Background: According to currently accepted diagnostic criteria, ultrasonography confirms hypertrophic pyloric stenosis (HPS) when the pyloric muscle thickness (MT) is greater than 4 mm and the pyloric channel length (CL) is greater than 15 mm. Hypertrophic pyloric stenosis frequently presents in newborns younger than 21 days; yet, the diagnostic criteria in this younger population remain poorly defined. We, therefore, sought to define the diagnostic criteria for HPS in newborns younger than 21 days.
Methods: Ultrasonographic measures of pyloric MT and CL were obtained by retrospective chart review (2000-2006) at a single institution for all newborns (aged 10 days to 6 weeks) with an intraoperatively proven diagnosis of HPS. Demographic characteristics and ultrasonographic measurements were collected, and features differentiating younger (21 days or younger) from older newborns were assessed. Measures of pyloric MT and CL were analyzed in 7-day increments, and comparisons were made between newborns aged 21 days or less and newborns 22 to 42 days of age. Based upon these features, a set of ultrasonographic parameters to establish the diagnosis of HPS in younger patients was defined.
Results: Three hundred fourteen newborns (83% male) underwent pyloromyotomy of whom 64% (n = 200) had a preoperative pyloric ultrasound. Sixty newborns (19%) were younger than 21 days, of whom 51 (85%) had preoperative ultrasonography. The ultrasound measurement of HPS was significantly decreased in younger vs older newborns: (MT, 3.7 +/- 0.65 vs 4.6 +/- 0.82 mm, P < .05; CL, 16.9 +/- 2.8 vs 18.2 +/- 3.4 mm, P < .05). Importantly, the mean ultrasound measurement for young newborns with HPS typically fell within the currently defined "normal" or "borderline" range. A linear relationship was determined to exist between pyloric MT and CL and patient age, suggesting the use of 3.5 mm as a "cutoff" in younger patients.
Conclusions: These findings suggest that current guidelines to diagnose HPS do not accurately diagnose HPS in children younger than 3 weeks, and these findings raise the need to evaluate the decision analysis algorithm using prospective studies.