【Abstract】Objective The sleep quality in maintenance hemodialysis (MHD) patients was investigated and its influencing factors were analyzed. Methods The sleep quality of 112 MHD patients was investigated using the Pittsburg Sleep Quality Index (PSQI) scale, Social Support Rating Scale (SSRS), Self-rating Anxiety Scale (SAS) and Self-rating Depression Scale (SDS). According to the PSQI score, patients with scores less than or equal to 10 were classified as a group without severe poor sleep quality (G1), and those with scores of >10 were classified as a group with severe poor sleep quality (G2). SSRS, SAS and SDS were used to evaluate the influences of anxiety and depression on sleep quality. Results Thirty-eight of the 112 MHD
patients (33.93%) had a PSQI score of >10. Four factors were significant different between the two groups: age (62.20 ± 10.62 years vs. 51.51±15.12 years, t=4.253, P＜0.001), months of dialysis (78.46±88.96 months vs. 46.12±45.68 months, t=2.028, P=0.048), albumin (37.94±2.86 g/L vs. 39.37±2.49 g/L, t=2.542, P=0.013), and PTH (243.32±158.85 pg/ml vs. 327.54±209.34 pg/ml, t=2.324, P=0.022). SDS score was also significant different between the two groups (χ2=7.110, P=0.029). A logistic regression model was established to predict the sleep quality: f(x) =1/(1 + e0.074289x1 + 0.055897x2 + 0.99036) (x1 is age; Z =- 2.600, P =0.009, and x2 is SDS score, Z =-2.004, P=0.028). The model had the accuracy of 71.9% and the area under curve of 0.748. Conclusions Sleep quality is a common problem in MHD patients. Severe sleep quality problem usually exists in elderly patients, patients with long dialysis period, low albumin, and low PTH. Patients with depression are easily to have severe sleep problem. We have established a logistic regression model to predict sleep quality, which may be clinically useful.
