HUI, for use with preschool-aged children in Brazil. The widely utilized HUI2 and HUI3 instruments are appropriate for assessing the health of older children (and adults).19 Indeed, the HUI3 is the most frequently used generic, preference-based measure of HRQL in children.28 Instruments in the HUI inventory include 16 versions of HUI questionnaires in the English language and many of these are available in other languages, including Brazilian Portuguese, for the measurement of the HRQL in survivors of cancer in childhood who were old enough (aged more than 13 years) to self-complete the questionnaires and, thus, to self-report on their health status.22,23 The quality of the translation was assessed by acceptability, face validity, inter-rater reliability and convergent validity using patients, as well as physician and nurse proxy respondents.22 That study was conducted by the Grupo Especializado em Pediatria dos Efeitos Tardios do Tratamento Oncologico (GEPETTO) at the Centro de Tratamento e Pesquisa of Hospital AC Camargo in São Paulo. The GEPETTO is a multi-disciplinary team established to follow, monitor and treat long-term cancer survivors. For the study, these included survivors of acute leukemias, malignant lymphomas, malignancies of the central nervous system and extra-cranial solid tumors. Of the 138 subjects, one-third reported cognitive disability or pain and one quarter reported problems with vision, speech or emotion, results similar to those reported in other countries, including those in Central and South America, using the HUI instruments.23 The CHSCS-PS (a forerunner of the HuPS) in English has been used in a community sample of 4,546 preschool children in the Netherlands,20 in which good discriminant validity was demonstrated, as well as among infants with very low birth weight,29 children with cerebral palsy 29 and children with solid tumors.10,30 A strength of this instrument is that it can identify almost 40 million different health states, based on a factorial of the number of domains/attributes and the number of levels.22

The translation and cultural adaptation of the HuPS classification system into Brazilian Portuguese is the first step in the validation of the instrument in Brazil. This will make it possible to measure the HRQL of preschool-aged children in Brazil with a generic preference-based instrument. This instrument will be useful in the other 9 countries in which Portuguese is an official language, the majority of which are in Africa, and it will be especially useful for preschoolers with cancer, as the incidence of malignant disease in children is highest in the first 5 years of life.24