Field testing of the psychometric properties of the oral symptom assessment scale-Portuguese version: a methodological study

Abstract

Abstract Background Oral symptoms are common among patients with different types of cancer and cancer sites. Patient report is the gold-standard in symptom management, and no multi-oral symptom assessment instrument is yet available for the Portuguese patients with advanced cancer. This study aimed to translate the Oral Symptom Assessment Scale (OSAS) into Portuguese, evaluate the structural validity of the Portuguese version and its psychometric properties. Methodology A methodological, quantitative, descriptive, multicentred, cross-sectional study was conducted. The translation, cross-cultural adaptation and structural validity procedures followed COSMIN guidelines. The original instrument was translated to Portuguese, and the final translated version was pre-tested in a sample of 30 patients with advanced (locally or metastatically) and incurable cancer, and an estimated survival of months. The final translated version was then administered to a sample of 101 patients with advanced (locally or metastatically) and incurable cancer, and an estimated survival of months, attending oncology outpatient consultations or receiving hospital-based palliative care. Three hypothesised factor structures were analysed using confirmatory factor analysis, in accordance with COSMIN guidelines. Results The Portuguese version of the OSAS was well understood, accepted, and considered relevant and feasible by the participants. A three-factor model with 17 items was retained, demonstrating good overall fit ( $$\:scaled\:CFI=0.979$$ ; $$\:SRMR=0.067$$ ; $$\:RMSEA=0.060$$ , $$\:CI\:95\%\:[0.033,\:0.082]$$ $$\:p\:=\:.244$$ ). No inadmissible parameter estimates were identified in the three-factor model. All factor loadings were statistically significant and within acceptable ranges (0.57 to 0.98). Conclusions Evidence supports a three-factor structure model, as the most appropriate representation of the instrument’s latent structure in this population, consisting of changes in oral functions, changes in oral mucosa, and dental changes. The Portuguese version of the OSAS demonstrated good structural validity and reliability and appears suitable for research and clinical use in Portuguese patients with advanced cancer.