The measurement performance of 13 biochemistry parameters (CEA, CA 19-9, amylase, lipase, sodium, potassium, chloride, creatinine, glucose, protein, albumin, LDH, triglycerides) was tested in a panel of biological fluids other than blood and urine (peritoneal, pleural, pancreatic fluids ...). Our protocol, based on a risk analysis, allowed us to justify our choices and compare the performance obtained with those of the serum or plasma matrix already validated. Thus, the coefficients of variation obtained in body fluids are comparable. The assessment of accuracy (spiking and dilution tests) shows the absence of bias, which is consistent with the absence of matrix effect. The linearity studied by dilution tests shows that the upper limits of the measurement interval communicated by the supplier are applicable to body fluids. The absence of contamination and stability have been also confirmed. All analytes are stable for 3 days at room temperature, 7 days between 2 and 8̊C, and 6 months at -20̊C; except LDH and lipase. For most analytes, at least one interference (hemolysis, icterus, lipemia) was found. Finally, a bibliographical study, confronted with the experience of prescribers, led us to define optimal thresholds to help interpret patients' results. In conclusion, this work has allowed us to validate analytical methods for body fluids testing after relying on their comparability to the blood matrix. We have also been able to adapt our practices and finally be accredited according to the standard NF IN ISO 15189.
Keywords: analytical validation; pancreatic fluid; peritoneal fluid; pleural fluid.