Collider bias in trauma comparative effectiveness research

The stratification blues for systematic reviews

on behalf of the ROC Investigators

Research output: Contribution to journalArticle

10 Citations (Scopus)

Abstract

Background Collider bias, or stratifying data by a covariate consequence rather than cause (confounder) of treatment and outcome, plagues randomised and observational trauma research. Of the seven trials of prehospital hypertonic saline in dextran (HSD) that have been evaluated in systematic reviews, none found an overall between-group difference in survival, but four reported significant subgroup effects. We hypothesised that an avoidable type of collider bias often introduced inadvertently into trauma comparative effectiveness research could explain the incongruous findings. Methods The two most recent HSD trials, a single-site pilot and a multi-site pivotal study, provided data for a secondary analysis to more closely examine the potential for collider bias. The two trials had followed the a priori statistical analysis plan to subgroup patients by a post-randomisation covariate and well-established surrogate for bleeding severity, massive transfusion (MT), ≥10 unit of red blood cells within 24 h of admission. Despite favourable HSD effects in the MT subgroup, opposite effects in the non-transfused subgroup halted the pivotal trial early. In addition to analyzing the data from the two trials, we constructed causal diagrams and performed a meta-analysis of the results from all seven trials to assess the extent to which collider bias could explain null overall effects with subgroup heterogeneity. Results As in previous trials, HSD induced significantly greater increases in systolic blood pressure (SBP) from prehospital to admission than control crystalloid (p = 0.003). Proportionately more HSD than control decedents accrued in the non-transfused subgroup, but with paradoxically longer survival. Despite different study populations and a span of over 20 years across the seven trials, the reported mortality effects were consistently null, summary RR = 0.99 (p = 0.864, homogeneity p = 0.709). Conclusions HSD delayed blood transfusion by modifying standard triggers like SBP with no detectable effect on survival. The reported heterogeneous HSD effects in subgroups can be explained by collider bias that trauma researchers can avoid by improved covariate selection and data capture strategies.

Original languageEnglish (US)
Pages (from-to)775-780
Number of pages6
JournalInjury
Volume46
Issue number5
DOIs
StatePublished - 2015

Fingerprint

dextran - saline drug combination
Comparative Effectiveness Research
Wounds and Injuries
Blood Pressure
Survival
Plague
Random Allocation
Blood Transfusion
Meta-Analysis
Erythrocytes
Research Personnel
Hemorrhage

Keywords

  • Bias
  • Comparative effectiveness research
  • Massive transfusion
  • Observational
  • Randomised clinical trial
  • Resuscitation
  • Subgroup
  • Survival
  • Trauma

ASJC Scopus subject areas

  • Emergency Medicine
  • Orthopedics and Sports Medicine

Cite this

Collider bias in trauma comparative effectiveness research : The stratification blues for systematic reviews. / on behalf of the ROC Investigators.

In: Injury, Vol. 46, No. 5, 2015, p. 775-780.

Research output: Contribution to journalArticle

on behalf of the ROC Investigators. / Collider bias in trauma comparative effectiveness research : The stratification blues for systematic reviews. In: Injury. 2015 ; Vol. 46, No. 5. pp. 775-780.
@article{023622857e9f44ab916fc44d38a71377,
title = "Collider bias in trauma comparative effectiveness research: The stratification blues for systematic reviews",
abstract = "Background Collider bias, or stratifying data by a covariate consequence rather than cause (confounder) of treatment and outcome, plagues randomised and observational trauma research. Of the seven trials of prehospital hypertonic saline in dextran (HSD) that have been evaluated in systematic reviews, none found an overall between-group difference in survival, but four reported significant subgroup effects. We hypothesised that an avoidable type of collider bias often introduced inadvertently into trauma comparative effectiveness research could explain the incongruous findings. Methods The two most recent HSD trials, a single-site pilot and a multi-site pivotal study, provided data for a secondary analysis to more closely examine the potential for collider bias. The two trials had followed the a priori statistical analysis plan to subgroup patients by a post-randomisation covariate and well-established surrogate for bleeding severity, massive transfusion (MT), ≥10 unit of red blood cells within 24 h of admission. Despite favourable HSD effects in the MT subgroup, opposite effects in the non-transfused subgroup halted the pivotal trial early. In addition to analyzing the data from the two trials, we constructed causal diagrams and performed a meta-analysis of the results from all seven trials to assess the extent to which collider bias could explain null overall effects with subgroup heterogeneity. Results As in previous trials, HSD induced significantly greater increases in systolic blood pressure (SBP) from prehospital to admission than control crystalloid (p = 0.003). Proportionately more HSD than control decedents accrued in the non-transfused subgroup, but with paradoxically longer survival. Despite different study populations and a span of over 20 years across the seven trials, the reported mortality effects were consistently null, summary RR = 0.99 (p = 0.864, homogeneity p = 0.709). Conclusions HSD delayed blood transfusion by modifying standard triggers like SBP with no detectable effect on survival. The reported heterogeneous HSD effects in subgroups can be explained by collider bias that trauma researchers can avoid by improved covariate selection and data capture strategies.",
keywords = "Bias, Comparative effectiveness research, Massive transfusion, Observational, Randomised clinical trial, Resuscitation, Subgroup, Survival, Trauma",
author = "{on behalf of the ROC Investigators} and {Del Junco}, {Deborah J.} and Bulger, {Eileen M.} and Fox, {Erin E.} and Holcomb, {John B.} and Karen Brasel and Hoyt, {David B.} and Grady, {James J.} and Sarah Duran and Patricia Klotz and Dubick, {Michael A.} and Wade, {Charles E.}",
year = "2015",
doi = "10.1016/j.injury.2015.01.043",
language = "English (US)",
volume = "46",
pages = "775--780",
journal = "Injury",
issn = "0020-1383",
publisher = "Elsevier Limited",
number = "5",

}

TY - JOUR

T1 - Collider bias in trauma comparative effectiveness research

T2 - The stratification blues for systematic reviews

AU - on behalf of the ROC Investigators

AU - Del Junco, Deborah J.

