Sahoo N, Dhochak N, Jat KR, Sankar J, Lodha R, Sethuraman G, Kabra M, Kabra SK. Development of algorithm for diagnosis of cystic fibrosis in absence of sweat chloride testing in resource-limited setting.
Pediatr Pulmonol 2022;
57:3077-3083. [PMID:
36062940 DOI:
10.1002/ppul.26137]
[Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [Abstract] [Key Words] [Track Full Text] [Journal Information] [Submit a Manuscript] [Subscribe] [Scholar Register] [Received: 03/04/2022] [Revised: 08/11/2022] [Accepted: 08/14/2022] [Indexed: 01/11/2023]
Abstract
OBJECTIVE
To develop a diagnostic algorithm for cystic fibrosis (CF) in the setting of unavailability of sweat chloride, based on clinical features and basic laboratory investigations.
METHODS
In a prospective observational study, we enrolled children with recurrent/persistent pneumonia with either malabsorption or poor growth, undergoing a sweat chloride test, between January 2019 and December 2020. They were simultaneously evaluated for aquagenic wrinkling of hands, stool fat globules, sputum for bacterial culture, blood gas, and serum electrolytes. Sensitivity and specificity were calculated for parameters having a significant difference between CF and non-CF groups. Scoring systems and algorithms for the diagnosis of CF were developed.
RESULTS
Of 134 children enrolled, 46 (34%) had CF. The sensitivity and specificity of various parameters to diagnose CF was: sibling death due to respiratory illness (30.43%, 96.59%), aquagenic wrinkling (76.74%, 47.67%), metabolic alkalosis (17.78%, 94.12%), hyponatremia (28.89%, 89.41%), stool fat globules (38.46%, 81.18%), and presence of Pseudomonas in sputum culture (23.68%, 98.80%). Using coefficients of significant parameters on stepwise logistic regression, the composite score for diagnosis of CF was calculated as: 3X sibling death due to respiratory illness + 1.5X hyponatremia + 1.5X metabolic alkalosis + 1.5X aquagenic wrinkling + 1X stool fat globules + 2.5X presence of Pseudomonas in sputum culture (each of the variables scores 0 or 1 for absence and presence, respectively). The cut-off of ≥2.5 had sensitivity and specificity of 81.82% and 76.83%, respectively.
CONCLUSIONS
In resource-limited settings, the proposed diagnostic algorithm can be used for the diagnosis of presumptive CF with fair sensitivity and specificity.
Collapse