Standardized clinical criteria and sweat test combined as a tool to diagnose Cystic Fibrosis

Abstract

Context: CF is under-diagnosed in Ecuador; one out of every 11,252 live births born in Ecuador could have CF.

Aim: To analyze the clinical findings, based on previously established criteria, with the results of the sweat test, in circumstances where we do not have the routine molecular study.

Methods: Epidemiological, observational, analytic, cross-sectional study. It analyzed 180 patients clinically suspected of CF. Inclusion criteria: children of both sexes older than 30 days and younger than 12 years, who meet at least three clinical criteria suggestive for CF, outpatient and referred by a specialist physician who made a preliminary diagnosis. This is a pilot study.

Results: The combination of criteria pneumonia, chronic cough and chronic obstructive bronchial syndrome is the most frequent, with not a significant relationship with a positive sweat test. On the contrary, a significant relationship was found between the clinical combinations of pneumonia with cough and rhinosinusitis; pneumonia with cough; presence of Pseudomonas aeruginosa; and pneumonia with digital cough and clubbing, so it is recommended to perform the test in all these associations. The most frequent clinical criterion for the reference and performance of the electrolyte test in sweat is pneumonia to repeat for two or more episodes.

Conclusion: Clinical combinations of pneumonia with cough and rhinosinusitis; pneumonia with cough; presence of Pseudomonas aeruginosa; and pneumonia with digital cough and clubbing are pathognomonic for CF and indication for the sweat test. The predictive performance in CF diagnosis, defined as compatible clinical presence plus high values of chloride in sweat test, was 91.1%.

Keywords: Evidence-based medicine; Internal medicine; Paediatrics.

Fig. 1

ROC curve for predictive performance in CF diagnosis, defined as compatible clinical presence plus positive pilocarpine test, model evaluated; the total yield reached 91.1%, with an AUC: 0.963. The best cut point was reached with a probability p ≥ 0.25, the prediction model reached a total yield of 91.1%, with a sensitivity of 93.3% (95% CI: 77.9% to 99.2%); specificity of 90.7% (95% CI: 84.8% to 94.8%); a positive predictive value (PPV) of 66.7% (95% CI: 54.6% to 76.9%) and a negative predictive value of 98.6% (95% CI: 94.7% a 99.6%), total area under the curve (AUC) was calculated at 0.963.