Bedtime sliding scale insulin is unnecessary for hospitalized patients with bedtime glucose < 300 mg/dL: A nudge-based quasi-experiment Journal Article


Authors: Flory, J. H.; Vertosick, E. A.; Kuperman, G.; Ancker, J. S.; Kim, S.; Fitzpatrick, C.; Gould, K.; Weiss, E.; Vickers, A. J.
Article Title: Bedtime sliding scale insulin is unnecessary for hospitalized patients with bedtime glucose < 300 mg/dL: A nudge-based quasi-experiment
Abstract: <p>Aims: Bedtime rapid-acting insulin is commonly given to hospitalized patients with moderate hyperglycemia. Its effectiveness has been evaluated in one randomized controlled trial (RCT), the results of which might not generalize to specific populations such as cancer patients. Methods: An order set encouraging use of rapid-acting bedtime insulin was replaced with one defaulting to rapid-acting bedtime insulin only for glucoses >= 300 mg/ dL, deployed over a two-month period in which approximately half of providers were arbitrarily assigned to the new order set. Analysis assessed the effect of the new order set on whether a less-aggressive bedtime sliding scale was actually ordered and mean morning glucose levels. Results: An inpatient insulin order set was placed for 458 patients. Exposure to the new order set was associated with a 91 % (95 % CI 87 % to 95 %) increase in use of the less-aggressive sliding scale. Mean morning glucose decreased by 16 mg/dL (95 % CI-16 to-1.1 mg/dL, p = 0.035). Conclusion: These results show it is unnecessary to give rapid-acting bedtime insulin for glucoses <300 mg/dL. Further research should be conducted using changes in the order set as a way of conducting well-controlled trials at low cost with minimum disruption to usual clinical care.</p>
Keywords: hyperglycemia; trial; order set
Journal Title: Diabetes Research and Clinical Practice
Volume: 228
ISSN: 0168-8227
Publisher: Elsevier Ireland Ltd.  
Date Published: 2025-10-01
Language: English
ACCESSION: WOS:001562095300004
DOI: 10.1016/j.diabres.2025.112428
PROVIDER: wos
Notes: Article -- 112428 -- Source: Wos
Altmetric
Citation Impact
BMJ Impact Analytics