Objective: To assess the influence of somatic symptoms of the Hamilton Depression Rating Scale (HAMD) and Montgomery-Asberg Depression Rating Scale (MADRS) on the clinimetric performance of these scales in patients with Parkinson's disease (PD).
Methods: A total of 224 patients underwent a protocolized mental status examination, consisting of the Structured Clinical Interview for DSM-IV depressive disorder (SCID-D), as well as the HAMD and MADRS. Sensitivity, specificity, positive, and negative predictive values for a range of cut-off scores were calculated for both rating scales and for modified versions of these scales in which all somatic items were eliminated. In addition, receiver operating characteristic (ROC) curves were obtained for both the modified and unmodified scales.
Results: Elimination of the somatic items of depression from the HAMD and MADRS resulted in a reduced specificity of both the HAMD and the MADRS, and an increased sensitivity of the MADRS.
Conclusion: The authors recommend the full version of the HAMD and MADRS if used for diagnostic purposes; for screening purposes, the abbreviated version without somatic items can be used. Additional advantages of using full rating scales, with somatic items included, are that these provide more information on the severity of depression and allow for easier comparison across studies.
Copyright 2010 Elsevier Inc. All rights reserved.