[Relationship between blood glucose trajectory during intensive care unit stay and mortality in patients with sepsis-associated acute respiratory distress syndrome].

Summary

Using MIMIC-IV and group-based trajectory modeling in 3,869 sepsis-associated ARDS patients, three 7-day glucose trajectories were identified: low-normal (6.1–7.0 mmol/L), moderate (7.8–8.9), and persistent hyperglycemia (10.6–13.1). Persistent hyperglycemia was associated with significantly higher 28-day and 1-year mortality compared with lower trajectories after adjustment.

Key Findings

• Three 7-day glucose trajectories identified via GBTM: low-normal (6.1–7.0 mmol/L; n=1,523), moderate (7.8–8.9; n=1,452), persistent hyperglycemia (10.6–13.1; n=894)
• Persistent hyperglycemia trajectory showed significantly higher 28-day and 1-year mortality than lower trajectories
• Associations persisted after multivariable adjustment; hypoglycemia incidence across groups was assessed

Clinical Implications

Avoid persistent hyperglycemia in sepsis-associated ARDS; implement frequent monitoring and protocols to maintain glucose in lower target ranges while minimizing hypoglycemia. Findings should inform design of RCTs testing trajectory-guided glycemic control.

Limitations

• Retrospective single-database study with potential residual confounding and missing data bias
• Glycemic management protocols and insulin dosing heterogeneity not standardized

Future Directions

Prospective trials testing trajectory-guided glycemic targets in sepsis-associated ARDS and mechanistic studies of glucose variability, inflammation, and lung injury.