Does strict glycemic control lead to better outcomes in non-icu type II diabetic patients with sepsis

Document Type

Conference Proceeding

Publication Date


Publication Title

Endocrine Practice


Objective: Studies have proven better outcomes in surgical patients with better glycemic control, but not among medical patients. The purpose of this research is to investigate if strict glycemic control leads to shorter hospital stay, lower mortality and re-admission in diabetic patients with sepsis. Methods: This retrospective study included 395 type 2 diabetics with sepsis between 2015-2017. Recorded data included Charlson's comorbidity score, qSOFA score, A1C, percentage of glucose controlled, length of stay,30-day mortality,3-month re-admission and type of insulin used. Pre-admission A1C< 7.5 was defined as controlled diabetic. Hospital glucose that was<180 for >50% of the time was defined as well controlled. Results: 51.6% of total were controlled diabetics and 21.5% of those became uncontrolled during the hospital stay. 48.4% of total were uncontrolled diabetics and 56.5% of those remained uncontrolled during the hospital stay. The well-controlled group length of stay was 6.41 days versus 5.95 days(p=0.145) in the uncontrolled group. 30- day mortality was 3.7% in the well-controlled group and 1.3% in the uncontrolled group(p=0.266). Re-admission rate at 3 months was 17.4% in the well-controlled group and 20.9% in the uncontrolled group(p=0.503). Sliding scale had glucose controlled success rate of 59.9% compared to 33.5%(p<0.001) with basal + bolus. Readmission at 3 months was strongly associated with high Charlson's comorbidity score of 4.14(p=0.008). Discussion: A study of 32,851 non-selctive patients found tight blood glucose control in non-critically ill patients decreased length of stay by 0.31(P<0.01), decreased mortality by 0.47(p<0.00001), and decreased re-admission by 2.47(p<0.0000). The poorly controlled group was composed of uncontrolled diabetics with multiple comorbidities which explained the longer length of stay and mortality. The data was considered skewed by adding all the non-diabetics in the study. Our study was based on a similar study design, but only type 2 diabetics were included and the results didn't support the strong finding in the previous research. This study validated the importance of Charlson's comorbidity score in prediction of re-admission. Our study suggested sliding scale is superior to basal + bolus in glucose control. However, 73% of patients placed on sliding scale were well controlled diabetics and took an oral agent at home. In this group of patients, the glucose will fall in well-controlled range even without sliding scale. Therefore, there is no benefit of using sliding scale over oral agents. Conclusion: Our study could not support better outcomes with tight glycemic control from previous study after limiting the study objects to only type 2 diabetics.



First Page


Last Page


This document is currently not available here.