Searchable abstracts of presentations at key conferences in endocrinology
Endocrine Abstracts (2014) 35 P488 | DOI: 10.1530/endoabs.35.P488

ECE2014 Poster Presentations Diabetes therapy (40 abstracts)

Basal bolus insulin therapy in hospitalised patients with diabetes mellitus type 2 using two algorithms embedded in a tablet PC

Julia Mader 1 , Katharina Neubauer 1 , Felix Aberer 1 , Lukas Schaupp 1 , Klaus Donsa 2 , Stephan Spat 2 , Hoell Bernhard 2 , Thomas Augustin 2 , Peter Beck 2 , Johannes Plank 2 & Thomas Pieber 1,


1Endoccrinology and Metabolism, Medical University of Graz, Graz, Austria; 2Joanneum Research, Graz, Austria.


Current guidelines recommend pre-meal blood glucose (BG) levels of <140 mg/dl in the hospital. The aim of this analysis was to compare two versions of a workflow-integrated algorithm for basal-bolus insulin therapy (REACTION algorithm) for glycemic management in patients with diabetes mellitus type 2 (T2D) hospitalized at the general ward.

For both algorithms BG measurements were performed four times daily (pre-breakfast, pre-lunch, pre-dinner, and at bedtime); insulin injections were given according to the algorithm advice algorithm. A basal-bolus regimen with advice for total daily dose (TDD) (50% basal insulin, 50% pre-meal bolus insulin with additional corrective dose if necessary) was generated once daily. In the refined algorithm bolus insulin dose was redistributed over the day wheareas total daily dose and the 50:50 ratio remained unchanged.

Each algorithm was applied in 15 T2D patients (initial algorithm: four females, age 69±10 years, HbA1c 76±30 mmol/mol, BMI 29±6 kg/m2; refined algorithm: seven females, age 73±11 years, HbA1c 62±18 mmol/mol, and BMI 30±7 kg/m2). Mean BG was 163±34 mg/dl (initial algorithm) vs. 148±25 mg/dl (refined algorithm). 6/456 (1.3%) and 7/457 (1.5%) measurements were in the hypoglycaemic range (<70 mg/dl), initial and refined algorithm respectively. In both groups no value was below 40 mg/dl. Mean TDD was 47±28 U (basal: 20±13 U and bolus: 27±16 U) for the initial algorithm and 47±27 U (basal: 22±12 U and bolus: 25±15 U) for the refined algorithm. Adherence to the insulin advices by the algorithm was 113/115 (98.3%) and 110/111 (99.1%) for TDD, 104/106 (98.1%) and 101/107 (94.4%) for basal insulin and 374/393 (95.2%) and 269/279 (96.4%) for bolus insulin (initial and refined algorithm respectively).

The refined version of the Reaction algorithm could improve glycaemic control without increased risk of hypoglycaemia. Adherence to insulin dosing advices generated by both algorithms was high. Insulin doses were comparable for the two versions of the algorithm. The reaction algorithm has the potential to improve glycaemic management in the hospital setting.

Article tools

My recent searches

No recent searches.