Practical issues encountered while determining Minimal Clinically Important Difference in Patient-Reported Outcomes.
Longitudinal modeling
Methodology
Minimal clinically important difference
Minimal important difference
Missing data
Patient-reported outcomes
Journal
Health and quality of life outcomes
ISSN: 1477-7525
Titre abrégé: Health Qual Life Outcomes
Pays: England
ID NLM: 101153626
Informations de publication
Date de publication:
27 May 2020
27 May 2020
Historique:
received:
26
03
2019
accepted:
07
05
2020
entrez:
29
5
2020
pubmed:
29
5
2020
medline:
25
9
2020
Statut:
epublish
Résumé
Using a real dataset, we highlighted several major methodological issues raised by the estimation of the Minimal Clinically Important Difference (MCID) of a Patient-Reported Outcomes instrument. We especially considered the management of missing data and the use of more than two times of measurement. While inappropriate missing data management and inappropriate use of multiple time points can lead to loss of precision and/or bias in MCID estimation, these issues are almost never dealt with and require cautious considerations in the context of MCID estimation. We used the LIGALONGO study (French Randomized Controlled Trial). We estimated MCID on the SF-36 General Health score by comparing many methods (distribution or anchor-based). Different techniques for imputation of missing data were performed (simple and multiple imputations). We also consider all measurement occasions by longitudinal modeling, and the dependence of the score difference on baseline. Three hundred ninety-three patients were studied. With distribution-based methods, a great variability in MCID was observed (from 3 to 26 points for improvement). Only 0.2 SD and 1/3 SD distribution methods gave MCID values consistent with anchor-based methods (from 4 to 7 points for improvement). The choice of missing data imputation technique clearly had an impact on MCID estimates. Simple imputation by mean score seemed to lead to out-of-range estimate, but as missing not at random mechanism can be hypothesized, even multiple imputations techniques can have led to an slight underestimation of MCID. Using 3 measurement occasions for improvement led to an increase in precision but lowered estimates. This practical example illustrates the substantial impact of some methodological issues that are usually never dealt with for MCID estimation. Simulation studies are needed to investigate those issues. NCT01240772 (ClinicalTrials.gov) registered on November 15, 2010.
Sections du résumé
BACKGROUND
BACKGROUND
Using a real dataset, we highlighted several major methodological issues raised by the estimation of the Minimal Clinically Important Difference (MCID) of a Patient-Reported Outcomes instrument. We especially considered the management of missing data and the use of more than two times of measurement. While inappropriate missing data management and inappropriate use of multiple time points can lead to loss of precision and/or bias in MCID estimation, these issues are almost never dealt with and require cautious considerations in the context of MCID estimation.
METHODS
METHODS
We used the LIGALONGO study (French Randomized Controlled Trial). We estimated MCID on the SF-36 General Health score by comparing many methods (distribution or anchor-based). Different techniques for imputation of missing data were performed (simple and multiple imputations). We also consider all measurement occasions by longitudinal modeling, and the dependence of the score difference on baseline.
RESULTS
RESULTS
Three hundred ninety-three patients were studied. With distribution-based methods, a great variability in MCID was observed (from 3 to 26 points for improvement). Only 0.2 SD and 1/3 SD distribution methods gave MCID values consistent with anchor-based methods (from 4 to 7 points for improvement). The choice of missing data imputation technique clearly had an impact on MCID estimates. Simple imputation by mean score seemed to lead to out-of-range estimate, but as missing not at random mechanism can be hypothesized, even multiple imputations techniques can have led to an slight underestimation of MCID. Using 3 measurement occasions for improvement led to an increase in precision but lowered estimates.
CONCLUSION
CONCLUSIONS
This practical example illustrates the substantial impact of some methodological issues that are usually never dealt with for MCID estimation. Simulation studies are needed to investigate those issues.
TRIAL REGISTRATION
BACKGROUND
NCT01240772 (ClinicalTrials.gov) registered on November 15, 2010.
Identifiants
pubmed: 32460882
doi: 10.1186/s12955-020-01398-w
pii: 10.1186/s12955-020-01398-w
pmc: PMC7251729
doi:
Banques de données
ClinicalTrials.gov
['NCT01240772']
Types de publication
Journal Article
Randomized Controlled Trial
Langues
eng
Sous-ensembles de citation
IM
Pagination
156Subventions
Organisme : Agence Nationale de la Recherche (FR)
ID : Jeunes Chercheurs 2016-2020 N° ANR-15-CE36-0003-01
Références
J Clin Epidemiol. 2008 Feb;61(2):102-9
pubmed: 18177782
Arch Intern Med. 1993 Jun 14;153(11):1337-42
pubmed: 8507124
Expert Rev Pharmacoecon Outcomes Res. 2011 Apr;11(2):163-9
pubmed: 21476818
J Clin Epidemiol. 2014 Apr;67(4):433-40
pubmed: 24447591
J Clin Epidemiol. 1996 Nov;49(11):1215-9
pubmed: 8892486
Med Care. 1999 May;37(5):469-78
pubmed: 10335749
Med Care. 1992 Jun;30(6):473-83
pubmed: 1593914
COPD. 2005 Mar;2(1):157-65
pubmed: 17136977
Spine (Phila Pa 1976). 2000 Dec 15;25(24):3192-9
pubmed: 11124736
Phys Ther. 1996 Oct;76(10):1109-23
pubmed: 8863764
J Clin Epidemiol. 1996 Jul;49(7):711-7
pubmed: 8691219
Health Qual Life Outcomes. 2006 Oct 11;4:79
pubmed: 17034633
Med Care. 1990 Jul;28(7):632-42
pubmed: 2366602
J Clin Epidemiol. 2017 Feb;82:128-136
pubmed: 27986610
Arthritis Rheum. 2001 Aug;45(4):384-91
pubmed: 11501727
J Clin Epidemiol. 2010 Jul;63(7):804-5; author reply 806
pubmed: 20399609
J Neurol Sci. 2014 Dec 15;347(1-2):268-74
pubmed: 25455299
Health Qual Life Outcomes. 2010 Nov 24;8:138
pubmed: 21106059
J Clin Epidemiol. 1999 Sep;52(9):861-73
pubmed: 10529027
J Clin Epidemiol. 2012 Mar;65(3):253-61
pubmed: 22014888
Qual Life Res. 2008 Feb;17(1):75-85
pubmed: 18027107
Ann Surg. 2016 Nov;264(5):710-716
pubmed: 27741005
J Clin Epidemiol. 2010 May;63(5):524-34
pubmed: 19926446
Control Clin Trials. 1989 Dec;10(4):407-15
pubmed: 2691207
Stat Methods Med Res. 2007 Jun;16(3):219-42
pubmed: 17621469
J Neurol Neurosurg Psychiatry. 2015 Mar;86(3):251-6
pubmed: 24987050
J Spine Surg. 2018 Mar;4(1):102-108
pubmed: 29732429
Med Care. 1989 Mar;27(3 Suppl):S178-89
pubmed: 2646488
Pharmacoeconomics. 2000 Nov;18(5):419-23
pubmed: 11151395
RMD Open. 2018 Oct 8;4(2):e000685
pubmed: 30402264
Spine J. 2007 Sep-Oct;7(5):541-6
pubmed: 17448732
Pharmacoeconomics. 2014 Dec;32(12):1157-70
pubmed: 25069632
J Rheumatol. 2005 Oct;32(10):2025-9
pubmed: 16206363