Spanish version of Jenkins Sleep Scale in physicians and nurses: psychometric properties from a Peruvian nationally representative sample

Abstract

ObjectiveTo evaluate the psychometric properties of the Spanish version of Jenkins Sleep Scale with 4 items (JSS-4) of the Peruvian health system's (PHS) nurses and physicians. MethodsWe carried out a psychometric study based on secondary analysis in a sample from a nationally representative survey that used acomplex sampling design. The participants were physicians and nurses aged 18–65 years, working in PHS private and public facilities, who have fulfilled all JSS-4 items. We performed a confirmatory factor analysis. Reliability was evaluated via two estimates - classic alpha (α) and categorical omega (ω) coefficients. Also, we tested the invariance across groups of variables. The convergent validity was evaluated based on the relation between JSS-4 and PHQ-2 using Pearson's correlation coefficient and effect size (Cohen's d). Also, we designed normative values based on percentiles. ResultsWe included 2100 physicians and 2826 nurses in the analysis. We observed that the unidimensional model has adequate goodness-of-fit indices and values of α and ω coefficients.No measurement invariance was found between the groups of professionals and age groups; however, invariance was achieved between sex, monthly income, work-related illness, and chronic illness groups. Regarding the relation with other variables, the JSS-4 has a small correlation with PHQ-2. Also, profession and age-specific normative values were proposed. ConclusionJSS-4 Spanish version has adequate psychometric properties in PHS nurses and physicians.