Dosage of calcium (Ca) and magnesium (Mg) by microwave plasma atomic emission spectrometry (MP-AES) in the control of individualized parenteral nutrition (IPN): routine implementation

Introduction
To guarantee the quality of our IPN, various checks (visual, gravimetric and analytical) are carried out. Among these controls, the sodium (Na) and potassium (K) assay is performed routinely by MP-AES. This analytical technique also allows the determination of other ions such as Ca and Mg present in our IPNs. Method validation (repeatability, linearity, intermediate precision, accuracy, inter-sample contamination, accuracy) has already been carried out for the 4 ions at 3 concentrations and 2 different wavelengths per element. However, after several tests, we encountered difficulties, particularly in the calibration of Ca and Mg with the 3 points range in place. We would like to add Ca and Mg assays to our routine to increase the level of security of our controls and further guarantee the quality of the IPNs.

Methods
We established a new calibration range with 5 points from the theoretical data of the last few months to target the range to potential results. Several analysis parameters (sampling time, stabilization and rinsing time, nebulization flow rates) have been redesigned to optimize analysis duration while maintaining efficient sample analysis. The analysis durations before and after modifications were measured. We then analyzed 30 samples under the same conditions as in routine. The concentrations obtained were compared with the theoretical value using the same compliance standards as for Na and K, that to say ± 10% (or ± 3 mmol/L if higher). The results are presented as the median [1st quartile; 3rd quartile].

Results
Of the 30 samples analyzed, we find a deviation from the theoretical value of 5.73% [3.29; 10.31] for Ca and 3.42% [1.26; 6.42] for Mg, that to say a concentration difference of 0.64 mmol/L [0.45; 1.41] for Ca and 0.06 mmol/L [0.03; 0.14] for Mg. There are 9 samples above 10% for Ca and 4 for Mg, but these samples are below the fixed 3 mmol/L (maximum 2.44 mmol/L for Ca, 0.71 mmol/L for Mg). All results comply with the applicable release conditions. The analysis duration goes from 140 seconds for the analysis of Na and K to 190 seconds for the 4 ions, it means 50 additional seconds per sample.

Conclusion
The results are conclusive but pose new constraints of time and organization of the routine. The 5 point range allows a better calibration. We will need to assess the stability of our calibration range to be able to prepare it a few hours in advance. We will discuss with the pediatricians of the neonatology department to decide on acceptable deviations and set the conditions for releasing IPNs for Ca and Mg. These new analyzes will be communicated to the outsourced establishments. The quality documents (procedures, working documents, etc.) will be updated.

See the poster