AU - Bulger, Eileen M.

AU - Fox, Erin E.

AU - Holcomb, John B.

AU - Brasel, Karen

AU - Hoyt, David B.

AU - Grady, James J.

AU - Duran, Sarah

AU - Klotz, Patricia

AU - Dubick, Michael A.

AU - Wade, Charles E.

PY - 2015

Y1 - 2015

N2 - Background Collider bias, or stratifying data by a covariate consequence rather than cause (confounder) of treatment and outcome, plagues randomised and observational trauma research. Of the seven trials of prehospital hypertonic saline in dextran (HSD) that have been evaluated in systematic reviews, none found an overall between-group difference in survival, but four reported significant subgroup effects. We hypothesised that an avoidable type of collider bias often introduced inadvertently into trauma comparative effectiveness research could explain the incongruous findings. Methods The two most recent HSD trials, a single-site pilot and a multi-site pivotal study, provided data for a secondary analysis to more closely examine the potential for collider bias. The two trials had followed the a priori statistical analysis plan to subgroup patients by a post-randomisation covariate and well-established surrogate for bleeding severity, massive transfusion (MT), ≥10 unit of red blood cells within 24 h of admission. Despite favourable HSD effects in the MT subgroup, opposite effects in the non-transfused subgroup halted the pivotal trial early. In addition to analyzing the data from the two trials, we constructed causal diagrams and performed a meta-analysis of the results from all seven trials to assess the extent to which collider bias could explain null overall effects with subgroup heterogeneity. Results As in previous trials, HSD induced significantly greater increases in systolic blood pressure (SBP) from prehospital to admission than control crystalloid (p = 0.003). Proportionately more HSD than control decedents accrued in the non-transfused subgroup, but with paradoxically longer survival. Despite different study populations and a span of over 20 years across the seven trials, the reported mortality effects were consistently null, summary RR = 0.99 (p = 0.864, homogeneity p = 0.709). Conclusions HSD delayed blood transfusion by modifying standard triggers like SBP with no detectable effect on survival. The reported heterogeneous HSD effects in subgroups can be explained by collider bias that trauma researchers can avoid by improved covariate selection and data capture strategies.

AB - Background Collider bias, or stratifying data by a covariate consequence rather than cause (confounder) of treatment and outcome, plagues randomised and observational trauma research. Of the seven trials of prehospital hypertonic saline in dextran (HSD) that have been evaluated in systematic reviews, none found an overall between-group difference in survival, but four reported significant subgroup effects. We hypothesised that an avoidable type of collider bias often introduced inadvertently into trauma comparative effectiveness research could explain the incongruous findings. Methods The two most recent HSD trials, a single-site pilot and a multi-site pivotal study, provided data for a secondary analysis to more closely examine the potential for collider bias. The two trials had followed the a priori statistical analysis plan to subgroup patients by a post-randomisation covariate and well-established surrogate for bleeding severity, massive transfusion (MT), ≥10 unit of red blood cells within 24 h of admission. Despite favourable HSD effects in the MT subgroup, opposite effects in the non-transfused subgroup halted the pivotal trial early. In addition to analyzing the data from the two trials, we constructed causal diagrams and performed a meta-analysis of the results from all seven trials to assess the extent to which collider bias could explain null overall effects with subgroup heterogeneity. Results As in previous trials, HSD induced significantly greater increases in systolic blood pressure (SBP) from prehospital to admission than control crystalloid (p = 0.003). Proportionately more HSD than control decedents accrued in the non-transfused subgroup, but with paradoxically longer survival. Despite different study populations and a span of over 20 years across the seven trials, the reported mortality effects were consistently null, summary RR = 0.99 (p = 0.864, homogeneity p = 0.709). Conclusions HSD delayed blood transfusion by modifying standard triggers like SBP with no detectable effect on survival. The reported heterogeneous HSD effects in subgroups can be explained by collider bias that trauma researchers can avoid by improved covariate selection and data capture strategies.

KW - Bias

KW - Comparative effectiveness research

KW - Massive transfusion

KW - Observational

KW - Randomised clinical trial

KW - Resuscitation

KW - Subgroup

KW - Survival

KW - Trauma

UR - http://www.scopus.com/inward/record.url?scp=84929914472&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84929914472&partnerID=8YFLogxK

U2 - 10.1016/j.injury.2015.01.043

DO - 10.1016/j.injury.2015.01.043

M3 - Article

VL - 46

SP - 775

EP - 780

JO - Injury

JF - Injury

SN - 0020-1383

IS - 5

ER -