The European Diagnostic Working Group presented comprehensive diagnostic algorithms for cystic fibrosis (CF) and confirmed the fundamental role of the sweat test for the diagnosis of CF.¹ However, several important differences between well-accepted guidelines for sweat testing^2,3 and the recommendations of the Working Group need to be discussed.

An adequate sweat sampling volume depends on the sampling area and not on the body surface area of the patient.^2,3 The unit therefore has to be cited as “g/m² sampling surface area/min sweat sampling time”^2,3 instead of “g/m² body surface area/min”.¹

For stimulation and sampling of sweat the authors recommend only the Gibson and Cooke technique and do not even mention the widely used Macroduct collection method which is well accepted by the National Committee for Clinical Laboratory Standards (NCCLS) and UK guidelines.^2,3 The authors do not give any reason for this limitation. Mastella et al⁴ have shown an acceptable agreement between both collection systems with a mean (SD) difference for chloride of 1.205 (6.47) mmol/l,⁴ comparable to Denning’s results which showed a mean (SD) difference between sequential Gibson-Cooke tests of −0.05 (8.6) mmol/l.⁵ Different failure rates, especially in patients under 4 months of age, should not be misused to condemn the Macroduct collection system⁴ because this problem can be overcome by experience.

The most important difference is the extension of the intermediate sweat chloride range up to 30–60 mmol/l from 40–60 mmol/l. This recommendation is based on the work of Lebecque and co-workers⁶ who investigated patients with sweat chloride levels of 30–60 mmol/l by extensive genetic testing and nasal potential difference measurements. Adults accounted for 30% of all patients with intermediate sweat chloride levels but were excluded from the analysis. Lebecque et al presented 10 children with intermediate sweat chloride levels and a diagnosis of CF. However, only 2 of the 10 patients (sweat chloride levels of 34 and 45 mmol/l) fulfilled the clinical criteria and laboratory evidence of CFTR dysfunction, according to the diagnostic criteria of the Cystic Fibrosis Consensus Panel.⁷ The other 8 patients had no clinical features of CF and/or no laboratory evidence of CFTR dysfunction in accordance with the diagnostic criteria of the Cystic Fibrosis Consensus Panel.⁷ As shown by Lebesque⁶ and Denning,⁵ the extension of the intermediate range can more than double the number of patients who will need further diagnostic investigations. Before such a far-reaching recommendation of the expansion of the intermediate chloride range is implemented in daily routine, prospective (not only retrospective) studies are urgently needed to define the specificity and sensitivity of this modification. As we all know, even a chloride level of <30 mmol/l cannot exclude the diagnosis of CF.