Multivariate meta-analysis of critical care meta-analyses: a meta-epidemiological study.

Publisher: BioMed Central Country of Publication: England NLM ID: 100968545 Publication Model: Electronic Cited Medium: Internet ISSN: 1471-2288 (Electronic) Linking ISSN: 14712288 NLM ISO Abbreviation: BMC Med Res Methodol Subsets: MEDLINE

Original Publication: London : BioMed Central, [2001-

Critical Care* ; Research Design*; Epidemiologic Studies ; Humans ; Multivariate Analysis ; Odds Ratio

Background: Meta-analyses typically consider multiple outcomes and report univariate effect sizes considered as independent. Multivariate meta-analysis (MVMA) incorporates outcome correlation and synthesises direct evidence and related outcome estimates within a single analysis. In a series of meta-analyses from the critically ill literature, the current study contrasts multiple univariate effect estimates and their precision with those derived from MVMA.
Methods: A previous meta-epidemiological study was used to identify meta-analyses with either one or two secondary outcomes providing sufficient detail to structure bivariate or tri-variate MVMA, with mortality as primary outcome. Analysis was performed using a random effects model for both odds ratio (OR) and risk ratio (RR); borrowing of strength (BoS) between multivariate outcome estimates was reported. Estimate comparisons, β coefficients, standard errors (SE) and confidence interval (CI) width, univariate versus multivariate, were performed using Lin's concordance correlation coefficient (CCC).
Results: In bivariate meta-analyses, for OR (n = 49) and RR (n = 48), there was substantial concordance (≥ 0.69) between estimates; but this was less so for tri-variate meta-analyses for both OR (n = 25; ≥ 0.38) and RR (≥ -0.10; n = 22). A variable change in the multivariate precision of primary mortality outcome estimates compared with univariate was present for both bivariate and tri-variate meta-analyses and for metrics. For second outcomes, precision tended to decrease and CI width increase for bivariate meta-analyses, but was variable in the tri-variate. For third outcomes, precision increased and CI width decreased. In bivariate meta-analyses, OR coefficient significance reversal, univariate versus MVMA, occurred once for mortality and 6 cases for second outcomes. RR coefficient significance reversal occurred in 4 cases; 2 were discordant with OR. For tri-variate OR meta-analyses reversal of coefficient estimate significance occurred in two cases for mortality, nine cases for second and 7 cases for third outcomes. In RR meta-analyses significance reversals occurred for mortality in 2 cases, 6 cases for second and 3 cases for third; there were 7 discordances with OR. BoS was greater in trivariate MVMAs compared with bivariate and for OR versus RR.
Conclusions: MVMA would appear to be the preferred solution to multiple univariate analyses; parameter significance changes may occur. Analytic metric appears to be a determinant.
(© 2021. The Author(s).)

