Prevalence, clinical predictors and diagnosis of gastro-oesophageal reflux in children with persistent respiratory symptoms

Objectives: This was a cross sectional study conducted in the Paediatric Institute among infants and children with chronic respiratory symptoms with the following objectives: i) to determine the prevalence of gastro-oesophageal reflux in children with persistent respiratory symptoms, ii) to identify the clinical predictors of GOR (Gastro-oesophageal reflux) in children with persistent respiratory symptoms and iii) assess the validity of abdominal ultrasound, barium oesophagogram and chest radiograph in diagnosing GOR in these patients.
Materials and Methods: Forty-four patients were recruited over a period of six months. All the presenting symptoms were identified. The patients were subjected to chest radiograph, abdominal ultrasound, barium oesophagogram and 24-hour pH oesophageal monitoring.
The predictive validity of clinical symptoms, chest radiograph, abdominal ultrasound and barium oesophagogram were assessed. Twenty-four hours oesophageal pH was the gold standard to diagnose GOR.
Results: The mean age of patients was 9.1 months (1-58 months). Thirty-one patients (70.5%) were confirmed to have GOR by pH study. Respiratory symptoms alone were not useful to predict GOR. Cough had the highest sensitivity of 51.6%. stridor, wheeze and choking each had a specificity of 76%. Wheeze, vomiting, choking and stridor were identified to have high specificity (90-100%) in diagnosing GOR when any two symptoms were taken in combination.
Collapse/consolidation was the commonest radiological abnormality but had low sensitivity (35.5%) and specificity (53.8%). However hyperinflation on chest radiograph had a specificity of 92.3% with positive predictive value at 80% in diagnosing GOR. Barium oesophagogram has low sensitivity (37.9%) and moderate specificity (75%) in diagnosing GOR in children with respiratory symptoms.
Abdominal ultrasound was a valid mode of diagnosing GOR when there were three or more reflux episodes demonstrated during the screening period with a specificity of 90.9%. However the sensitivity was low ie 20-25%. The specificity increased to 90-100% when two positive tests were taken in combination (abdominal ultrasound and barium oesophagogram). However the sensitivity remained low (10-20%). Chest radiograph did not improve the predictive value when considered with the above tests. Combination of clinical symptoms were useful as clinical predictors of GOR. In the absence of a pH oesophageal monitoring, a combination of barium oesophagogram and ultrasound may be helpful in diagnosing GOR.