Background and importance Glycaemic alterations are highly prevalent in premature newborns during the first days of life. Particularly, hyperglycaemia has been reported as an independent risk factor for increased mortality and morbidity. This condition requires the use of insulin infusions for its treatment. However, this drug presents problems of adsorption to plastic that is intensified with low insulin concentrations and infusion rhythms used in neonates, conditioning the decrease and also the ignorance of the doses actually administrated to the newborn. There is no consensus on the appropriate insulin preparation and administration.

Aim and objectives To determine the combination of variables for the preparation and administration of insulin infusions that provides higher accuracy and lower probability of error.

Material and methods An experimental study was carried out with the aim of determining which variable (additive (albumin yes/no), solution (sodium chloride solution, NaCl 0.9%/glucose 5%), operator (1/2), preconditioning (yes/no), purge (yes/no), concentration (0.05–0.1 UI/mL), infusion rate (0.3–0.7 mL/h) and infusion duration (1 hour/24 hours)) most influences the concentration and dose of insulin administrated. The determinations were made with immunoassay using IMMULITE 1000 equipment. Previously, an ad hoc calibration was developed, adjusted to the range of doses commonly used in neonatal insulin infusions. Finally, a screening model using Plackett-Burman designs was developed to calculate insulin recovery and determine the variables with the most influence.

Results 24 experimental infusions were made, using combinations of different variables. After analysing the total of the samples, each of the recovery values obtained were entered in the screening model. The variables that achieved higher insulin recovery values were the additive (albumin - yes) and the solution (NaCl 0.9%). The model can explain 48.16% of the variation in insulin recovery, in which the additive has a standardised effect four times greater and the solution two times greater than the rest of the variables that do not exceed 1. (Figure 1: Pareto diagram)

Conclusion and relevance The additive and the solution seem to be the most important determining factors for the recovery of insulin in the preparation of the infusions. The addition of albumin and preparing the infusions with sodium chloride solution 0.9% as solution results in a greater recovery of insulin.

Conflict of interest No conflict of interest