Intravenous Neonatal Central Access Safety Trial

BACKGROUND Particulate contamination due to infusion therapy (administration of parenteral nutrition and medications) carries a potential health risk for infants in neonatal intensive care units (NICU).

This particulate consists of metals, drug crystals, glass fragments or cotton fibres and can be generated by drug packaging, incomplete reconstitution and chemical incompatibilities.

Filters have been shown to remove micro-organisms, endotoxin, air and particles in critically-ill adults and older infants, but its benefits in newborn remain to be demonstrated.

Although recommendations for the use of in-line intravenous filters have been published, there is no consensus on their use.

Moreover, 50% of inflammatory episodes in the setting of NICU are blood culture-negative. These episodes could be partly related to the presence of particles in the infusion lines.

PROJECT AIMS AND DESIGN:

Aim of this multicenter trial is to evaluate the effectiveness of in-line filtration in reducing culture-negative inflammatory episodes in infants admitted to NICU. Further aim is to evaluate the efficacy in the reduction of main CVC-associated complications when using filters.

PATIENTS:

All infants admitted to the NICUs are considered eligible for inclusion into the trial if prolonged infusion therapy (one week or more) is expected, with either umbilical vein catheters (UVC) in a central position or percutaneously inserted central venous catheter. Infants in whom a peripheral UVC is placed but for whom infusion therapy via central venous access is planned will be enrolled too.

Exclusion criteria: Infants will be excluded from participation in the trial if they have clinical characteristics requiring transfer to units not participating in the study before discontinuation of infusion therapy (neurological or surgical diseases, chromosomal abnormalities, and major malformations)..

PROTOCOL:

After randomization each infant will be subsequently allocated to experimental group (Filter) or to control group (Control) as per randomization. Each research unit will refer to its own protocols for infection sulveillance and prevention, although respecting some minimal standard criteria and indications, common and approved by all research units.

In the filter group, all infusions, with the exception of some solutions (eg blood products), will be subjected to filtration. The aqueous solutions (parenteral therapy and drugs) will be administered through 0.2 μm filters which will be replaced every 96 h; the lipid emulsions will be administered through 1.2 μm filters which will be replaced every 24 h. In case of emergency, life-saving drugs will be administered with bolus modality though the infusion line closer to the patient without the need for filtration. In case of drugs/solutions not supported by filtration (eg blood products), they will be administered through a dedicated unfiltered access, which will be removed as soon as the drug is no longer needed.

In the control group, all infusion will be administered through unfiltered accesses.

Data will be collected daily from enrolment up to 48 h after discontinuation of infusion therapy.

In case of inflammatory episode, all patients will undergo to defined specimens, as always in correct clinical practice: complete blood count, sepsis biomarkers (C- reactive protein, procalcitonin, presepsin) and blood culture.

At discharge information regarding the main neonatal pathologies will be recorded.

MAIN OUTCOME Frequency of patients with at least one inflammatory episode sepsis-like, defined by alteration of the biomarkers of inflammation in a negative-culture contest.

SECONDARY OUTCOMES

• Frequency of patients with at least one inflammatory episode defined by alteration of the biomarkers of inflammation in a positive-culture contest.
• Occurrence of phlebitis or local cutaneous inflammation
• Occurrence of luminal obstruction and/or extravascular fluid effusion
• Duration of mechanical ventilation
• Number of catheter days
• Length of stay
• Neonatal mortality

SAMPLE SIZE:

The baseline risk of inflammatory states in the target population remains undetermined. However, we hypothesize a range between 30% and 35%. Consequently, a median risk of 32.5% was assumed for the control group. With the application of filters, a plausible 30% risk reduction is anticipated, resulting in an estimated risk of 22.75% in the intervention arm. Utilizing Fisher's exact test to compare two independent proportions, with an alpha of 0.05, a power of 0.80, and a 1:1 group allocation, the calculated minimum sample size required for significance is 349 infants for each arm, leading to a total of N=698 infants. Accounting for an estimated 5% dropout rate during follow-up, the adjusted minimum sample size becomes N=736 infants"

DATA ANALYSIS:

Data will be analysed according to an intention-to-treat model. Therefore, data from all infants enrolled into the study will be considered for the analysis. Death and transfer to another hospital before discontinuation of infusion therapy, are the only two reasons for exclusion. The primary outcome will be evaluated by Fisher's exact test. Secondary outcomes will be evaluated by Fisher's exact test or appropriate generalized linear models.

EXPECTED RESULTS AND IMPACT ON CLINICAL PRACTICE:

If the use of in-line filters resulted in a significant decrease in negative-culture inflammatory episodes and/or in any other complications, the use of in-line filters in all intravenous administration systems may be recommended in NICU.