Clinical outcomes associated with the use of subcutaneous insulin-by-glucose sliding scales to manage hyperglycemia in hospitalized patients with pneumonia

OBJECTIVE: Despite its ubiquitous use, the value of subcutaneous insulin-by-glucose sliding scales (SISS) for the management of in-patient hyperglycemia has not been carefully assessed. METHODS: The medical charts of 391 patients >45 years of age admitted with pneumonia from January 2003 to May 2004 who had a recorded glucose within 24h of admission and who did not have active cancer, tuberculosis or AIDS were reviewed. Abstracted data included demographics, clinical characteristics, admission and daily glucose levels, medications, SISS use and clinical outcomes. The primary outcome was pre-defined as a composite of in-hospital mortality, cardiovascular complications, sepsis or ICU admission. RESULTS: Compared to patients not prescribed an SISS during the admission, the 47 patients prescribed an SISS had a higher rate of the following outcomes: primary outcome (OR=2.55; 95% CI 1.38-4.73); cardiovascular complications or death (OR=1.86; 95% CI 0.99-3.49), sepsis or ICU admission (OR=4.98; 95% CI 2.38-10.42). The relationship between sliding scale use and the primary outcome was statistically significant, even after controlling for age, sex, diabetes, steroids, CHF and COPD (P<0.0001). Patients receiving a sliding scale had mean in-hospital glucose values of 11.83 mmol/L versus 7.2 mmol/L (P<0.0001) in patients not receiving an insulin sliding scale. CONCLUSION: Among patients admitted to a medical ward with pneumonia, an SISS is associated with higher glucose levels and poorer clinical outcomes.