Abstract
Background: This study aimed to assess the accuracy of the Short-Form Health Survey (SF-36) mental health subscale (MH) and mental component summary (MCS) scores in identifying the presence of probable major depressive or anxiety disorder in patients with rheumatoid arthritis.
Methods: SF-36 data were collected in 100 hospital outpatients with rheumatoid arthritis. MH and MCS scores were compared against depression and anxiety data collected using validated measures as part of routine clinical practice. Sensitivity and specificity of the SF-36 were established using receiver operating characteristic (ROC) curve analysis, and area under the curve (AUC) compared the performance of the SF-36 components with the 9-item Patient Health Questionnaire (PHQ9) for depression and the 7-item Generalised Anxiety Disorder (GAD7) questionnaire for anxiety.
Results: The MH with a threshold of ≤52 had sensitivity and specificity of 81.0% and 71.4% respectively to detect anxiety, correctly classifying 73.5% of patients with probable anxiety disorder. A threshold of ≤56 had sensitivity and specificity of 92.6% and 73.2% respectively to detect depression, correctly classifying 78.6% of patients, and the same threshold could also be used to detect either depression or anxiety with a sensitivity of 87.9%, specificity of 76.9% and accuracy of 80.6%. The MCS with a threshold of ≤35 had sensitivity and specificity of 85.7% and 81.9% respectively to detect anxiety, correctly classifying 82.8% of patients with probable anxiety disorder. A threshold of ≤40 had sensitivity and specificity of 92.3% and 70.2% respectively to detect depression, correctly classifying 76.3% of patients. A threshold of ≤38 could be used to detect either depression or anxiety with a sensitivity of 87.5%, specificity of 80.3% and accuracy of 82.8%.
Conclusion: This analysis may increase the utility of a widely-used questionnaire. Overall, optimal use of the SF-36 for screening for mental disorder may be through using the MCS with a threshold of ≤38 to identify the presence of either depression or anxiety.
Methods: SF-36 data were collected in 100 hospital outpatients with rheumatoid arthritis. MH and MCS scores were compared against depression and anxiety data collected using validated measures as part of routine clinical practice. Sensitivity and specificity of the SF-36 were established using receiver operating characteristic (ROC) curve analysis, and area under the curve (AUC) compared the performance of the SF-36 components with the 9-item Patient Health Questionnaire (PHQ9) for depression and the 7-item Generalised Anxiety Disorder (GAD7) questionnaire for anxiety.
Results: The MH with a threshold of ≤52 had sensitivity and specificity of 81.0% and 71.4% respectively to detect anxiety, correctly classifying 73.5% of patients with probable anxiety disorder. A threshold of ≤56 had sensitivity and specificity of 92.6% and 73.2% respectively to detect depression, correctly classifying 78.6% of patients, and the same threshold could also be used to detect either depression or anxiety with a sensitivity of 87.9%, specificity of 76.9% and accuracy of 80.6%. The MCS with a threshold of ≤35 had sensitivity and specificity of 85.7% and 81.9% respectively to detect anxiety, correctly classifying 82.8% of patients with probable anxiety disorder. A threshold of ≤40 had sensitivity and specificity of 92.3% and 70.2% respectively to detect depression, correctly classifying 76.3% of patients. A threshold of ≤38 could be used to detect either depression or anxiety with a sensitivity of 87.5%, specificity of 80.3% and accuracy of 82.8%.
Conclusion: This analysis may increase the utility of a widely-used questionnaire. Overall, optimal use of the SF-36 for screening for mental disorder may be through using the MCS with a threshold of ≤38 to identify the presence of either depression or anxiety.
Original language | English |
---|---|
Article number | 224 |
Journal | BMC MUSCULOSKELETAL DISORDERS |
Volume | 17 |
DOIs | |
Publication status | Published - 23 May 2016 |