Causal Bayesian machine learning to assess treatment effect heterogeneity by dexamethasone dose for patients with COVID-19 and severe hypoxemia.
Journal
Scientific reports
ISSN: 2045-2322
Titre abrégé: Sci Rep
Pays: England
ID NLM: 101563288
Informations de publication
Date de publication:
21 04 2023
21 04 2023
Historique:
received:
03
01
2023
accepted:
12
04
2023
medline:
25
4
2023
pubmed:
22
4
2023
entrez:
21
04
2023
Statut:
epublish
Résumé
The currently recommended dose of dexamethasone for patients with severe or critical COVID-19 is 6 mg per day (mg/d) regardless of patient features and variation. However, patients with severe or critical COVID-19 are heterogenous in many ways (e.g., age, weight, comorbidities, disease severity, and immune features). Thus, it is conceivable that a standardized dosing protocol may not be optimal. We assessed treatment effect heterogeneity in the COVID STEROID 2 trial, which compared 6 mg/d to 12 mg/d, using a causal inference framework with Bayesian Additive Regression Trees, a flexible modeling method that detects interactive effects and nonlinear relationships among multiple patient characteristics simultaneously. We found that 12 mg/d of dexamethasone, relative to 6 mg/d, was probably associated with better long-term outcomes (days alive without life support and mortality after 90 days) among the entire trial population (i.e., no signals of harm), and probably more beneficial among those without diabetes mellitus, that were older, were not using IL-6 inhibitors at baseline, weighed less, or had higher level respiratory support at baseline. This adds more evidence supporting the use of 12 mg/d in practice for most patients not receiving other immunosuppressants and that additional study of dosing could potentially optimize clinical outcomes.
Identifiants
pubmed: 37085591
doi: 10.1038/s41598-023-33425-3
pii: 10.1038/s41598-023-33425-3
pmc: PMC10120498
doi:
Substances chimiques
Dexamethasone
7S5I7G3JQL
Types de publication
Journal Article
Research Support, N.I.H., Extramural
Research Support, Non-U.S. Gov't
Langues
eng
Sous-ensembles de citation
IM
Pagination
6570Subventions
Organisme : NHLBI NIH HHS
ID : R01 HL126802
Pays : United States
Organisme : Department of Health
ID : CS-2016-16-011
Pays : United Kingdom
Informations de copyright
© 2023. The Author(s).
Références
Stat Med. 2018 Oct 15;37(23):3309-3324
pubmed: 29862536
J Clin Epidemiol. 2005 Mar;58(3):261-8
pubmed: 15718115
Intensive Care Med. 2022 Jan;48(1):45-55
pubmed: 34757439
Acta Anaesthesiol Scand. 2022 Sep;66(8):987-995
pubmed: 35781689
Ann Intern Med. 2007 Oct 16;147(8):573-7
pubmed: 17938396
Intensive Care Med. 2022 Nov;48(11):1525-1538
pubmed: 36102943
Science. 2020 Sep 4;369(6508):
pubmed: 32669297
JAMA. 2021 Nov 9;326(18):1807-1817
pubmed: 34673895
JAMA. 2022 Feb 15;327(7):683
pubmed: 35166802
Nat Med. 2020 Oct;26(10):1623-1635
pubmed: 32807934
Stat Med. 2021 Sep 20;40(21):4691-4713
pubmed: 34114252
Nature. 2020 Aug;584(7821):463-469
pubmed: 32717743
Nature. 2021 Feb;590(7847):635-641
pubmed: 33429418
Stat Methods Med Res. 2019 Apr;28(4):1079-1093
pubmed: 29254443
Acta Anaesthesiol Scand. 2021 Jul;65(6):834-845
pubmed: 33583034
JAMA. 2021 Aug 10;326(6):499-518
pubmed: 34228774
N Engl J Med. 2021 Feb 25;384(8):693-704
pubmed: 32678530
Stat Med. 2021 May 20;40(11):2665-2691
pubmed: 33751659
BMJ. 2020 Sep 4;370:m3379
pubmed: 32887691
Nat Med. 2022 Jan;28(1):39-50
pubmed: 35064248
Crit Care. 2006 Feb;10(1):103
pubmed: 16420653
J Comput Graph Stat. 2018;27(1):209-219
pubmed: 29706752
Acta Anaesthesiol Scand. 2021 Nov;65(10):1421-1430
pubmed: 34138478
Health Technol Assess. 2001;5(33):1-56
pubmed: 11701102
JAMA. 2020 Oct 6;324(13):1330-1341
pubmed: 32876694
Trends Pharmacol Sci. 2019 Jan;40(1):38-49
pubmed: 30497693
Acta Anaesthesiol Scand. 2023 Feb;67(2):195-205
pubmed: 36314057
BMJ. 2015 Nov 04;351:h5651
pubmed: 26537915