Depression affects more than 400 million adults worldwide and contributes substantial morbidity and mortality. Clinical trials commonly report overall depression scale totals, which can obscure how individual symptoms respond to treatment. Although collaborative care is an evidence-based approach for depression in primary care, its impact on individual depressive symptoms—especially among racially and socioeconomically diverse patients—has been less well characterized. This report is a secondary, item-level analysis of the eIMPACT randomized controlled trial designed to address that gap.
The eIMPACT trial randomized 216 primary care patients to receive 12 months of modernized collaborative care or usual primary care. The study sample was diverse: mean age was 59 years, 78% were women, 50% identified as Black, and 46% reported annual income below $10,000. The modernized collaborative care intervention integrated internet-delivered cognitive behavioral therapy (CBT), telephonic CBT, and/or selected antidepressant medications.
This secondary analysis used two path analytic models to estimate simultaneous, item-level treatment effects on two commonly used measures of depressive symptoms: the Hopkins Symptom Checklist-20 (SCL-20) and the Patient Health Questionnaire-9 (PHQ-9). The analysis focused on between-group differences in individual symptom items rather than aggregate scale scores.
Details such as randomization procedures, exact model specifications, handling of missing data, duration and adherence to each intervention component, and statistical adjustment covariates were not reported in the PubMed abstract and therefore are not described here.
Across the two instruments, the modernized collaborative care intervention outperformed usual care on a large majority of individual symptom items. On the SCL-20, the intervention showed superiority on 15 of 20 items. On the PHQ-9, 7 of 9 items favored the intervention.
Seventeen symptoms demonstrated moderate-to-large between-group improvements, with reported standardized beta coefficients in the range β = -0.50 to -0.70. Notable symptoms in this category included hopelessness, anxious distress, depressed mood, self-blame, feelings of worthlessness, disturbed sleep, and fatigue.
An additional six symptoms showed smaller but still statistically significant improvements, with β values from -0.35 to -0.49.
Several symptoms did not show a significant advantage with the intervention. No between-group benefit was observed for suicidality, loneliness, sexual dysfunction, appetite change, or psychomotor symptoms; reported β values for these items ranged from -0.03 to -0.37.
The item-level outcomes indicate that modernized collaborative care produced broad symptom improvement across cognitive, affective, and somatic domains for most items on the SCL-20 and PHQ-9. The largest effects were seen for cognitive-affective symptoms such as hopelessness, depressed mood, self-blame, and worthlessness, as well as for common somatic complaints including sleep disturbance and fatigue. By contrast, symptoms that remained comparatively resistant to change included suicidality and several somatic and interpersonal domains (loneliness, sexual dysfunction, appetite change, psychomotor symptoms).
These contrasts suggest that total scale improvements can mask persistent symptom domains. Item-level analysis therefore provides additional granularity to identify which symptom clusters improve with a given intervention and which may require targeted or adjunctive treatments.
For clinicians and primary care teams using collaborative care models or digital/telephonic CBT approaches, these results suggest that modernized collaborative care can yield substantial improvement across most depressive symptoms in a socioeconomically and racially diverse outpatient population. Important takeaways include:
Monitor symptom-level change in addition to total scale scores to detect persistent or treatment-resistant domains.
Expect moderate-to-large improvements in cognitive-affective symptoms (e.g., hopelessness, depressed mood, self-blame) and common somatic complaints such as sleep disturbance and fatigue under modernized collaborative care.
Recognize that symptoms such as suicidality, loneliness, sexual dysfunction, appetite change, and psychomotor disturbances may persist despite collaborative care and could warrant targeted adjunctive interventions or specialty referral.
Use symptom-level data to personalize follow-up, select adjunctive therapies, and guide stepped-care decisions.
This summary is based on the PubMed abstract of the secondary analysis; specific methodological details were not reported there. Information not provided in the abstract includes randomization and blinding procedures for this secondary analysis, detailed statistical methods, adherence and engagement rates with internet or telephonic CBT, dose and selection criteria for antidepressant use, and longer-term follow-up outcomes beyond the 12-month intervention period. The abstract does not provide numerical confidence intervals, p-values for each item, or subgroup analyses beyond the reported demographic descriptors. Readers seeking full methodological and statistical detail should consult the full-text article.
In this secondary analysis of the eIMPACT randomized trial, modernized collaborative care incorporating internet and telephonic CBT plus selective antidepressant treatment improved most individual depressive symptoms on the SCL-20 and PHQ-9 in a racially and socioeconomically diverse primary care sample. Symptom-level analysis highlighted clinically important domains that were less responsive to the intervention—information that can inform targeted adjunctive treatment and optimization of recovery strategies.