Predictors of remission from body dysmorphic disorder after internet-delivered cognitive behavior therapy: a machine learning approach.
Body dysmorphic disorder
Cognitive behaviour therapy
Internet
Machine learning
Predictor
Journal
BMC psychiatry
ISSN: 1471-244X
Titre abrégé: BMC Psychiatry
Pays: England
ID NLM: 100968559
Informations de publication
Date de publication:
19 05 2020
19 05 2020
Historique:
received:
03
12
2019
accepted:
05
05
2020
entrez:
21
5
2020
pubmed:
21
5
2020
medline:
13
11
2020
Statut:
epublish
Résumé
Previous attempts to identify predictors of treatment outcomes in body dysmorphic disorder (BDD) have yielded inconsistent findings. One way to increase precision and clinical utility could be to use machine learning methods, which can incorporate multiple non-linear associations in prediction models. This study used a random forests machine learning approach to test if it is possible to reliably predict remission from BDD in a sample of 88 individuals that had received internet-delivered cognitive behavioral therapy for BDD. The random forest models were compared to traditional logistic regression analyses. Random forests correctly identified 78% of participants as remitters or non-remitters at post-treatment. The accuracy of prediction was lower in subsequent follow-ups (68, 66 and 61% correctly classified at 3-, 12- and 24-month follow-ups, respectively). Depressive symptoms, treatment credibility, working alliance, and initial severity of BDD were among the most important predictors at the beginning of treatment. By contrast, the logistic regression models did not identify consistent and strong predictors of remission from BDD. The results provide initial support for the clinical utility of machine learning approaches in the prediction of outcomes of patients with BDD. ClinicalTrials.gov ID: NCT02010619.
Sections du résumé
BACKGROUND
Previous attempts to identify predictors of treatment outcomes in body dysmorphic disorder (BDD) have yielded inconsistent findings. One way to increase precision and clinical utility could be to use machine learning methods, which can incorporate multiple non-linear associations in prediction models.
METHODS
This study used a random forests machine learning approach to test if it is possible to reliably predict remission from BDD in a sample of 88 individuals that had received internet-delivered cognitive behavioral therapy for BDD. The random forest models were compared to traditional logistic regression analyses.
RESULTS
Random forests correctly identified 78% of participants as remitters or non-remitters at post-treatment. The accuracy of prediction was lower in subsequent follow-ups (68, 66 and 61% correctly classified at 3-, 12- and 24-month follow-ups, respectively). Depressive symptoms, treatment credibility, working alliance, and initial severity of BDD were among the most important predictors at the beginning of treatment. By contrast, the logistic regression models did not identify consistent and strong predictors of remission from BDD.
CONCLUSIONS
The results provide initial support for the clinical utility of machine learning approaches in the prediction of outcomes of patients with BDD.
TRIAL REGISTRATION
ClinicalTrials.gov ID: NCT02010619.
Identifiants
pubmed: 32429939
doi: 10.1186/s12888-020-02655-4
pii: 10.1186/s12888-020-02655-4
pmc: PMC7238519
doi:
Banques de données
ClinicalTrials.gov
['NCT02010619']
Types de publication
Journal Article
Research Support, Non-U.S. Gov't
Langues
eng
Sous-ensembles de citation
IM
Pagination
247Subventions
Organisme : Karolinska Institutet
ID : Clinical Scientist Training Programme
Pays : International
Organisme : Fredrik och Ingrid Thurings Stiftelse
ID : 2018-00390
Pays : International
Références
J Psychiatr Res. 2006 Mar;40(2):95-104
pubmed: 16229856
J Clin Psychiatry. 1998;59 Suppl 20:22-33;quiz 34-57
pubmed: 9881538
Psychopharmacol Bull. 1994;30(2):179-86
pubmed: 7831453
Methods Inf Med. 2012;51(1):74-81
pubmed: 21915433
BMJ Open. 2014 Sep 25;4(9):e005923
pubmed: 25256187
Clin Psychol Rev. 2015 Aug;40:156-69
pubmed: 26117062
Int J Psychiatry Clin Pract. 2016 Sep;20(3):191-8
pubmed: 27314665
Bioinformatics. 2012 Jan 1;28(1):112-8
pubmed: 22039212
Ann Med. 2001 Jul;33(5):337-43
pubmed: 11491192
Behav Ther. 2015 Nov;46(6):775-85
pubmed: 26520220
Int J Methods Psychiatr Res. 2015 Jun;24(2):156-69
pubmed: 25994109
Body Image. 2016 Sep;18:168-86
pubmed: 27498379
Acta Psychiatr Scand. 1994 Jan;89(1):21-8
pubmed: 8140903
Psychol Med. 2019 Oct 30;:1-7
pubmed: 31662124
Mol Psychiatry. 2017 Jan;22(1):37-43
pubmed: 27843153
JAMA Psychiatry. 2018 Dec 1;75(12):1289-1297
pubmed: 30347013
Psychol Med. 2013 May;43(5):1109-17
pubmed: 23171833
Compr Psychiatry. 2015 Apr;58:108-15
pubmed: 25617963
Clin Psychol Rev. 2016 Nov;49:55-66
pubmed: 27607741
BMJ. 2016 Feb 02;352:i241
pubmed: 26837684
Psychopharmacol Bull. 1997;33(1):17-22
pubmed: 9133747
Compr Psychiatry. 2005 Jul-Aug;46(4):254-60
pubmed: 16175755
Compr Psychiatry. 2008 Nov-Dec;49(6):561-9
pubmed: 18970904
Psychol Med. 2019 Oct;49(14):2330-2341
pubmed: 30392475
Int J Methods Psychiatr Res. 2018 Mar;27(1):
pubmed: 28752937
J Clin Epidemiol. 1999 Oct;52(10):935-42
pubmed: 10513756
BMJ Open. 2019 Jan 15;9(1):e024307
pubmed: 30647044
Behav Res Ther. 2007 Sep;45(9):2232-40
pubmed: 17306222
Behav Ther. 2014 May;45(3):314-27
pubmed: 24680228
J Psychiatr Res. 2004 Jan;38(1):63-72
pubmed: 14690771
Eur Psychiatry. 2006 Mar;21(2):75-80
pubmed: 16360307
Clin Psychol Rev. 2016 Aug;48:43-51
pubmed: 27393916
Am J Psychiatry. 1997 Nov;154(11):1593-8
pubmed: 9356570
J Consult Clin Psychol. 2010 Apr;78(2):184-9
pubmed: 20350029
Psychother Psychosom. 2014;83(6):341-53
pubmed: 25323062