Abstract:Background : The aim of this study was to investigate the optimal CVP range in sepsis and septic shock patients admitted to intensive care unit. Methods : We performed a retrospective study with adult sepsis patients with CVP records based on the eICU Collaborative Research Database. Multivariable logistic regression was performed to explore the associations between CVP level and hospital mortality. Non-linear correlations and optimal CVP range were explored using restricted cubic splines (RCS). Results : 5302 sepsis patients were included in this study. Patients in 4-8 mmHg group owned the lowest odds ratio for raw hospital mortality (19.7%). The logistic regression analyses revealed that hospital death risk increased significantly when mean CVP level exceeds 12 mmHg compared to 4-8 mmHg level. U-shaped association of CVP with hospital mortality was revealed by RCS model in septic shock patients and the optimal range was 5.6-12 mmHg. While, there was a J-shaped trend for non-septic shock patients. For non-septic shock patients, patients had an increased risk of hospital death only if CVP exceeded 11 mmHg. Conclusions : We observed U-shaped association between mean CVP level and hospital mortality in septic shock patients and J-shaped association in non-septic shock patients. This may imply that patients with different severity of sepsis have different CVP requirements. We need to monitor and manage CVP according to the circulatory status of the sepsis patient.

Exploring the Optimal Range of Central Venous Pressure in Sepsis and Septic Shock Patients: A Retrospective Study in 208 Hospitals