Summary: In a crossover trial in adolescents with type 1 diabetes on pumps, two extended-bolus algorithms for a mixed meal both kept post-meal glucose within target range; the Sieradzki method gave numerically longer time in tight range and fewer late lows, but that difference was not statistically significant, while one-hour glucose was modestly higher with the Pankowska method.
PICO Summary
| Element | Detail |
|---|---|
| Population | 58 adolescents with type 1 diabetes (median age 15.5) on insulin pumps and CGM in non-automated mode; randomised, double-blind, crossover trial, Poland. |
| Intervention | Sieradzki Equation (30% × carbohydrate units × insulin-to-carb ratio) as a 4-hour extended bolus for a standardised mixed meal. |
| Comparison | Pankowska Equation (fat-protein units × insulin-to-carb ratio) as a 4-hour extended bolus. |
| Outcome | Both methods kept postprandial glucose within the recommended time in range. The Sieradzki method gave longer time in tight range (82.5% vs 70.5%) and fewer hypoglycaemic episodes at 180 and 300 minutes, but the tight-range difference was not statistically significant (p=0.63). Glucose at 60 minutes was higher after Pankowska (136 vs 124 mg/dL; p=0.016). |
Sieradzki vs Pankowska extended bolus for mixed meals
RCT crossover · pediatric type 1 diabetes · 4-hour bolus
Both extended-bolus algorithms kept post-meal glucose in target. The simpler Sieradzki method trended toward longer tight-range time but the difference was not significant; Pankowska gave a modestly higher one-hour glucose.
Expert Commentary
This is a well-conducted, gold-standard crossover trial on a genuinely difficult problem, the delayed glucose rise after fat- and protein-rich meals that simple carbohydrate counting cannot cover. The honest headline, though, is that both algorithms worked and that the trial does not establish one as clearly superior. The eye-catching tight-range difference, 82.5% versus 70.5%, did not reach statistical significance at p=0.63, so it should be read as a numerical trend rather than a proven advantage, and the only statistically significant between-method finding was a modestly higher one-hour glucose with the Pankowska approach. What can be said fairly is that the simpler Sieradzki calculation performed at least as well as the more complex fat-protein-unit method, which matters because simplicity aids real-world adoption by families. Limitations include standardised meals that do not capture real-life variability, a single mixed-meal composition, and non-automated pump mode, so results may differ with automated insulin delivery. Can I use this with my patients? Yes, pragmatically. For adolescents on pumps struggling with high-fat meals, it supports using an extended bolus and suggests the easier 30% method is a reasonable default, while I would individualise the dose and duration using CGM feedback rather than treating either equation as definitively better.
References
Dymińska M, Kowalczyk-Korcz E, Piechowiak K, Szypowska A. The impact of two different insulin dose calculation methods on postprandial glycemia after a mixed meal in children with type 1 diabetes: a randomized study. Nutrients. 2025;17(20):3287. doi:10.3390/nu17203287