Validação para o uso no Brasil do Pediatric Quality of Life (PedsQLTM): um estudo envolvendo famílias com crianças e adolescentes acometidos por câncer

Abstract

The aims of the present study were the transcultural adaptation and analysis of the psychometric properties of the Pediatric Quality of LifeTM (PedsQL) Cancer Module for use in Brazil, using a sample of children and adolescents affected by cancer, as well as their families. The study was carried out in 6 steps, following pre-established criteria: 1- Translation; 2- Review by an examining committee (group of specialists); 3- Backtranslation; 4- Pre-test with a sub-sample of the study target population (35 individuals); 5- Field test (190 families); and 6- Study of the psychometric properties of the instruments (reliability and validity). The SPSS 12.0 and Microsoft Excel programs were used for the statistical analysis. The author of the instrument and the institute, which coordinates the validation and utilization of the instrument, accompaniedall phases of the validation process. Regarding semantic equivalence, the instruments proved to maintain equivalency, as 22 (81,5%) of the 27 items presented complete agreement, according to the visual analogue scale. Furthermore, it was observed that the general meaning of the items remained unaltered in 74,0% of the pairs of statements. The points of theoreticalconceptual discussions were considered satisfactory with regard to conceptual equivalences and items. Regarding the analysis of the psychometric properties, test-retest reliability demonstrated the good stability of the instrument according to the analysis of the sub-scale scores (0.69-0.90 for the children/adolescents; 0.71-0.93 for the guardians). The measurement of agreement of the items exhibited values of 0.26-0.85 for the children / adolescents and 0.25-0.87 for theguardians, thereby ranging from regular to almost perfect. For the overall scale, internal consistency demonstrated adequate indices in the comparison between groups (= 0.78-0.80 for children/adolescents, 0.68-0.88 for guardians). In relation to discriminant validity, the subscales pain and hurt, nausea, anxiety over procedures and anxiety over treatment proved capable of distinguishing the groups of individuals "on treatment" and "off treatment (p<0.05). The convergent validity analysis demonstrated significant correlations between the Pediatric Quality of Life Cancer Module scores and those of the Pediatric Quality of Life Generic Core. Individuals in treatment presented lower scores on both instruments, confirming the negative impact of the disease on the quality of life. A low correlation was found between the subscales scores ofthe versions of the children/adolescents and of the guardians. A good level of conceptual equivalence was therefore observed regarding the items and semantics between the version-synthesis in Portuguese and the original, as well as excellent acceptability of the adapted instrument. Furthermore, the Pediatric Quality of Life Cancer Module exhibited adequate properties regarding the reproducibility and validity of the construct, which suggests its use as an additional parameter in studies that assess the impact of neoplasms on the quality of life of children and adolescents. However, the low internal consistency verified in the different subscales suggests the need to carry out further studies with the aim of improving the homogeneity of the instrument.