CMAJ. 2007 Apr 10;176(8):1091-6. (PMID: 17420491)
BMC Med Res Methodol. 2007 Jan 12;7:3. (PMID: 17222330)
BMJ. 2010 Feb 15;340:c365. (PMID: 20156912)
Stat Med. 2002 Feb 28;21(4):589-624. (PMID: 11836738)
Crit Care Med. 2002 Mar;30(3):555-62. (PMID: 11990914)
Infection. 2009 Dec;37(6):497-507. (PMID: 19826761)
Control Clin Trials. 1986 Sep;7(3):177-88. (PMID: 3802833)
Stat Med. 2011 Sep 10;30(20):2481-98. (PMID: 21268052)
J Clin Epidemiol. 2009 Dec;62(12):1292-300. (PMID: 19447007)
Ann Intensive Care. 2014 Oct 26;4:32. (PMID: 25593748)
Crit Care. 2017 Oct 17;21(1):253. (PMID: 29041948)
Crit Care Resusc. 2017 Jun;19(2):117-127. (PMID: 28651507)
J Clin Epidemiol. 2020 Nov 7;:. (PMID: 33171273)
Brain Inj. 2018;32(6):693-703. (PMID: 29580096)
Crit Care Med. 2011 Aug;39(8):1985-91. (PMID: 21478738)
Crit Care. 2012 Dec 12;16(2):R66. (PMID: 22534505)
Minerva Anestesiol. 2010 Apr;76(4):276-86. (PMID: 20332741)
Cochrane Database Syst Rev. 2015 Aug 04;(8):CD008875. (PMID: 26241698)
Arch Intern Med. 2009 Feb 9;169(3):219-29. (PMID: 19204211)
Stat Med. 1993 Dec 30;12(24):2273-84. (PMID: 7907813)
J Stat Softw. 2014 Jan 1;56(11):11. (PMID: 24904241)
Eur J Epidemiol. 2016 Apr;31(4):337-50. (PMID: 27209009)
Anesth Analg. 2018 Feb;126(2):513-521. (PMID: 29189271)
Biostatistics. 2008 Jan;9(1):172-86. (PMID: 17626226)
Stat Med. 2019 Oct 15;38(23):4772-4787. (PMID: 31338853)
Biometrics. 2020 Dec;76(4):1240-1250. (PMID: 32720712)
Stat Med. 2017 Jan 30;36(2):301-317. (PMID: 27714841)
Intensive Care Med. 2008 Nov;34(11):1969-79. (PMID: 18575838)
Br J Anaesth. 2015 Mar;114(3):396-405. (PMID: 25534400)
Eur Heart J. 2009 May;30(10):1245-53. (PMID: 19336434)
CMAJ. 2009 Apr 14;180(8):821-7. (PMID: 19318387)
Intensive Care Med. 2006 May;32(5):639-49. (PMID: 16570149)
Res Synth Methods. 2015 Jun;6(2):157-74. (PMID: 26099484)
J Clin Epidemiol. 2015 May;68(5):542-50. (PMID: 25537265)
Lancet. 2008 Dec 6;372(9654):1962-76. (PMID: 19012955)
Crit Care. 2016 Apr 29;20(1):117. (PMID: 27129307)
Cochrane Database Syst Rev. 2017 Jul 13;7:CD004104. (PMID: 28702957)
JAMA. 2005 Dec 28;294(24):3124-30. (PMID: 16380593)
Ann Intern Med. 2008 Jul 15;149(2):109-22. (PMID: 18626050)
Crit Care. 2016 Aug 19;19:262. (PMID: 27538711)
J Intensive Care Med. 2021 May 4;:8850666211014479. (PMID: 33942659)
Intensive Care Med. 2013 Aug;39(8):1396-404. (PMID: 23685609)
J Am Coll Cardiol. 2009 May 5;53(18):1677-89. (PMID: 19406344)
Res Synth Methods. 2019 Sep;10(3):440-451. (PMID: 31058440)
Am J Kidney Dis. 2007 Jan;49(1):56-68. (PMID: 17185146)
J Clin Epidemiol. 2019 Feb;106:41-49. (PMID: 30342086)
J Clin Epidemiol. 2017 Mar;83:8-17. (PMID: 28063915)
Br J Surg. 2016 Jan;103(1):14-26. (PMID: 26522616)
Crit Care Med. 2018 Jul;46(7):1049-1056. (PMID: 29629984)
Stat Med. 2012 Dec 20;31(29):3821-39. (PMID: 22807043)
Crit Care Med. 2018 Feb;46(2):252-263. (PMID: 29189348)
Stat Methods Med Res. 2017 Dec;26(6):2853-2868. (PMID: 26546254)
Intensive Care Med. 2008 Nov;34(11):1980-90. (PMID: 18626628)
Stat Med. 1998 Nov 30;17(22):2537-50. (PMID: 9839346)
Medicine (Baltimore). 2015 Aug;94(31):e1319. (PMID: 26252319)
J Crit Care. 2019 Oct;53:114-119. (PMID: 31228761)
BMC Nephrol. 2017 Aug 7;18(1):264. (PMID: 28784106)
Int J Antimicrob Agents. 2014 May;43(5):403-11. (PMID: 24657044)
Psychol Methods. 2006 Jun;11(2):193-206. (PMID: 16784338)
J R Stat Soc Ser C Appl Stat. 2018 Nov;67(5):1177-1205. (PMID: 30344346)
PLoS One. 2013 Jul 05;8(7):e66844. (PMID: 23861749)
J Hosp Infect. 2007 Mar;65(3):187-203. (PMID: 17244516)
Crit Care Med. 2018 Feb;46(2):236-243. (PMID: 29116995)
Stat Med. 2013 Mar 30;32(7):1191-205. (PMID: 23208849)
Stat Med. 2012 Sep 10;31(20):2179-95. (PMID: 22532016)
Clin Trials. 2012 Oct;9(5):610-20. (PMID: 22872546)
J Neurotrauma. 2008 Jan;25(1):62-71. (PMID: 18355159)
Res Synth Methods. 2018 Jul 3;:. (PMID: 29971966)
JAMA. 2009 Jun 10;301(22):2362-75. (PMID: 19509383)
Ann Intensive Care. 2017 Dec;7(1):58. (PMID: 28585162)
BMJ. 2007 Apr 28;334(7599):889. (PMID: 17387118)
BMC Anesthesiol. 2016 Jul 18;16(1):39. (PMID: 27430341)
Res Synth Methods. 2018 Mar;9(1):73-88. (PMID: 29055096)
Stat Med. 2014 Apr 30;33(9):1441-59. (PMID: 24285290)
Crit Care Resusc. 2016 Mar;18(1):17-24. (PMID: 26947412)
Stat Med. 2007 Jan 15;26(1):78-97. (PMID: 16526010)
Eur Heart J. 2008 Nov;29(21):2592-600. (PMID: 18664462)
Arch Intern Med. 2010 May 10;170(9):751-8. (PMID: 20458081)
Res Synth Methods. 2019 Sep;10(3):398-419. (PMID: 30854785)
BMJ. 2017 Sep 13;358:j3932. (PMID: 28903924)
J Clin Epidemiol. 2000 Nov;53(11):1119-29. (PMID: 11106885)
J Gastrointest Surg. 2008 Apr;12(4):739-55. (PMID: 17939012)
Lipids Health Dis. 2015 Mar 31;14:23. (PMID: 25889853)
Crit Care Med. 2010 Mar;38(3):954-62. (PMID: 20016381)
Crit Care Resusc. 2018 Jun;20(2):150-163. (PMID: 29852854)
Crit Care. 2015 Aug 18;19:294. (PMID: 26283217)

Keywords: Borrowing of strength; Critical care; Metric; Multivariate meta-analysis; Random effects

Date Created: 20210719 Date Completed: 20210809 Latest Revision: 20210809

20231215

PMC8286437

10.1186/s12874-021-01336-4

34275460