Table 3.

Logistic regression and Cox regression analyses evaluating mania symptom instability as predictor of likelihood of recovery and time until recovery from depression^a

Step and predictor	B	Wald	OR	P	ΔR2b
Logistic regression: predicting recovery
Step 1: treatment group	0.12	0.14	1.13 (0.59–2.19)	0.71	0.02
Step 2
Time in study	0.01	35.15	1.01 (1.01–1.02)	<0.01	0.42
Number of assessments	−0.14	19.17	0.87 (0.82–0.93)	<0.01
Initial depression symptoms	0.09	1.17	1.09 (0.93–1.29)	0.28
Initial mania symptoms	0.14	0.57	1.15 (0.80–1.65)	0.45
Average depression symptoms	−0.44	14.97	0.65 (0.52–0.81)	<0.01
Average mania symptoms	−0.09	0.11	0.91 (0.52–1.60)	0.75
Step 3: mania symptom instability	−0.53	4.21	0.59 (0.36–0.98)	0.04	0.03
Cox regression: predicting time until recovery
Step 1: treatment group	0.16	0.92	1.17 (0.85–1.63)	0.34	0.02
Step 2
Initial depression symptoms	0.08	3.15	1.08 (0.99–1.18)	0.08	0.41
Initial mania symptoms	0.09	0.83	1.09 (0.91–1.31)	0.36
Average depression symptoms	−0.29	23.05	0.75 (0.67–0.85)	<0.01
Average mania symptoms	0.02	0.02	1.02 (0.75–1.39)	0.89
Number of assessments	−0.93	21.10	0.40 (0.27–0.59)	<0.01
Number of assessments × time	0.16	17.09	1.17 (1.09–1.26)	<0.01
Step 3: mania symptom instability	−0.33	6.34	0.72 (0.56–0.93)	0.01	0.03

^a.n = 252. Treatment group is intensive psychosocial treatment (coded 1) v. collaborative care (coded 0). Average symptoms represent mean Clinical Monitoring Form (CMF) symptoms across follow-up period prior to recovery (for those who recovered) or end of study (for those who did not recover). Number of assessments is the number of CMFs completed prior to recovery (for those who recovered) or end of study (for those who did not recover). Symptom instability is the mean square successive difference in symptoms occurring prior to recovery (for those who recovered) or end of study (for those who did not recover). The proportionality of risk assumption was not upheld for survival analyses involving number of assessments, so the time-dependent covariates (interaction terms between time and these predictors) are included in Cox regression analyses.⁶¹

^b.R² for logistic regressions represents Nagelkerke R² change since previous step, an estimate of the increment in variance in the probability of recovery accounted for by the predictors tested since the previous step. Change in R² for Cox regressions represents Cox-Snell R² change since previous step, an estimate of the relative association between survival and the predictors tested since the previous step.⁶¹