149 Interplay Between Diabetes and Outcomes of Sepsis: A Nationwide Cohort Study With Laboratory Data

Abstract

The association between diabetes mellitus (DM) and sepsis hospitalization outcome was inconclusive. Also, studies using nationwide database with linked hospital information including laboratory data were rare. Understanding the trajectory of diabetic patients in sepsis course was important to improve medical care. To determine the effect of DM on the outcome of hospitalized sepsis patients. In nationwide database, diabetic severity score, socioeconomic status, and antidiabetic drugs were evaluated. Important clinical information, such as HbA1c, initial glucose level, APACHII score, and blood cultures were collected from hospital-based database to make further understanding. This is an observational cohort study of hospitalized sepsis patients using the nationwide database and linked hospital-based clinical information of a medical center. We collected the information using the national insurance claims database in a period of one year before the index hospitalization for sepsis and further included the hospitalization outcomes such as acute organ failure and complications. Propensity score matching was used to constitute the study (DM) and comparison (non-DM) cohorts. The major outcome was the association between DM complication severity and hospital mortality. Laboratory data including blood cultures of diabetic sepsis patients who were enrolled in the nationwide epidemiologic study were retrieved after link back to a medical center. We initially included 120,439 hospitalized sepsis patients. After propensity score matching, there was 19,719 diabetic and equal number of non-diabetic sepsis patients included in the study, respectively. Diabetic patients had an increased odds ratio (OR) of 1.14 (95% CI, 1.1-1.19) of mortality in the multivariate analysis. Furthermore, diabetic patients with adjusted diabetes complication severity index score (aDCSI score) of 0, 1, 2, 3, 4,and ≥5 had an OR of 0.91 (95% CI, 0.85-0.97), 0.87 (95% CI, 0.8-0.96), 1.14 (95% CI, 1.07-1.22), 1.25 (95% CI, 1.13-1.38), 1.56 (95% CI, 1.43-1.7), and 1.77 (95% CI, 1.61-1.96) in mortality, respectively. DPP-4 inhibitors and TZD use were associated with decreased OR of mortality of 0.82 (95% CI, 0.7-0.96) and 0.86 (95% CI, 0.76-0.98), respectively. A linked 1143 diabetic sepsis patients were retrieved after link to the hospital database. Initial blood glucose level in survived and dead diabetic septic patients did not have obvious difference: 217.5 (mg/dL). (150-315.8) versus 205 (138-318) (P=0.198). Also, the survived diabetic septic patients did not have a lower HbA1C (%) than the dead diabetic septic patients: 7.5 (6.5-9.1) versus 7.1 (6.1-8.8) (P=0.013*). There was no difference in blood culture classifications, that is, gram-positive coccus and gram-negative bacillus, etc, in survived and dead patients. Higher aDCSI score diabetic patients had an increased mortality rate during sepsis course. Complication prevention in diabetic patients play an important role. However, initial blood glucose and HbA1c levels did not affect the further sepsis course.