King's College London

Research portal

One (effect) size does not fit at all: Interpreting clinical significance and effect sizes in depression treatment trials

Research output: Contribution to journalArticle

Fredrik Hieronymus, Sameer Jauhar, Søren Dinesen Østergaard, Allan H Young

Original languageEnglish
Pages (from-to)1074-1078
Number of pages5
JournalJournal of Psychopharmacology
Volume34
Issue number10
DOIs
Published1 Oct 2020

Documents

  • Manuscript (clean) v1.5

    Manuscript_clean_v1.5.docx, 95.1 KB, application/vnd.openxmlformats-officedocument.wordprocessingml.document

    Uploaded date:08 Apr 2020

King's Authors

Abstract

The efficacy of antidepressants in major depressive disorder has been continually questioned, mainly on the basis of studies using the sum-score of the Hamilton Depression Rating Scale as a primary outcome parameter. On this measure antidepressants show a standardised mean difference of around 0.3, which some authors suggested is below the cut-off for clinical significance. Prompted by a recent review that, using this argument, concluded antidepressants should not be used for adults with major depressive disorder, we (a) review the evidence in support of the cut-off for clinical significance espoused in that article (a Hamilton Depression Rating Scale standardised mean difference of 0.875); (b) discuss the limitations of average Hamilton Depression Rating Scale sum-score differences between groups as measure of clinical significance; (c) explore alternative measures of clinical importance; and (d) suggest future directions to help overcome disagreements on how to define clinical significance. We conclude that (a) the proposed Hamilton Depression Rating Scale cut-off of 0.875 has no scientific basis and is likely misleading; (b) there is no agreed upon way of delineating clinically significant from clinically insignificant; (c) evidence suggests the Hamilton Depression Rating Scale sum-score underestimates antidepressant efficacy; and (d) future clinical trials should consider including measures directly reflective of functioning and wellbeing, in addition to measures focused on depression psychopathology.

View graph of relations

© 2020 King's College London | Strand | London WC2R 2LS | England | United Kingdom | Tel +44 (0)20 7836 5454