Validation of a nutrition screening tool for pediatric patients with cystic fibrosis / Miriam Isabel Souza dos Santos Simon, Gabriele Carra Forte, Juliane da Silva Pereira, Elenara da Fonseca Andrade Procianoy, Michele Drehmer

Background In cystic fibrosis (CF), nutrition diagnosis is of critical relevance because the early identification of nutrition-related compromise enables early, adequate intervention and, consequently, influences patient prognosis. Up to now, there has not been a validated nutrition screening tool that takes into consideration clinical variables. Objective To validate a specific nutritional risk screening tool for patients with CF based on clinical variables, anthropometric parameters, and dietary intake. Design Cross-sectional study. The nutrition screening tool was compared with a risk screening tool proposed by McDonald and the Cystic Fibrosis Foundation criteria. Participants/setting Patients aged 6 to 18 years, with a diagnosis of CF confirmed by two determinations of elevated chloride level in sweat (sweat test) and/or by identification of two CF-associated genetic mutations who were receiving follow-up care through the outpatient clinic of a Cystic Fibrosis Treatment Center. Main outcome measures Earlier identification of nutritional risk in CF patients aged 6 to 18 years when a new screening tool was applied. Statistical analyses performed Agreement among the tested methods was assessed by means of the kappa coefficient for categorical variables. Sensitivity, specificity, and accuracy values were calculated. The significance level was set at 5% (P<0.05). Statistical analyses were carried out in PASW Statistics for Windows version 18.0 (2009, SPSS Inc). Results Eighty-two patients (49% men, aged 6 to 18 years) were enrolled in the study. The agreement between the proposed screening tool and the tool for screening nutritional risk for CF by the McDonald method was good (κ=0.804; P<0.001) and the sensitivity and specificity was 85% and 95%, respectively. Agreement with the Cystic Fibrosis Foundation criteria was lower (κ=0.418; P<0.001), and the sensitivity and specificity were both 72%. Conclusions The proposed screening tool with defined clinical variables promotes earlier identification of nutritional risk in pediatric patients with CF.

Nutrition.