Table 2.

Process-Outcome Relationships in Previous Studies of Depressed Patients

Study	Population	Indicator	Source	Time frame	Outcome	Source	Time frame	Correction for selection bias	Indicator–outcome relationship
Rost, Williams, Wherry, & Smith, 1995	38 primary-care patients with current major depression identified by screening	75% or greater compliance with therapeutic dose of AD for 8+ weeks	Patient report	6 month window following index visit	2 or fewer depressive symptoms	Patient report	6 months following index visit	Covariates	Significant but strength not estimated
Simon et al., 1995	88 primary-care patients identified by AD prescription for new treatment episode	Average daily dose diagnosed within 10% of recommended dose	Administrative database report	6 months following referral to study	No current major depression at 1, 4, or 7 months AND 50% or more reduction in depression severity	Patient report	7 months following index visit	Covariates	Significant improvement with adequate doses compared to inadequate doses (84% versus 64%)
Melfi et al., 1998	4052 patients beginning new treatment episode identified by depression diagnosis and AD prescription	Filled AD prescriptions 4+ times	Administrative database	6 months following initial prescription	New AD treatment episode OR hospital admission for depression OR electro-convulsive shock therapy OR emergency room visit for mental health OR attempted suicide	Administrative database	18 months following treatment period	Covariates	Significantly reduces risk of relapse/remission by 56%
Katon et al., 2000	1,599 primary care patients identified by initial AD prescription for new treatment episode	AD prescription refill	Administrative database	6 weeks following initial prescription	Persistent depressive symptoms	Patient report	6–8 weeks following initial prescriptions	Covariates	Significantly reduces probability of persistent symptoms (r = −.18)
Fortney et al., 2001	106 primary-care patients with current major and minor depression identified by screening	75% or greater compliance with therapeutic dose of AD for 8+ weeks OR 8+ specialty care visits over 12 week period	Patient report Administrative database report	6 month window not keyed to index visit	Depression severity MCS PCS	patient report	6 months following baseline interview	Covariates IV sensitivity analysis	Significant 10.1–12.2 point improvement in depression severity in patients with major depression
Bull et al., 2002	401 primary-case patients identified by AD prescription for new episode of major depression	≥3 depression related follow-up visits with physician	Administrative database report	Not specified	Continuation of AD medicationa	Patient report	75–105 days following AD initiation	Covariates	Significantly more patients who continued ADs had 3+ follow-up visits (37% versus 18%)
Schoenbaum et al., 2002	938 primary are patients with current major and minor depression identified by screening	Any AD use at therapeutic level OR 4+ specialty counseling visits	Patient report	6 months following index visit	Probable depression MCS	Patient report	6 months following index visit	IV analysis	Significantly increases remission (76% versus 30%) and improves MCS by 15.4 points

Note. IV = Instrument variable. MCS = Mental Component Scale of SF-12 or SF-36. AD = Antidepressant medication. PCS = Physical Component Scale of SF-36.

^aRelated to improvement in depressive symptoms at 3 months.