A retrospective, observational study of patient outcomes for critically ill patients receiving parenteral nutrition.

Objective: To evaluate health care-related utilization for critically ill patients receiving parenteral nutrition (PN) administered via a premixed multichamber bag (MCB) or compounded solutions (COM).

Design: A retrospective database analysis of critically ill patients (intensive care unit stay ≥ 3 days) receiving PN and discharged between January 1, 2010, and June 30, 2011, using the Premier Hospital Database. Patients were identified as receiving MCB or COM on the basis of product description codes. Primary outcomes were length of stay (LOS) and total costs. Comorbidities and clinical outcomes were identified using International Classificaion of Diseases, Ninth Revision diagnosis codes. All costs reported were for inpatient services only. Patients receiving MCB and COM were matched on key patient and hospital characteristics using a propensity score methodology. Multivariate regression models for cost and LOS used generalized linear models with a log link and gamma distribution.

Results: A total of 42,631 patients met the inclusion criteria (MCB = 5,679; COM = 36,952), and the final matched population included 3,559 patients from each cohort. Baseline patient and hospital characteristics were well matched between groups. Adjusted multivariate models demonstrated a small difference between groups for LOS (MCB = 9.40 days vs. COM = 9.65 days; P = 0.014). In addition, patients receiving MCB incurred approximately 9.1% less in total costs (MCB = $37,790 vs. COM = $41,569; P < 0.001).

Conclusions: Overall, patients receiving MCB and COM experienced similar LOS, though patients receiving MCB had significantly lower overall costs. Interpretation of the study findings is subject to several limitations, and additional studies that include explicit identification of the method for compounding are needed.