Background: The sweat test is the current gold standard for the diagnosis of cystic fibrosis (CF). CF is unlikely when sweat chloride (Clsw) is lower than 30mmol/L, Clsw>60 is suggestive of CF, with intermediate values between 30 and 60mmol/L. To correctly interpret a sweat chloride value, the biological variability of the sweat chloride has to be known.
Methods: Sweat tests performed in two centers using the classic Gibson and Cooke method were retrospectively reviewed (n=5904). Within test variability of Clsw was measured by comparing results from right and left arm collected on the same day. Between test variability was calculated from subjects with sweat tests performed on more than one occasion.
Results: Within test variability of Clsw calculated in 1022 subjects was low with differences between -3.2 (p5) and +3.6mmol/L (p95). Results from left and right arm were classified differently in only 3 subjects. Between test variability of Clsw in 197 subjects was larger, with differences between -18.2mmol/L (p5) and +14.1mmol/L (p95) between repeat tests. Changes in diagnostic conclusion were seen in 55/197 subjects, the most frequent being changing from indeterminate to 'CF unlikely' range (48/102).
Conclusion: Variability of sweat chloride is substantial, with frequent changes in diagnostic conclusion, especially in the intermediate range.
Keywords: Cystic fibrosis; Diagnosis; Variability.
Copyright © 2016 European Cystic Fibrosis Society. Published by Elsevier B.V. All rights reserved.