OBJECTIVE: To evaluate the efficacy of daily insulin dose increases in managing inpatient hyperglycemia. METHODS: Retrospective study of patients discharged from 2 urban academic medical centers and 3 large suburban community hospitals between 2015 and 2019 who received ≥10 units of basal insulin on any day. On hyperglycemic days (mean glucose ≥180 mg/dL), we categorized the relative insulin dose increases into 4 categories based on percentage changes from the previous day. We further subclassified these categories according to the average blood glucose (BG), total daily dose (TDD), and weight-based dosing quartiles. The primary goal was achieving an average BG of ≤160 mg/dL without subsequent hypoglycemia (≤70 mg/dL) on the following day. RESULTS: From 25 186 hospital admissions, we collected data on 240 556 hospital days and 63 033 hyperglycemic index days. The median age was 64, with 53.4% being male and 52.1% White. The median BG level was 222.7 mg/dL. Type 2 diabetes was coded in 54.7%, while 36.3% lacked a diabetes code but received basal insulin. Insulin dose adjustments showed a strong correlation with glycemic control
specifically, a 44% to 100% increase in TDD was significantly more likely to achieve the primary outcome, compared to a TDD increase of 10% to 22%. This trend remained consistent across varied BG ranges and dosing categories. CONCLUSION: More intensive insulin adjustments may be required for inpatient hyperglycemia compared to the typical 10% to 20% recommendation. Prospective studies are needed to validate and build upon these retrospective findings.