Predictors of sliding scale insulin use by housestaff physicians in the management of hospitalized patients with diabetes mellitus

Abstract

Objective: To identify predictive factors for the use of sliding scale insulin (SSI) by the housestaff physicians for in-hospital management of diabetes mellitus (DM). Design: Prospective cohort study. Materials & methods: A total of 215 consecutive patients admitted to the medical or surgical wards of two urban University-affiliated hospitals, with DM as a primary or secondary diagnosis, were prospectively followed to discharge. Demographic, laboratory and clinical data were obtained from in-hospital records. A survey was administered to the primary housestaff physicians regarding the potential reasons for prescribing SSI versus proactive antihyperglycemic therapy (standing insulin dose and/or oral antidiabetic agents). Results: SSI was prescribed for 71.2% of the patients and the lowest blood glucose (BG) at which insulin was given was recorded at 150–199 mg/dl (13.2% [of patients]), 200–249 mg/dl (81.1%) and 250–299 mg/dl (6.7%). Factors that predicted the use of SSI by housestaff physicians included the admission service, surgery versus medicine (odds ration [OR]: 6.0, 95% confidence interval: 5.5–23.3; p = 0.01), concern regarding wide swings of BG (OR: 5.56 [1.8–16.8]; p < 0.01), using the SSI results to estimate the standing dose of insulin (OR: 5.22 [1.8–14.7]; p < 0.01) and high BG on admission (OR: 3.92 [1.3–12.3]; p < 0.02). Conclusion: SSI is commonly prescribed for hospitalized patients with DM. It is more likely to be prescribed on the surgical wards compared with medical service. Perception among house staff regarding the wide swings of BG and the perception of the utility of the SSI to calculate standing-dose insulin were significant predictors for its use. Given the previous reports indicating higher in-hospital BG with the use of SSI, which leads to several complications, increased understanding of the reasons behind the use of SSI by house staff would help develop educational programs aimed at changing this practice in favor of more physiologic insulin regimens.