Adjusting Automated Insulin Delivery Pumps During Short-Course Steroids in Chemotherapy
For patients on AID systems receiving short courses of steroids during chemotherapy cycles, increase total daily insulin by 50-100% (targeting 1.0-1.3 units/kg/day), distributed as 25-30% basal and 70-75% prandial insulin, with daily titration based on glucose patterns. 1
Immediate Adjustments at Steroid Initiation
Increase total daily insulin dose substantially at the start of each steroid cycle:
- Target 1.0-1.3 units/kg/day total insulin requirement 1
- This represents approximately a 50-100% increase from baseline for most patients
- The distribution should shift toward prandial coverage: 70-75% as bolus/prandial insulin and 25-30% as basal 1
Modify AID system parameters that can be adjusted:
- Increase the glucose target setpoint if your system allows this adjustment 2
- Increase insulin-to-carbohydrate ratios (more aggressive) by 30-50% 1
- Decrease correction factors (more aggressive corrections) by 30-50% 1
- Some systems allow adjustment of "insulin sensitivity factor" - decrease this value to deliver more insulin 2
System-Specific Considerations
For hybrid closed-loop systems (Tandem Control-IQ, Omnipod 5):
- These systems automatically adjust basal insulin but require manual meal boluses 3
- You must manually increase meal bolus doses by 50-100% during steroid cycles 1
- The automated basal adjustments will help but are insufficient alone for steroid-induced hyperglycemia 2
For Beta Bionics iLet (minimal user input system):
- This system makes autonomous dosing decisions with minimal carbohydrate counting 3
- The algorithm should adapt over 2-3 days, but may lag behind rapid steroid-induced insulin resistance 3
- Consider temporarily switching to a system requiring more manual control, or reverting to manual mode if hyperglycemia persists 2
Daily Titration Protocol
Titrate insulin daily during the steroid cycle:
- Review glucose data every 24 hours 1
- If >80% of glucose values remain >250 mg/dL, increase total daily insulin by 20% 1
- If time-in-range (70-180 mg/dL) is <50%, increase insulin by 10-20% 4
- Continue daily adjustments until achieving >60% time-in-range 1
Monitor for hypoglycemia as steroids wear off:
- Hypoglycemia risk increases as steroid effects diminish (typically 12-24 hours after last dose) 1
- In the study of hyper-CVAD patients, only 1.3% of glucose values were <70 mg/dL despite aggressive insulin dosing 1
- The AID system's automatic basal suspension provides some protection, but manual boluses cannot be automatically reduced 2
Cycle-to-Cycle Learning
Use data from previous cycles to preemptively adjust:
- Glucose control improves with subsequent steroid cycles (81% hyperglycemia in first cycle vs 60% in later cycles) 1
- Insulin requirements remain relatively stable across cycles (1.0-1.3 units/kg/day) 1
- Preemptively increase insulin doses at the start of subsequent cycles based on prior cycle requirements 1
Critical Pitfalls to Avoid
Do not rely solely on automated adjustments:
- AID systems are designed for typical day-to-day variability, not the dramatic insulin resistance caused by high-dose steroids 2
- The algorithms will increase insulin delivery, but not rapidly or aggressively enough for steroid-induced hyperglycemia 2
- Manual intervention with increased meal boluses is essential 1
Do not underdose prandial insulin:
- The 70-75% prandial distribution is critical because steroids cause predominantly postprandial hyperglycemia 1
- Standard 50/50 basal-bolus splits are inadequate during steroid therapy 1
Prepare for system reversion:
- Have a backup plan for manual insulin injection therapy if the AID system cannot keep pace with insulin requirements 2
- Stock rapid-acting and long-acting insulin, syringes/pens, and glucose monitoring supplies 2
- Patients may need to temporarily switch to manual mode or open-loop pump therapy 2
Monitoring Intensity
Increase glucose monitoring frequency:
- Check CGM data every 4-6 hours during waking hours 1
- AID systems rely on CGM accuracy, which can be compromised during chemotherapy 2
- Confirm CGM readings with fingerstick glucose if values seem discordant with symptoms 2
Watch for infusion site failures: