Sleep quality in cancer patients: a common metric for several instruments measuring sleep quality.

Purpose: Sleep problems are frequently observed in cancer patients. Multiple questionnaires for assessing sleep quality have been developed. The aim of this study was to present transfer rules that allow the conversion of the patients' scores from one questionnaire to another. In addition, we anchored this common metric to the general population.

Methods: A sample of 1,733 cancer patients completed the following questionnaires: Pittsburgh Sleep Quality Index, Insomnia Sleep Index, Jenkins Sleep Scale, EORTC QLQ-C30, and the sleep scale of the EORTC QLQ-SURV100. The methods for establishing a common metric were based on Item Response Theory.

Results: The main result of the study is a figure that allows the conversion from one of the above-mentioned sleep scales into another. Furthermore, the scores of the questionnaires can be transferred to theta scores that indicate the position within the group of cancer patients and also to T scores that indicate the position in relation to the general population. The correlations between the sleep scales ranged between 0.70 and 0.85.

Conclusions: The conversion rules presented in the study enable researchers and clinicians to directly compare single scores or mean scores across studies using different sleep scales, to assess the degree of sleep problems with regard to the general population, and to relate cutoff scores from one questionnaire to another.