In an interim evaluation on baseline data of the German PADRE observational study the correlation between physician- and patient assessment of emotional and physical symptoms of depressed patients was evaluated.

This multicenter, prospective, 6-month observational study focused on adult outpatients with a depressive episode as diagnosed according to ICD-10 criteria, chosen by their physician to start new anti-depressive treatment with duloxetine. Correlations between the applied depression and/or pain scales were calculated via Spearman's correlation coefficient. Symptoms were evaluated via clinician rated 'Inventory for Depressive Symptomatology' ([IDS-C], total score, including item 25), patient rated 5-item scale 'KUSTA', (rating mood, activity, tension-relaxation, sleep and appetite on visual analog scales [VAS]), and patient rated VAS for 'Pain'.

All participating physicians are psychiatrists/neurologists. 2.748 patients (71% female, mean age 52.7 yrs) were evaluated. Any pain symptoms were documented in 88.9% of patients at baseline. When comparing patient- with physician-assessments, correlation of PPS scales was low to moderate and varied for different pain types: IDS-C item 25 (="somatic disorders") vs. overall pain-VAS: r=0.421 (95% CI 0.390, 0.452), IDS-C item 25 vs. abdominal pain: r=0.189; IDS-C item 25 vs. chest-pain: r=0.179. When comparing IDS-C total vs. the KUSTA items, correlation was moderate in all cases (e.g.: r= -0.510 for IDS total vs. KUSTA mood).

Only a low to moderate correlation was observed between physician- and patient assessment for PPS in depressed patients. Therefore, patient pain ratings should explicitly be included in the assessment of depressed patients.