Infantile hypertrophic pyloric stenosis: Has anything changed?


  • Nicole D Taylor,

    1. Douglas Cohen Department of Paediatric Surgery, The Children's Hospital at Westmead, Sydney Medical School, The University of Sydney, Sydney, New South Wales, Australia
    Search for more papers by this author
  • Daniel T Cass,

    1. Douglas Cohen Department of Paediatric Surgery, The Children's Hospital at Westmead, Sydney Medical School, The University of Sydney, Sydney, New South Wales, Australia
    Search for more papers by this author
  • Andrew JA Holland

    Corresponding author
    • Douglas Cohen Department of Paediatric Surgery, The Children's Hospital at Westmead, Sydney Medical School, The University of Sydney, Sydney, New South Wales, Australia
    Search for more papers by this author

  • Conflict of interest: None declared.

Correspondence: Professor Andrew JA Holland, Douglas Cohen Department of Paediatric Surgery, The Children's Hospital at Westmead, Locked Bag 4001, Westmead, NSW 2145, Australia. Fax: +61 2 9845 3346; email:



The study aims to review the contemporary presentation, diagnosis and treatment of children with infantile hypertrophic pyloric stenosis (IHPS) at an Australian paediatric tertiary centre.


A retrospective case review of patients with IHPS presenting to our institution between January 2004 and December 2010 was performed. Results were compared with a previous study evaluating two earlier 6-year series of patients treated between January 1984 and November 1995.


Of 362 infants presenting over the current 7-year review, 84.8% were male. Mean age of admission fell from 5.9 weeks in the previous study to 5.4 weeks. An ‘olive’ was palpated on examination in 48%, visible peristalsis seen in 25% and 16% of patients presented with haematemesis. Diagnosis solely on clinical examination fell from 74% to 9% and was associated with a corresponding increase in use of ultrasound from 16% to 91%. The frequency of serum chloride values less than 85 mmol/L declined from 26% to 9%. A variety of open and minimally invasive surgical approaches were used with similar outcomes, although laparoscopic pyloromyotomy was associated with significantly higher rates of wound infection (χ2 = 4.6, P = 0.03). The frequency of major complications remained low at 1%.


Contemporary patients with IHPS typically present earlier with a reduction in the incidence of metabolic derangement. Diagnosis based on clinical examination alone appears uncommon, with the majority of suspected cases confirmed by ultrasound. There was no clear difference in overall outcome based on the surgical approach used, although minor variations may reflect our institution's initial experience with laparoscopic pyloromyotomy.

What is already known on this topic

  1. Infantile hypertrophic pyloric stenosis (IHPS) is commoner in males and the first born child.
  2. Typically occurring in the first few months of life, infants present with non-bilious, forceful vomiting.
  3. Following correction of any electrolyte imbalance, pyloromyotomy remains the standard treatment.

What this paper adds

  1. Infants with infantile hypertrophic pyloric stenosis (IHPS) appear to present earlier than previously reported, with corresponding reductions in metabolic derangement.
  2. Ultrasound is presently utilised in achieving the correct diagnosis in over 90% of cases of IHPS..
  3. A variety of surgical approaches can be utilised to perform pyloromyotomy with generally equivalent outcomes.

The aetiology of infantile hypertrophic pyloric stenosis (IHPS), a condition characterised by abnormal thickening of the pylorus resulting in a gastric outlet obstruction, remains unknown. IHPS typically presents with progressive projectile vomiting that commences between the second and eighth week of life.[1] Traditionally, the diagnosis had been made clinically by feeling the thickened pylorus or ‘olive’ and observing gastric peristaltic waves.[2] Untreated, the infant becomes dehydrated, developing a hypokalaemic, hypochloraemic metabolic alkalosis. This must be corrected prior to surgical intervention.[2]

Several previous studies in the late 1990s and early 2000s had identified a trend towards earlier presentation, decreased metabolic derangement and greater use of diagnostic ultrasound (US).[1, 3-5] We sought to describe the more recent presentation, diagnosis, management and outcomes of children with IHPS at an Australian paediatric tertiary centre.


Patient population and setting

A retrospective case-note review was performed of all patients with IHPS that presented to The Children's Hospital at Westmead (CHW), a paediatric tertiary institution in Western Sydney, over a 7-year period between January 2004 and December 2010. Medical records for all 362 patients listed with the International Classification of Diseases Code (ICD-10-AM) of IHPS (Q40.0) were evaluated.

Patients from this current series were compared with those in a previous study from our institution that included 303 patients over a 12-year period, which had been divided into two 6-year cohorts.[1] The first included 151 patients treated between January 1984 and December 1989 (Series One) and the second 152 patients treated between January 1990 and November 1995 (Series Two).

Data for the current series were collected by a single author (NDT) using a standardised data collection sheet and included: gender, gestation, birth order, family history, age at time of first presentation to an Emergency Department (ED), time and date of admission, duration of symptoms prior to first ED presentation, presenting symptoms and signs, electrolyte profile on admission, method of diagnosis, operative approach, time to commencement of operative procedure, complications and time to hospital discharge. Wound infection was defined on the basis of a purulent wound discharge and a positive wound swab culture.

