| Purpose The aim of this study was to test the hypothesis that two blood glucose variability measures have effect on the outcome in sepsis within72hours.Method A prospectively collected data set including all glucose measures and other lab date and the outcome was analyzed in the ICU within72hours. Then used the logistic regression to assess the association between hospital mortality and standard deviation (SD),"hyper-hypo" glucose variability and other lab data.Result There were841glucose measurements included100patients admitted in ICU. We found the SD (p=0.031, OR=0.333(0.123-0.907)) and "hyper-hypo" glucose variability (p=0.046, OR=0.170(0.030-0.966)) and the glucose level>10mmol/L has association with the mortality (p=0.031OR=0.084(0.009-0.798)).Conclusion Both of the SD and "hyper-hypo" glucose variability have effect on the mortality of sepsis, and the glucose level>10mmol/L has association with the mortality. |
