Skip to main content
. 2026 Aug 3;34(8):824. doi: 10.1007/s00520-026-11038-6

Table 3.

Intragroup variations and intergroup comparisons of clinical and sleep outcomes

Intervention group (n = 38) Control group (n = 38) p-value Treatment effect (95% CI)
Baseline Post-intervention Change (SD) Baseline Post-intervention Change (SD)
Mean (SD) Mean (SD) Mean (SD) Mean (SD)
Primary outcomes
PSQI
Sleep Quality (points) 9.37 (3.04) 9.97 (3.48) 0.60 (1.97) 8.74 (3.75) 10.79 (3.86) 2.05 (2.09) 0.006 −1.45 (−2.36; −0.56)
Secondary outcomes
PSQI
Time in bed (hours) 7.70 (1.12) 7.76 (1.33) 0.06 (0.79) 8.00 (1.28) 7.86 (1.28) −0.14 (0.61) 0.20 (−0.12; 0.53)
Sleep latency (min) 27.46 (33.16) 19.84 (19.75) −5.61 (19.7) 18.58 (18.74) 17.82 (14.95) −0.76 (14.3) −4.84 (−12.97; 2.50)
Nighttime sleep (hours) 5.55 (1.61) 5.86 (1.36) 0.31 (0.97) 6.01 (1.57) 5.76 (1.60) −0.24 (0.61) 0.55 (0.19; 0.94)
Sleep efficiency (%) 72.80 (18.80) 76.90 (18.10) 4.17 (11.9) 76.20 (19.90) 74.60 (20.10) −1.62 (7.52) 5.79 (1.50; 10.42)
ESS (points) 5.97 (4.46) 4.29 (4.45) −1.68 (2.47) 4.42 (4.16) 5.47 (4.94) 1.05 (3.15) −2.74 (−4.11 −0.56)
Accelerometer-derived sleep variables
Time in bed (hours) 7.74 (0.93) 7.77 (0.92) 0.03 (1.12) 7.94 (1.21) 7.75 (0.97) −0.19 (1.04) 0.22 (−0.26;0.71)
Nighttime sleep (hours) 6.63 (0.83) 6.66 (0.78) 0.03 (0.99) 6.68 (1.01) 6.60 (0.74) −0.08 (0.93) 0.11 (−0.32; 0.54)
Sleep efficiency (%) 86.10 (4.50) 86.40 (4.5) 0.30 (5.31) 84.30 (4.80) 86.12 (4.79) 1.80 (4.80) −1.53 (−3.82; 0.67)
Sleep interruptions (counts) 15.12 (3.79) 14.33 (2.76) −0.79 (2.83) 14.13 (3.00) 14.610 (2.38) 0.48 (2.71) −1.27 (−2.50; −0.01
CSQ (points) 27.61 (3.32) 24.63 (4.11) 2.97 (1.27; 4.68)

CI confidence interval, ESS Epworth Sleepiness Scale, PSQI Pittsburgh Sleep Quality Index, CSQ-8 8-item Client Satisfaction Questionnaire, SD Standard Deviation. p-value derived from the Wilcoxon rank-sum test comparing the mean change scores (variation from baseline) between the intervention and control groups, except for CSQ, which is derived from an independent Student’s t-test comparing post-intervention scores between groups. Treatment effect values represent the difference in mean change scores between the study groups. The corresponding 95% CI for change parameters was constructed using a non-parametric percentile bootstrap method based on 5000 replications to robustly accommodate skewed distributions without assuming population normality, while the CSQ CI was derived using standard parametric independent t-test assumptions. Significant differences are highlighted in bold