Validity and reliability of the Taiwanese version of the Pittsburgh Sleep Quality Index in cancer patients

Abstract

Background Sleep disturbance remains a common symptom among cancer patients. Assessment of sleep disturbance in cancer patients is hindered by infrequent use of standardized tools for sleep management. Objectives The purposes of this study were to validate the Taiwanese version of the Pittsburgh Sleep Quality Index (PSQI-T) and to determine detection cut-off points in cancer patients. Design A cross-sectional and descriptive correlational design. Participants A sample of 205 Taiwanese patients with various cancer diagnoses. Methods The survey included the Pittsburgh Sleep Quality Index-Taiwanese version, the MD Anderson Symptom Inventory-Taiwanese version, the Brief Fatigue Inventory-Taiwanese version, the seven-day sleep log, and Manual of Mental Disorders, Fourth Edition. Results The internal consistency Cronbach's alpha for the PSQI was 0.79. Test–retest reliability was 0.91 for the global score over a 20- to 28-day interval in a sample of 16 patients. Construct validity was established by a significant relationship of the PSQI-T global score to the total symptom severity score and the fatigue severity score. Convergent validity was examined by correlating the PSQI-T scores and scores of the DSM-IV and scores on the seven-day sleep log. Known-group validity was established by comparing PSQI-T scores for patients having low fatigue levels and those having high fatigue levels. Receiver-operating characteristic (ROC) curves were used to determine the detection cut-off points. Conclusions We found that a PSQI-T global score of 8 generates the best sensitivity and specificity for measuring sleep disturbance in cancer patients. The PSQI-T is a reliable, valid, and sensitive instrument for measuring sleep quality among Taiwanese cancer patients.