Detection and reduction of errors in parenteral nutrition compounding through gravimetric and product control.

OBJECTIVE: To analyze the errors in the preparation of parenteral nutrition in a Pharmacy Service, detected through an already consolidated gravimetric and product quality control, and compare them with those detected during the initial years of implementing this quality control.

METHODS: All errors detected through quality control in the compounding of pediatric and adult parenteral nutrition between 2019 and 2021 were prospectively analyzed. This quality control consisted of 3 sequential processes: a visual check, a gravimetric control, and a product control. Errors were classified as gravimetric, when the nutrition had a deviation of more than 5% from the theoretical weight, or as product errors when a qualitative or quantitative error was detected upon reviewing the remainder of the components used. These errors were analyzed in terms of type and the component involved. A comparison was made with the errors detected during the implementation phase of this quality control from 2016 to 2018.

RESULTS: A total of 41,809 parenteral nutritions were reviewed, and 345 errors were detected (0.83% of the preparations); of these, 59 errors were found in pediatric nutritions (0.68% of them), and 286 in adult nutritions (0.86% of them). Among these errors, 193 were of gravimetric nature, while 152 were detected through product control. The main components involved in product errors were electrolytes, primarily due to the addition of excessive volumes and the use of incorrect components. A significant absolute reduction of 0.71% (p < 0.05) in the total number of errors was observed when compared to the implementation phase. This reduction was consistent in both gravimetric errors (-0.59%) and product-related errors (-0.12%) (p < 0.05).

CONCLUSIONS: Comprehensive quality control of parenteral nutrition preparation is an easily implementable tool that effectively detected and prevented significant errors. Furthermore, its widespread adoption contributed to a reduction in the overall error count.