Ethics approval was granted by the CHW Research Ethics Committee.

Statistical analysis

Statistical analysis was performed using SPSS Statistics software (Version 19, International Business Machines Corp., Armonk, NY, USA). A P-value of <0.05 was considered significant.


Current series

Of 362 infants, 84.8% were male, resulting in a male : female ratio of 5.6:1. Birth order was recorded in 59% of patients. Of these, 43% were firstborn, 28% second, 18% third and 6% fourth. There was a documented positive family history in 17% (Table 1), with the majority (83.7%) of patients term (Table 2).

Table 1. Family history
Relationshipn% of current series
  1. Note: Eleven patients (3%) had a family history in two or more family members.
Uncle or aunt154
Table 2. Gestation
Post-term (>42 weeks)3141.1
Term (>37)2604330383.7
Preterm (<37 but >32)268349.4
Extreme prematurity (<32)4041.1
Not recorded143174.7

Presentation and duration of symptoms

Mean age at time of admission was 5.4 weeks (minimum 3 days and maximum 207 days, Fig. 1). There were three early presentations before 2 weeks of age in the current series: at days 3, 9 and 10 of life. There were four late presentations: two at week 14, and one each at weeks 22 and 29. In contrast, the mean age at time of admission in the first and second series combined was 5.9 weeks.

Figure 1.

Age at admission.

The median duration of symptoms prior to ED presentation in the current series was 5 days (range 1 to 105 days, Fig. 2). The most common presenting symptoms were non-bilious vomiting and failure to thrive or gain weight; with a palpable olive, dehydration and visible peristalsis the most common signs (Table 3). Of the 116 patients documented to be dehydrated, severity was documented as mild in 33, moderate in 8 and severe in 6.

Figure 2.

Duration of symptoms.

Table 3. Presentation
Symptom or signn%
Non-bilious vomiting36199.7
Palpable ‘olive’17248
Failure to thrive or gain weight16646
Visible peristalsis9225
Bilious vomiting164

Investigations and diagnosis

Approximately two thirds (67%) of patients had an elevated bicarbonate level: 29% were hypochloraemic, 20% were hyponatraemic and 10% were hypokalaemic. Figure 3 illustrates the proportion of patients with serum chloride value less than 85 mmol/L in all three series. In the present study, all patients were commenced on intravenous fluids on admission. Of these, 31% were documented to require correction of either dehydration or electrolyte abnormalities.

Figure 3.

Frequency of hypochloraemia.

Diagnosis was confirmed solely by clinical examination in 9% of patients (n = 34) in the current series. Of those diagnosed clinically, 1% (n = 6) required a test feed to make the clinical diagnosis. US was used to either establish or confirm the diagnosis in 91% (n = 331) and fluoroscopy in 9% (n = 31), with some patients having more than one investigation. Compared with Series One and Two, there was a dramatic rise in the use of US for diagnosis (Fig. 4).

Figure 4.

Frequency of use of ultrasound (US) to confirm diagnosis.

Operative treatment and complications

Pyloromyotomy was performed using a variety of surgical approaches in the current series (Table 4). One of the 74 laparoscopic pyloromyotomies required conversion to an open procedure due to a mucosal perforation, treated with an omental patch, but was analysed on an intention to treat basis. Overall, there were 12 mucosal perforations: 7 open and 5 laparoscopic, all identified and repaired at the time of the initial procedure.

Table 4. Operative approach
Operative approachn%
Right upper quadrant transverse incision17849.2
Peri-umbilical incision (Bianchi)6217.1
Midline incision4713
Missing data10.3

In 11 patients, additional procedures were performed under the same anaesthetic: 6 had excision of an umbilical granuloma, polyp or stump, 2 bilateral inguinal herniotomies, 2 tongue-tie releases, 1 left inguinal herniotomy and 1 pre-auricular skin tag removal.

Major complications were uncommon, occurring in approximately 1% of patients (Table 5). Minor complications occurred more frequently, with prolonged post-operative vomiting that delayed the introduction of feeds and wound infection the most common. Other complications included: wound granulomas (n = 4), haematemesis (n = 4), serosal tear (n = 4), bradycardia and/or post-operative apnoea (n = 4), and one each of seizure and respiratory arrest, thiopentone overdose, difficulty weaning from ventilation, need for reoperation with later diagnosis of jejunal pseudo-obstruction, atelectasis, post-extubation stridor and delayed gastric emptying. There were no incomplete pyloromyotomies. In comparison, in the combined first and second series, wound infection had occurred in 19%, mucosal perforation in 4.6%, wound dehiscence in 2% and incisional hernia in 0.6%.

