Principal component analysis of coagulation assays in severely injured children

Christine M. Leeper, Matthew D. Neal, Christine McKenna, Timothy Billiar, Barbara A. Gaines*

*Corresponding author for this work

Research output: Contribution to journalArticlepeer-review

23 Scopus citations

Abstract

Background: Trauma-induced coagulopathy is common and associated with poor outcome in injured children. Our aim is to identify patterns of coagulation dysregulation after injury and associate these phenotypes with relevant clinical outcomes. Methods: We performed principal components analysis on prospectively collected data from children with the highest-level trauma activation June 2015–June 2016. Parameters included admission international normalized ratio, platelet count and thromboelastograms. Variables were reduced to principal components; principal component scores were generated for each subject and used in logistic regression with outcomes including mortality, disability, venous thromboembolism, and blood transfusion in the first 24 hours. Results: We included 133 subjects with median interquartile range age =10 (5–13 years), median interquartile range Injury Severity Score =17 (9–25), 73.5% boys, 70.8% blunt trauma. principal component analysis identified 3 significant principal components accounting for 75.0% of overall variance. Principal component 1 reflected clot strength; principal component 2 reflected abnormal fibrinolysis, both hyperfibrinolysis and fibrinolysis shutdown; principal component 3 reflected global clotting factor depletion. High principal component 1 score was associated with increased mortality (odds ratio =1.63) and blood transfusion (odds ratio 1.36). Principal component 2 score was correlated with Injury Severity Score (rho 0.4) and associated with venous thromboembolism (odds ratio 1.84), functional disability (odds ratio 1.66), mortality (odds ratio 2.07) and blood transfusion (odds ratio 2.79). PC3 score was associated with increased mortality (odds ratio 1.92) and blood transfusion (odds ratio 1.25). Conclusion: Principal component analysis detects 3 patterns of coagulation dysregulation using widely available laboratory parameters: (1) abnormalities in clot strength; (2) abnormalities in fibrinolysis, and (3) clotting factor depletion. While all were associated with mortality and transfusion, fibrinolytic dysregulation was associated with injury severity and portends particularly poor outcome including venous thromboembolism and disability.

Original languageEnglish
Pages (from-to)827-831
Number of pages5
JournalSurgery
Volume163
Issue number4
DOIs
StatePublished - Apr 2018
Externally publishedYes

Fingerprint

Dive into the research topics of 'Principal component analysis of coagulation assays in severely injured children'. Together they form a unique fingerprint.

Cite this