Variability in Parenteral Nutrition Use in US Children's Hospitals.

Background: The variability of parenteral nutrition (PN) use for pediatric inpatients is currently unknown. In this study, we aim to determine the variability in PN use in US children's hospitals and the association of PN initiation with inpatient PN use.

Methods: We performed a retrospective cohort study of children who received PN during an inpatient encounter in US children's hospitals. Hospitals were divided into tertiles based on their rates of PN use: low (<36.9 of 1000 encounters), medium (36.9-51.8 of 1000 encounters), and high (>51.8 of 1000 encounters). Multivariable regression models were developed to assess the associations between hospital PN use and time to PN initiation, PN duration, and encounter length of stay after adjustment for salient patient characteristics.

Results: The cohort included 82,142 patients receiving PN, and rates of hospital PN use ranged from 5.9 to 76.7 patients receiving PN per 1000 inpatient encounters. After multivariable adjustment, patients treated at high-use hospitals had a significantly shorter time to initiation of PN compared with low-use hospitals (incident rate ratio [95% CI]: 0.78 [0.69-0.89]; P < .001). There was no significant association between low- and medium- or high-use hospitals regarding PN duration or hospital length of stay.

Conclusion: Large variation in PN use exists among US children's hospitals. High-use hospitals are more likely to start PN earlier but do not have longer PN duration or encounter length of stay. This variability makes PN use an ideal target for hospital quality improvement efforts to improve adherence to PN evidence-based guidelines.