Table 4.

Logistic regression predicting CPAP compliance

Variables	B	SE	Wald chi-square	p	Odds ratio	95% CI
Constant	1.34	0.771	3.034	0.082
Apnea hypopnea index†	-0.02	0.009	3.494	0.062	0.984	0.967–1.001
Hypertension	0.66	0.358	3.433	0.064	1.94	0.962–3.916
Stage N1 (%)	0.04	0.014	6.516	0.011*	1.037	1.008–1.067
PSQI, overall sleep quality	-0.42	0.238	3.152	0.076	0.655	0.411–1.045
Covered by health insurance	-0.97	0.359	7.291	0.007**	0.379	0.188–0.767
Significance of the regression model				<0.001***

Nagelkerke R2=0.186 (the regression equation can explain the phenomenon by 18.6%). The final model predicted the actual data with a predictive value of 64.2%. Logistic regression analysis for compliance with independent variables such as age, hypertension, health insurance coverage, ISI (total score, dissatisfaction with sleep, how noticeable), overall sleep quality in the PSQI, and ratio of stage N1.

^*p<0.05;

^**p<0.01;

^***p<0.001;

^†entered as independent variables because of their clinical importance.

CPAP, continuous positive airway pressure; CI, confidence interval; PSQI, Pittsburgh Sleep Quality Inventory; SE, standard error; ISI, insomnia severity index