Handling of rescue and missing data affects synthesis and interpretation of evidence: The sodium-glucose cotransporter 2 inhibitor example

Catharine B. Stack, A. Russell Localio, Michael E. Griswold, Steven N. Goodman, Cynthia D. Mulrow

Research output: Contribution to journalArticle

Abstract

Confidence in evidence summarized in meta-analyses depends on the strength of the underlying studies. This inherent limitation of syntheses appears in the case of a meta-analysis of sodium-glucose cotransporter 2 inhibitors for the treatment of type 2 diabetes because many of the pertinent randomized trials did not handle patient dropout and "rescue" medication properly. Repudiated statistical methods, such as last observation carried forward, and unsophisticated methods for handling postrescue data produce unreliable summary estimates. Future reports of randomized studies and meta-analyses of those studies must focus on posing precise questions about the treatment effect of interest and then implement appropriate statistical methods to account for missing data, patient dropout, and use of rescue medication.

Original languageEnglish (US)
Pages (from-to)285-288
Number of pages4
JournalAnnals of Internal Medicine
Volume159
Issue number4
DOIs
StatePublished - Aug 20 2013
Externally publishedYes

Fingerprint

Sodium-Glucose Transport Proteins
Meta-Analysis
Patient Dropouts
Type 2 Diabetes Mellitus
Observation

ASJC Scopus subject areas

  • Internal Medicine

Cite this

Handling of rescue and missing data affects synthesis and interpretation of evidence : The sodium-glucose cotransporter 2 inhibitor example. / Stack, Catharine B.; Localio, A. Russell; Griswold, Michael E.; Goodman, Steven N.; Mulrow, Cynthia D.

In: Annals of Internal Medicine, Vol. 159, No. 4, 20.08.2013, p. 285-288.

Research output: Contribution to journalArticle

Stack, Catharine B. ; Localio, A. Russell ; Griswold, Michael E. ; Goodman, Steven N. ; Mulrow, Cynthia D. / Handling of rescue and missing data affects synthesis and interpretation of evidence : The sodium-glucose cotransporter 2 inhibitor example. In: Annals of Internal Medicine. 2013 ; Vol. 159, No. 4. pp. 285-288.
@article{3c7b9bc952634414b83eb28a9f7a59dd,
title = "Handling of rescue and missing data affects synthesis and interpretation of evidence: The sodium-glucose cotransporter 2 inhibitor example",
abstract = "Confidence in evidence summarized in meta-analyses depends on the strength of the underlying studies. This inherent limitation of syntheses appears in the case of a meta-analysis of sodium-glucose cotransporter 2 inhibitors for the treatment of type 2 diabetes because many of the pertinent randomized trials did not handle patient dropout and {"}rescue{"} medication properly. Repudiated statistical methods, such as last observation carried forward, and unsophisticated methods for handling postrescue data produce unreliable summary estimates. Future reports of randomized studies and meta-analyses of those studies must focus on posing precise questions about the treatment effect of interest and then implement appropriate statistical methods to account for missing data, patient dropout, and use of rescue medication.",
author = "Stack, {Catharine B.} and Localio, {A. Russell} and Griswold, {Michael E.} and Goodman, {Steven N.} and Mulrow, {Cynthia D.}",
year = "2013",
month = "8",
day = "20",
doi = "10.7326/0003-4819-159-4-201308200-00009",
language = "English (US)",
volume = "159",
pages = "285--288",
journal = "Annals of Internal Medicine",
issn = "0003-4819",
publisher = "American College of Physicians",
number = "4",

}

TY - JOUR

T1 - Handling of rescue and missing data affects synthesis and interpretation of evidence

T2 - The sodium-glucose cotransporter 2 inhibitor example

AU - Stack, Catharine B.

AU - Localio, A. Russell

AU - Griswold, Michael E.

AU - Goodman, Steven N.

AU - Mulrow, Cynthia D.

PY - 2013/8/20

Y1 - 2013/8/20

N2 - Confidence in evidence summarized in meta-analyses depends on the strength of the underlying studies. This inherent limitation of syntheses appears in the case of a meta-analysis of sodium-glucose cotransporter 2 inhibitors for the treatment of type 2 diabetes because many of the pertinent randomized trials did not handle patient dropout and "rescue" medication properly. Repudiated statistical methods, such as last observation carried forward, and unsophisticated methods for handling postrescue data produce unreliable summary estimates. Future reports of randomized studies and meta-analyses of those studies must focus on posing precise questions about the treatment effect of interest and then implement appropriate statistical methods to account for missing data, patient dropout, and use of rescue medication.

AB - Confidence in evidence summarized in meta-analyses depends on the strength of the underlying studies. This inherent limitation of syntheses appears in the case of a meta-analysis of sodium-glucose cotransporter 2 inhibitors for the treatment of type 2 diabetes because many of the pertinent randomized trials did not handle patient dropout and "rescue" medication properly. Repudiated statistical methods, such as last observation carried forward, and unsophisticated methods for handling postrescue data produce unreliable summary estimates. Future reports of randomized studies and meta-analyses of those studies must focus on posing precise questions about the treatment effect of interest and then implement appropriate statistical methods to account for missing data, patient dropout, and use of rescue medication.

UR - http://www.scopus.com/inward/record.url?scp=84882243403&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84882243403&partnerID=8YFLogxK

U2 - 10.7326/0003-4819-159-4-201308200-00009

DO - 10.7326/0003-4819-159-4-201308200-00009

M3 - Article

C2 - 24026261

AN - SCOPUS:84882243403

VL - 159

SP - 285

EP - 288

JO - Annals of Internal Medicine

JF - Annals of Internal Medicine

SN - 0003-4819

IS - 4

ER -