Depression affects over 400 million adults worldwide and is associated with substantial morbidity and mortality. Although symptoms vary in their responsiveness to treatment, clinical trials typically report total scores, obscuring symptom-level effects. Despite collaborative care’s effectiveness for depression in primary care, its impact on individual depressive symptoms among racially and socioeconomically diverse patients remains unclear.

We conducted a secondary analysis of the eIMPACT trial of 216 primary care patients (Mage = 59 years; 78% women; 50% Black; 46% with income <$10,000/year). Participants were randomized to 12 months of modernized collaborative care (internet cognitive-behavioral therapy [CBT], telephonic CBT, and/or select antidepressants) or usual primary care. Two path analytic models estimated simultaneous item-level effects on the Hopkins Symptom Checklist-20 (SCL-20) and Patient Health Questionnaire-9 (PHQ-9).

The intervention outperformed usual care on 15 of 20 SCL-20 items and 7 of 9 PHQ-9 items. Seventeen symptoms – including hopelessness, anxious distress, depressed mood, self-blame, worthlessness, disturbed sleep, and fatigue – showed moderate-to-large between-group improvement (βs = −0.50 to −0.70). Smaller but significant gains were observed for six other symptoms (βs = −0.35 to −0.49). No significant advantage emerged for suicidality, loneliness, sexual dysfunction, appetite change, or psychomotor symptoms (βs = −0.03 to −0.37).

Modernized collaborative care improves most depressive symptoms in diverse primary care patients, though cross-measure persistent symptoms highlight clinically consequential domains that may be obscured by total scores. Symptom-level analyses may help identify persistent symptom domains requiring targeted adjunctive treatment to optimize recovery.