Statistical analyses
Statistical analyses were performed with Stata software (version 15,
StataCorp, College Station, Texas, USA) and R 3.5.1
(http://cran.r-project.org/). All tests were two-sided with an alpha
level set at 5%. Categorical variables were expressed as number of
children and associated percentages, and continuous variables as
mean ± standard deviation or as median [25th;
75th percentiles], according to the statistical
distribution and the results previously reported in the literature.
Factors associated with OAHI, RS30 and RS10 were studied with:
Mann-Whitney test for binary variables, Kruskal-Wallis test for
categorical variables with more than two categories and Spearman’s rank
correlation coefficient for continuous variables. To assess whether RS30
and RS10 can be predictive of an OAHI ≥ 1.5 and ≥ 5, receiver operating
characteristic (ROC) curves were plotted. The area under the ROC curves
were presented with a 95% confidence interval (CI) obtained with the
technique of DeLong et al[21]. Finally, Youden’s index was
calculated to study the “optimal” threshold value of RS30 and RS10 to
predict an OAHI ≥ 1.5 and ≥ 5. It was presented with its specificity,
sensitivity, and negative and positive predictive values.