Inter-reflux and bolus clearance times in non-pathologic pediatric patients: data support computational models

Previous contributions suggested that gastroesophageal reflux can be modeled in terms of a Poisson process. This study aims to provide empirical data to validate this statement in pediatric patients so that computational models can be broadly used as an alternative for research. A retrospective review of 63 pediatric patients who underwent 24-hour impedance-pH monitoring to discard gastroesophageal reflux disease was conducted in this study. Patients were grouped by age as preterm (21), infants (21), and children (21). All the tracings were analyzed by a trained physician who identified the reflux entry and the bolus clearance time of each episode. The time between reflux episodes was tested against three probability distributions (gamma, exponential and inverse Gaussian) whereas the bolus clearance time was tested against a normal probability distribution. Parameters were estimated using the maximum likelihood method. The Kolmogorov-Smirnov test and the Kullback-Leibler divergence were computed to evaluate the goodness of fit. One-way analysis of variance was applied to compare results along the three groups. Exponential fitting for inter-reflux time was successful in 90.48% of children older than 1 year. The overall reflux rate was 57.58 reflux episodes per day whereas the mean bolus clearance time ranged between 10.87 in preterm subjects and 12.05 in children, showing a good Gaussian fitting. The time between reflux episodes can be modeled in terms of a Poisson process in non tube-fed patients, whereas the bolus clearance time follows a normal distribution in all cases.