Shielding parenteral nutrition from light: does the available evidence support a randomized, controlled trial?

BACKGROUND: Exposure of total parenteral nutrition to ambient light induces the generation of peroxides, creating oxidant stress, which potentially compounds complications of prematurity. Photograph protection of total parenteral nutrition reduces the peroxide load and has been shown to be associated with nutritional and biochemical benefits in animals and humans. It is unclear whether this reduction in peroxides from total parenteral nutrition leads to a reduction in the complications of prematurity, such as bronchopulmonary dysplasia. Our hypothesis was that shielding total parenteral nutrition from ambient light is linked to clinical benefits.
OBJECTIVE: The purpose of this work was to determine whether photograph protection of total parenteral nutrition (light protected), as compared with no photoprotection (light exposed), reduces the occurrence of bronchopulmonary dysplasia or death in preterm infants.
METHODS: The Canadian Neonatal Network provided data for infants born in 2006 at <28 weeks' gestation admitted to level 3 NICUs in Canada. A retrospective analysis was performed comparing bronchopulmonary dysplasia and death in infants who received light-exposed or light-protected parenteral nutrition. Data were analyzed by using logistic regression models. RESULTS. Thirteen NICUs offered partial light-protected (total parenteral nutrition bag only, intravenous tubing exposed) and 13 offered light-exposed parenteral nutrition; not a single NICU offered complete light-protected parenteral nutrition (total parenteral nutrition bag plus intravenous tubing). The incidence of bronchopulmonary dysplasia or death was 66% with light-protected (n = 428) vs 59% with light-exposed (n = 438) parenteral nutrition.
CONCLUSIONS: Partial photograph protection of total parenteral nutrition was not associated with a reduction in bronchopulmonary dysplasia or death as compared with no photograph protection; this relationship is confounded by covariates with strong associations with bronchopulmonary dysplasia. Partial photograph protection of total parenteral nutrition solutions confers no clinical benefit, while consuming valuable resources. A randomized, controlled trial is justified to determine whether there is a true "cause-and-effect" relationship between complete photoprotection of total parenteral nutrition and bronchopulmonary dysplasia or death.