Table 2. Logistic Regression and Cox Regression Analyses Evaluating sleep group as predictors of Likelihood of Recovery and Time until Recovery.

Predictor	b	Wald	OR	(95% CI)	p	ΔR2*
Logistic regression: predicting recovery						.097
Treatment group**	-0.49	3.06	0.62	(0.36-1.06)	.08
Short sleepers***	-0.24	0.50	0.79	(0.41-1.52)	.48
Long sleepers***	-0.28	0.68	0.76	(0.39-1.47)	.41
Baseline GAF	0.04	0.02	1.04	(1.01-1.07)	.01
Anticonvulsants	0.05	0.02	1.05	(0.52-2.11)	.89
Atypical Antipsychotics	-0.27	0.74	0.76	(0.41-1.41)	.39
Other Mood Stabilizers	0.77	3.92	2.16	(1.01-4.63)	.05
Cox regression: predicting time until recovery						-.076
Treatment group**	-0.22	1.59	0.80	(0.57-1.13)	.21
Short sleepers***	-0.10	0.25	0.90	(0.60-1.36)	.62
Long sleepers***	-0.12	0.33	0.88	(0.58-1.35)	.57
Baseline GAF	0.01	1.18	1.01	(0.99-1.03)	.28
Anticonvulsants	0.12	0.33	1.13	(0.74-1.73)	.57
Atypical Antipsychotics	0.04	0.04	1.04	(0.71-1.52)	.85
Other Mood Stabilizers	0.36	2.02	1.43	(0.87-2.36)	.16

Abbreviations: OR (Odds ratio); CI (Confidence interval)

Notes:

^*For logistic regressions, R² represents Nagelkerke R², an estimate of the increment in variance in the probability of recovery accounted for by the predictors tested. For Cox regressions, R² represents Cox–Snell R², an estimate of the relative association between survival and the predictors tested.

^**Treatment group: intensive psychotherapy (1) versus collaborative care (0).

^***Sleep group: dummy coded with the normal sleepers group coded as the reference group, so only coefficients relative to the reference group are shown.