OBJECTIVES: In secretin-stimulated pancreatic function testing, the standard technique for bicarbonate measurement is back titration (BT). Chemistry autoanalyzers (AAs) automate bicarbonate measurement and are universally available; however, this method has not been validated in pancreatic fluid. The aims of the study were (1) to compare the AA and BT for measurement of bicarbonate in pancreatic fluid and (2) to determine the effects of variable conditions on bicarbonate concentration. METHODS: Pancreatic fluid samples were analyzed for bicarbonate concentration using both BT and AA. Additional experiments were conducted to determine the effect of different sample conditions. RESULTS: The results obtained by BT and AA were highly concordant (Lin concordance coefficient, 0.96). An AA cutpoint of 84 mM optimized sensitivity (91.1%) and specificity (94.1%) compared with the BT reference standard. Blood contamination spuriously increased the bicarbonate, an effect that was eliminated by centrifugation. The bicarbonate levels were not significantly affected by delayed analysis, storage condition, nitrogen gas treatment, or the addition of mineral oil or protease inhibitors. CONCLUSIONS: The bicarbonate concentrations obtained by AA are highly concordant with those obtained by BT. Secretin pancreatic function testing fluid analysis may now be performed in any hospital.
OBJECTIVES: In secretin-stimulated pancreatic function testing, the standard technique for bicarbonate measurement is back titration (BT). Chemistry autoanalyzers (AAs) automate bicarbonate measurement and are universally available; however, this method has not been validated in pancreatic fluid. The aims of the study were (1) to compare the AA and BT for measurement of bicarbonate in pancreatic fluid and (2) to determine the effects of variable conditions on bicarbonate concentration. METHODS: Pancreatic fluid samples were analyzed for bicarbonate concentration using both BT and AA. Additional experiments were conducted to determine the effect of different sample conditions. RESULTS: The results obtained by BT and AA were highly concordant (Lin concordance coefficient, 0.96). An AA cutpoint of 84 mM optimized sensitivity (91.1%) and specificity (94.1%) compared with the BT reference standard. Blood contamination spuriously increased the bicarbonate, an effect that was eliminated by centrifugation. The bicarbonate levels were not significantly affected by delayed analysis, storage condition, nitrogen gas treatment, or the addition of mineral oil or protease inhibitors. CONCLUSIONS: The bicarbonate concentrations obtained by AA are highly concordant with those obtained by BT. Secretin pancreatic function testing fluid analysis may now be performed in any hospital.