Prediction of depression in systemic lupus erythematosus patients using SF-36 Mental Health scores

Abstract

Objective. As depression is common in systemic lupus erythematosus (SLE) patients, we investigated whether and how the Medical Outcome Survey Short Form 36 (SF-36) scores, routinely used in the assessment of SLE patients, would indicate the absence or presence of depression. Methods. The Depression subscale of the Hospital Anxiety and Depression Scale (HADS-D) and the SF-36 were applied in a cross-sectional cohort of 60 SLE patients [mean age 45 (S.D. 15) yr, disease duration 11 (9) yr, 90% female, 100% Caucasians]. The SF-36 domain score with the closest association with HADS-D was used for further analysis. On the basis of HADS-D scores, the patients were split into two groups: one without depression (score < 8) and the other with possible depression (score ≥ 8). Results. The SF-36 Mental Health score was most closely correlated to the depression score (p = -0.69, P < 0.0005). The calculated Mental Health score cut-off value which significantly differentiated possibly depressed from non-depressed SLE patients was 61. Its sensitivity for the detection of possible depression was 89%, its specificity 77% and its negative predictive value 97%. Conclusions. The present study contributes to knowledge of means of excluding depression and the prevention of underdiagnosis and undertreatment of depression in SLE patients.