Association between ABO blood groups and sepsis risk in severe burns: a retrospective cohort study.

INTRODUCTION

Severe burns cause both physical and psychological harm while significantly increasing susceptibility to infections. Previous studies have linked ABO blood types to various diseases; however, their role in the development of sepsis, particularly in patients with severe burns, remains unclear. This retrospective cohort study examined whether ABO blood type is associated with sepsis outcomes in patients with severe burns from the 2014 Kunshan explosion.

METHODS

After the 2014 Kunshan explosion incident, 185 hospitalized patients were studied. Patients lacking ABO information or essential demographic details were excluded. Kaplan-Meier survival analysis was used to compare sepsis-free survival (SFS) and septic shock-free survival (SSFS) across ABO groups. Multivariable Cox regression was used to assess whether ABO blood type served as an independent prognostic factor for SFS in severely burned patients. The multivariable model was assessed using receiver operating characteristic curves and calibration plots. Additional confounding factors were analyzed through subgroup analysis.

RESULTS

Significant variations in SFS (p = 0.006) and SSFS (p = 0.002) were observed across ABO blood groups in Kaplan-Meier survival analysis. In multivariable analyses, ABO blood type was not independently associated with sepsis-free survival. Subgroup analysis further revealed significant differences in acute kidney injury (p = 0.02) and complications (p < 0.05) suggesting potential trends that warrant further investigation.

CONCLUSION

On univariable analyses, SFS and SSFS showed statistical differences across ABO groups, with type O longest and type AB shortest. After adjustment for burn extent, age, and sex, ABO type was not independently associated with sepsis-free survival, consistent with the dominant effect of burn extent. This univariable pattern is biologically plausible given higher von Willebrand factor and factor VIII in non-O groups which favor microvascular immunothrombosis, blood type related differences in inflammatory mediators such as hepatocyte growth factor MCP 1 and IL 18, and pathogen binding to A B and H antigens on injured surfaces. Future multicentre prospective studies are needed to validate these findings and clarify clinical relevance. CLINICAL TRIAL: Not applicable.