Automated Self-Adjusting Insulin Algorithm Improves Inpatient Glycemic Control
Overview
A three-year retrospective study demonstrated that an automated self-adjusting subcutaneous insulin algorithm (SQIA) significantly improved glycemic control in hospitalized patients who were NPO, on tube feedings, or total parenteral nutrition compared to conventional insulin therapy. The SQIA reduced rates of hypoglycemia and severe hyperglycemia, including in patients receiving glucocorticoids.
Background
Inpatient hyperglycemia is common and associated with adverse outcomes, yet achieving optimal glycemic control is challenging due to variable caloric intake, illness-related metabolic changes, and medication effects. Patients who are nil per os (NPO) or receiving continuous enteral or parenteral nutrition are particularly at risk for both hyperglycemia and hypoglycemia. Current guidelines recommend basal-bolus insulin regimens, but clinical inertia and logistical barriers often limit timely insulin adjustments. The SQIA was developed to automate insulin dose titration every 4 hours using rapid-acting insulin to maintain glucose within target ranges, aiming to improve safety and efficacy.
Data Highlights
Glycemic Range
SQIA (%)
Conventional Insulin (%)
Difference (%)
P Value
Hypoglycemia (<70 mg/dL)
0.65
1.10
-0.45
<.001
Severe Hyperglycemia (>250 mg/dL)
5.40
6.65
-1.25
0.002
Severe Hyperglycemia with High-Dose Glucocorticoids
Lower by 11.1%
Reference
-11.1
Not specified
Key Findings
The SQIA group had significantly fewer hypoglycemic events (0.65%) compared to conventional insulin (1.10%), with a difference of -0.45% (P < .001).
Severe hyperglycemia occurred less frequently in the SQIA group (5.40%) versus conventional insulin (6.65%), difference -1.25% (P = .002).
Patients receiving glucocorticoids, especially high-dose, had markedly lower severe hyperglycemia rates when managed with SQIA (11.1% reduction).
The SQIA reduced the need for physician insulin order adjustments by over 12-fold compared to conventional insulin therapy.
The algorithm continuously titrates rapid-acting insulin every 4 hours based on prior doses and glucose values to maintain glucose between 120 and 180 mg/dL.
Clinical Implications
The SQIA offers a practical and effective approach to inpatient glycemic management in patients who are NPO or receiving continuous nutrition, reducing both hypoglycemia and severe hyperglycemia. Its integration into the EMR streamlines insulin dosing adjustments, potentially decreasing clinician workload and improving patient safety. Adoption of such automated algorithms may enhance glycemic control and outcomes in complex inpatient populations.
Conclusion
Implementation of an automated self-adjusting subcutaneous insulin algorithm significantly improves inpatient glycemic control compared to conventional insulin regimens, reducing both hypoglycemia and severe hyperglycemia over a three-year period. This approach represents a valuable tool for managing challenging inpatient populations requiring precise insulin titration.
References
UCSF Clinical Research Article 2023 -- Automated Self-Adjusting Subcutaneous Insulin Algorithm for Inpatients
