Diagnostic accuracy of ancillary tests in diagnosis of cystic fibrosis and development of cystic fibrosis clinical diagnostic score: A multicentre prospective cohort study

Abstract

Background: Diagnostic tests for cystic fibrosis (CF) are not readily available in resource-limited settings, which leads to delay in diagnosis and treatment of children with CF.

Methods: In a multicentric prospective study, children with recurrent/persistent pneumonia, failure to thrive, or steatorrhea with suspicion of CF were enrolled. Sweat chloride concentration was measured to confirm the diagnosis of CF. Diagnostic accuracy of various clinical features, ancillary laboratory investigations (serum electrolytes, blood gas, stool fat globules), and aquagenic wrinkling for CF diagnosis was estimated. CF clinical diagnostic score (CF-CDS) was developed by combining significant parameters in stepwise logistic regression.

Results: Of 860 children enrolled, 313 (36.7 %) were diagnosed with CF. History of a sibling with CF, clubbing, hyponatremia, metabolic alkalosis, stool fat globule positivity, sputum Pseudomonas isolation, and aquagenic wrinkling within 3 min were found to be independently associated with a diagnosis of CF. CF-CDS score developed by combining these parameters demonstrated excellent diagnostic accuracy for diagnosis of CF [AUROC of 0.923 (95 %CI: 0.899, 0.946)]. At a cut-off of ≥2.5, CF-CDS had sensitivity and specificity of 87.64 % and 81.02 %, respectively.

Conclusion: CF-CDS has excellent diagnostic accuracy for diagnosis of CF in children and can be used to decide on starting treatment of CF pending confirmatory tests when confirmatory tests are not readily available.