Table 2.

Disruption in antidepressant therapy

	Hazard Ratio [95% CI] (p-value)
	LIS		Non-LIS
	Aged	Disabled	Aged	Disabled
Generic initiation	0.88 [0.79, 0.98] (p=0.020)	0.84 [0.75, 0.93] (p=0.001)	0.78 [0.70, 0.87] (p<0.001)	0.71 [0.53, 0.96] (p=0.025)
Coverage gap (vs. pre-coverage gap)			1.27 [0.98, 1.65] (p=0.075)	2.05 [1.18, 3.55] (p=0.010)
Generic initiation * coverage gap			0.92 [0.65, 1.30] (p=0.621)	0.70 [0.27, 0.79] (p=0.455)

LIS – Low Income Subsidy; S.D. – standard deviation; CI - confidence interval

Results are based on Cox proportional hazard models that controlled for patient demographics (age, gender, race/ethnicity), measures of patient disease burden (indicators of 9 co-morbid conditions based on Chronic Condition Warehouse definition and the natural log of total prescription drug spending in 2007), and the specialty of the diagnosing clinician associated with the index outpatient visit (primary care, mental health specialty, or other medical specialty).