Background Polyvinyl chloride (PVC) infusion tubes are too widely used in our hospital. Even in intensive care units (ICU), PVC is commonly used although polyethylene (PE) is specifically recommended to limit container-content interactions. For example, insulin is used to achieve tight glucose control in ICU patients. Its adsorption on PVC is well known, but rarely considered in the choice of infusion device when nurses compose their infusion assemblies.

Purpose This model of interaction was used as a tool, to educate nurses in the issue of choosing the right equipment and to illustrate the clinical impact of an unsuitable medical device in the administration of a drug. For this, we showed healthcare workers how the insulin concentration changed by comparing the infusion between two devices available in our hospital.

Materials and methods ICU nurses prepared 40 mL insulin syringes (1 IU/mL) according to the ICU preparation protocol. The concentration was determined by UV-spectrophotometry at 4 points in the syringes to assess their homogeneity and dose accuracy. Then, PVC and PE syringe extension lines (SEL, L:150 cm, ID:1 mm) used in our hospital were evaluated. Firstly, each SEL was filled with 1 IU/mL insulin solution and closed with Luer-lock plugs. Measurements were achieved after 15 and 60 min of contact. Secondly, the insulin levels were measured at the egress of both SEL every 5 min over a 24 h time period at a flow rate of 2 mL/h. Results were expressed as a percentage of the theoretical value. All experiments were repeated 5 times. Statistical comparisons were performed with a Mann-Whitney test (p = 0.05).

Results Eleven nurses prepared 48 insulin syringes with an average proportion of 100.6 ± 3.9% of the theoretical value in all points of the syringes.

After 15 min of contact, the insulin levels dropped significantly in PVC SEL compared to PE SEL (46.2 ± 2.4% vs 98.4 ± 1.7%, p = 0.01, respectively). The level continued to drop at 60 min (30.1 ± 4.0% vs 96.0 ± 0.6%, p = 0.01, respectively).

During infusion tests, the insulin concentration after a 24 h infusion was only 64.6 ± 2.0% using PVC SEL whereas it was maintained at 101.7 ± 0.8% with PE SEL (p = 0.01).

Conclusions A paradoxical situation exists in the insulin infusion context. Although this interaction with PVC is well documented and confirmed by this study, ICUs still use it. This study allowed us to demonstrate that even if the preparation process is performed correctly and the concentration is homogeneous, PE SEL are really more suitable for insulin administration and must replace PVC SEL. The use of such a tool may help pharmacists in the ongoing education of nurses. Other drugs are being considered to enhance the study and to support our infusion training group. An evaluation of the impact of our training course on daily practice is planned.

No conflict of interest.