Table 5. Operative complications
  1. aSee text.
Incisional hernia (requiring repair)20.6
Wound dehiscence20.6
Total 1.2
Post-operative vomiting delaying feeding >24 h246.6
Wound infection236.4
Mucosal perforation (repaired intra-operatively)123.3
Incisional hernia (not requiring repair)92.5
Total 23.8

In the current series, the complication rates for laparoscopic and open pyloromyotomy were compared and found to be broadly comparable. The only statistically significant difference was a higher rate of wound infection following a laparoscopic approach (9 of 74 laparoscopic cases (12%) compared with 14 of 287 open cases (5%), χ2 = 4.6, P= 0.032). Laparoscopic rates of mucosal perforation and post-operative vomiting delaying feeding were also slightly higher than in the open approach, but these differences were not statistically significant (χ2 = 2.8, P = 0.097 and χ2 = 2.7, P = 0.100, respectively).

Time to surgery and length of stay

The mean time from ED presentation to commencement of surgery was 1.9 days, despite all significant electrolyte imbalances and dehydration having been corrected within the first 12 to 24 h of admission. The mean post-operative stay was 2.4 days, with the mean length of stay (LOS) of 4.3 days. Despite presenting earlier, only 21% of patients had surgery performed within 24 h of admission compared with 58% in the first and second series, with the overall LOS therefore also shorter in the first series and second series (3.7 days and 3.2 days, respectively).


This study highlights several important changes in the presentation and management of IHPS at our institution over the past three decades.[1, 3, 4] Infants currently present earlier, with corresponding reductions in the frequency of metabolic disturbances and moderate to severe dehydration.[3-5] Mean age at admission has fallen, consistent with other recent published data.[4]

Perhaps in part because of this earlier presentation, the majority of patients with IHPS now have the diagnosis made by US, rather than simple clinical examination augmented by a test feed.[2-4] While more vigorous, less dehydrated infants may be intrinsically more difficult to examine, it seems likely a combination of factors, including parental pressure, medico-legal concerns and the widespread availability of US have all contributed to this change.[4, 6] Rather than decrying the apparent loss of clinical skills associated with diagnosis of IHPS, the critical lens might be better directed on determining how care might be enhanced as a result of earlier diagnosis.[7] For example, outpatient investigation via US may prevent unnecessary hospital presentations and even admissions: as such, the additional costs of a US may in fact represent a cost saving.

In the current study, despite earlier presentation with minimal or no electrolyte imbalance, patients neither progressed to surgery nor were discharged any earlier compared with previous series.[1] This may reflect underlying caution in the medical management of infants, or more likely difficulties in access to operative time at a busy paediatric tertiary institution with an increasing non-tertiary surgical workload. It seems likely that the potential cost savings as a result of earlier diagnosis and avoidance of hospital admission of infants without IHPS would be offset by the financial consequences of their longer overall admission.

While the great majority of infants presented with non-bilious vomiting, 16 (4%) presented with bilious vomiting. In the majority of these infants, the bilious vomiting ceased promptly with pyloromyotomy. In the one case in which this persisted, a subsequent upper gastrointestinal contrast study was normal and the bilious vomiting ceased 4 days post-operatively. It is possible that, in one or two cases, a combination of haematemesis and old blood in the vomitus, combined with curdling of milk retained in the stomach for a period of time, may have been mistaken for bile. In the remaining cases, it would seem that the severity of the vomiting, as occasionally occurs in severe gastroenteritis, was sufficient to result in bile staining of the vomitus despite the narrowing of the pyloric canal. This has been reported previously, although less frequently than in our series, in 1.4% (5 of 354 infants) of cases and serves to emphasise that bilious vomiting does not absolutely exclude the diagnosis of IHPS.[8]

Our surgical outcomes revealed decreased rates of wound infection, mucosal perforation and wound dehiscence compared with earlier series, and on the whole were well within published rates.[1, 3-6] The dramatic reduction in wound infection rates in particular may have occurred as a result of earlier presentation and diagnosis, with patients less likely to be malnourished prior to surgery. While the overall complication rate appeared high, this reflected comprehensive data collection and the inclusion of more minor adverse events not clearly documented in other series.[1, 3-6] The complications from laparoscopic pyloromyotomy seen in our institution, together with no clear evidence of benefit in terms of earlier discharge, may represent the reported ‘learning curve’ associated with this technique, with somewhere between 20 to 35 cases required to achieve an expert level.[9, 10]

Several well-conducted trials have demonstrated that once sufficient experience has been obtained, at least equivalent results to open pyloromyotomy may be achieved with potentially earlier introduction of enteral feeds and discharge.[11, 12] At the same time, recent concerns over the potential adverse neurodevelopmental outcomes seen in infants following surgical procedures requiring general anaesthesia, including pyloromyotomy, need to be carefully considered given the potential for performing open pyloromyotomy under regional anaesthesia.[13-16]


IHPS remains a common paediatric surgical condition. While infants now typically present earlier, with less dehydration and metabolic abnormalities, this has not resulted in shorter times to operative intervention or reduced LOS. The majority of cases will now be diagnosed with US. Laparoscopic pyloromyotomy was associated with a higher rate of wound infection and no statistically significant evidence of earlier return to normal feeding or discharge. This may reflect our institution's early involvement with this technique and might be expected to improve with increasing experience.