Reliability and validity of Turkish translation of the pediatric sleep questionnaire

OBJECTIVE: Aim of this study was to investigate the validity and reliability of the Turkish translation of Pediatric Sleep Questionnaire (PSQ)-the extended version. The Pediatric Sleep Questionnaire (PSQ) has two versions: a shorter one, which has been validated for sleep-related breathing disorders, and an extended version, which deals with a wider range of sleep disturbances. METHOD: After translation and back-translation of the PSQ, it was filled in by the parents of 99 patients with Attention Deficit Hyperactivity Disorder (ADHD), 34 patients with Upper Airway Resistance and Snoring (UARS), and 42 healthy controls. Cronbach alpha was computed as a measure of internal consistency along with test-retest reliability and item-total score correlations. We used principal components analysis and varimax rotation to investigate the factor structure. The Receiver Operator Characteristic (ROC) curves were used to determine cut-off points, sensitivity, specificity, and negative and positive predictive values. RESULTS: Cronbach alpha coefficient was 0.77 for the whole questionnaire. Similar to the original PSQ study, we found four factors, namely "Snoring," "Behavioral Problems," "Breathing Problems," and "Other Problems.". However, different items loaded to these factors. While the first three factors were reliable and valid, this was not the case for the last factor. ADHD patients had higher "Breathing Problems" score than the controls, while all sleep related problems were higher in the UARS group than the other two groups. Cut-off points obtained from the ROC curves indicated that when a PSQ total score of 6 points was used as a cut-off, UARS group and the controls could be discriminated easily. In order to discriminate the ADHD and UARS groups, using the "Total Sleep Problems Score", which did not include the "Behavioral Problems" factor, resulted in greater specificty and sensitivity. DISCUSSION: These results indicated that the Turkish version of the PSQ was a valid and reliable measure, that the PSQ total score could discriminate patients with primary sleep disorder from healthy controls, and that the "Total Sleep Problems Score", which was obtained by omitting the Behavioral Problems factor score, could be used to discriminate primary sleep problems from secondary sleep problems associated with ADHD.