Population-level surgical outcomes for infantile hypertrophic pyloric stenosis

Abstract

Objectives: Determine national outcomes for pyloromyotomy; how these are affected by: (i) surgical approach (open/laparoscopic), or (ii) centre type/volume and establish potential benchmarks of quality.

Methods: Hospital Episode Statistics data were analysed for admissions 2002-2011. Data presented as median (IQR).

Results: 9686 infants underwent pyloromyotomy (83% male). Surgery was performed in 22 specialist (SpCen) and 39 nonspecialist centres (NonSpCen). The proportion treated in SpCen increased linearly by 0.4%/year (r=0.76, p=0.01). Annual case volume in SpCen vs. NonSpCen was 40 (24-53) vs. 1 (0-3). Time to surgery was shorter in SpCen (1day [1, 2] vs. 2 [1-3]), but total stay equal (4days [3-6]). 137 (1.4%) had complications requiring reoperation (wound problem 0.6%; repeat pyloromyotomy 0.5% and perforation, bleeding or obstruction 0.2%): pooled rates were similar between SpCen and NonSpCen (1.4% vs. 1.6%, p=0.52). Three NonSpCen had >5% reoperations (within 99.8% C.I. as small denominators). There was no relationship between reoperation and centre volume. Laparoscopic pyloromyotomy had increased risk of repeat pyloromyotomy (OR 2.28 [1.14-4.57], p=0.029).

Conclusions: Pyloric stenosis surgery shifted from centres local to patients, but outcomes were unaffected by centre type/volume. Modest reported benefits of laparoscopy appear offset by increased reoperations. Quality benchmarks could be set for reoperation <4%.

Type of study: Treatment Study.

Level of evidence: Level III.

Keywords: Hospital episode statistics; Infant; Outcomes; Pyloric stenosis; Pyloromyotomy; Volume-outcome.