Quality of life in individuals with neurofibromatosis type 1 associated cutaneous neurofibromas: validation of the Dutch cNF-Skindex

• Published in: Journal of patient-reported outcomes 2024-05, Vol.8 (1), p.52-8, Article 52
• Main Authors: Dhaenens, Britt A. E., van Dijk, Sarah A., Fertitta, Laura, Taal, Walter, Wolkenstein, Pierre, Oostenbrink, Rianne
• Format: Article
• Language: English
• Subjects: Adolescent; Adult; cNF-Skindex; Cutaneous neurofibroma; Discriminant analysis; Female; Genetic disorders; Humans; Male; Medicine; Medicine & Public Health; Middle Aged; Netherlands; Neurofibroma - psychology; Neurofibromatosis 1 - psychology; Neurofibromatosis type 1; Patient-reported outcome measure; Patients; Psychometrics - instrumentation; Psychometrics - methods; QoL; Quality of life; Quality of Life - psychology; Quality of Life Research; Questionnaires; Regression analysis; Reproducibility of Results; Skin Neoplasms - psychology; Surveys and Questionnaires; Young Adult
• ISSN: 2509-8020; 2509-8020
• DOI: 10.1186/s41687-024-00732-w

Background Almost all patients with Neurofibromatosis type 1 (NF1) develop cutaneous neurofibroma (cNF), benign dermal tumours that have a large impact on the patient’s Quality of Life (QoL). The French cNF-Skindex is the first questionnaire to specifically assess cNF-related QoL in patients with NF1. We aimed to adapt and validate a Dutch version of the cNF-Skindex. Methods The questionnaire was translated using forward and backwards translation, and subsequently administered to a sample of 59 patients on two separate occasions. Feasibility was evaluated by the presence of floor/ceiling effects. Reliability was assessed by evaluating internal consistency and test-retest reliability, by calculating Cronbach’s alpha and Spearman’s rank correlation coefficients. The EQ-5D-5L and SF-36 were used to evaluate convergent validity, using Spearman’s rank correlation coefficients. An exploratory factor analysis was performed to study the data’s internal structure. Multivariable linear regression was used to model the relationship between patient characteristics and the cNF-Skindex. Results The Dutch cNF-Skindex demonstrated excellent feasibility and reliability (Cronbach’s alpha 0.96, test-retest correlation coefficient 0.88). Convergent validity was confirmed for the EQ-5D-5L and relevant SF-36 scales. All items and subdomains from the original questionnaire were confirmed following exploratory factor analysis. The patient characteristics included in the multivariable linear regression were not significantly associated with the cNF-Skindex score. Conclusions The Dutch cNF-Skindex displayed excellent psychometric properties, enabling use in the Netherlands.