Computer-Based versus Paper-Based Insulin Infusion Algorithms in Diabetic Ketoacidosis

Area of Science:

• Endocrinology
• Critical Care Medicine
• Medical Informatics

Background:

• Diabetic ketoacidosis (DKA) management in intensive care units (ICUs) relies on insulin infusion algorithms.
• Limited data exist comparing paper-based and computer-based algorithms for DKA in the ICU.
• This study addresses the need for comparative data on DKA management strategies.

Purpose of the Study:

• To compare the efficacy and safety of computer-based versus paper-based insulin infusion algorithms in managing DKA.
• To determine differences in DKA resolution time and hypoglycemia incidence between the two algorithm types.

Main Methods:

• Retrospective review of 620 ICU patients with DKA from 2015-2018.
• Patients were categorized into paper-based (pre-March 2016) and computer-based (post-March 2016) groups.
• Statistical analyses included summary and univariate comparisons.

Main Results:

• The computer-based group (n=373) showed a significantly shorter mean time to DKA resolution compared to the paper-based group (n=247).
• Severe hypoglycemia (<50 mg/dL) occurred significantly less often in the computer-based group (1%) versus the paper-based group (8%).
• No significant baseline demographic differences were noted, except for older age in the computer-based group.

Conclusions:

• Computer-based insulin infusion algorithms are more effective and safer for DKA management in the ICU.
• These algorithms expedite DKA resolution and reduce hypoglycemic events.
• Implementation of computer-based systems offers an improved therapeutic option for DKA patients.