Traditionally, the diagnosis of infantile hypertrophic pyloric stenosis (IHPS) is based on a history of projectile, nonbilious vomiting, and palpation of a pyloric tumor.^1,2 The presence of hypochloremic alkalosis has also been associated with IHPS.³ Infants with historical but not physical findings consistent with IHPS undergo radiologic evaluation by either upper gastrointestinal series (UGI) or, more recently, abdominal ultrasound (U/S) to establish the diagnosis. The increased reliance of radiologic tests to diagnose IHPS has been ascribed to inexperienced examiners not palpating a pyloric tumor and subsequently proceeding with further evaluation.⁴ To evaluate temporal trends in the presentation and methods of diagnosis of IHPS, we performed a review of patients who underwent pyloromyotomy (PM) in the past 25 years.

PATIENTS AND METHODS

RESULTS

Analysis of the five time periods demonstrated that patients presented at a younger age, weighed more, and had a shorter length of illness in the most recent interval (Table 1). The prevalence of prematurity and positive family history of pyloric stenosis did not vary over time. The frequency of palpation of a pyloric tumor by the pediatric surgical staff decreased significantly over time (P < .001).

Table 1. Demographic and Physical Characteristics [View Table]

Electrolytes including sodium, potassium, and chloride were assessed at the time of admission in 98% of patients. Blood gas analysis was obtained in 10% of patients. Serum bicarbonate was added to the standard electrolyte panel in 1981 at one hospital and 1983 at the second hospital; therefore, analysis including bicarbonate is limited to the later half of the study. Metabolic derangements at admission were found less frequently over time (Table 2). Hypochloremia (defined as a serum chloride of less than 90 mEq/L) or alkalosis (defined as a pH of greater than 7.46) decreased in frequency in the later years (Fig 1).

Table 2. Admission Laboratory Data [View Table]

An imaging study, either an UGI or abdominal U/S was performed in 90% of the entire study population. Eighty percent of these tests were ordered by the referring physician, not by the pediatric surgeon. UGI was used exclusively until 1983 when U/S was introduced to both hospitals as an alternative diagnostic test in patients with suspected IHPS. One imaging study was performed in 61% of patients in the earliest time period compared with 96% in the most recent time period (Fig 2). Diagnostic imaging was ordered by the referring physician increasingly over time, ranging from 74% in the earliest time period to 86% in the most recent time period.

Preoperative length of stay (PLOS) varied throughout the study (Fig 3). The PLOS was initially greater than 1 day; however, it declined to .68 days during the middle portion of the study period. An increase in the PLOS to greater than 1 day was found again in the most recent time frame. The total length of stay (TLOS) initially declined then remained stable.

DISCUSSION

This analysis demonstrates that the aforementioned classical features of IHPS have decreased in frequency over time. The length of illness before admission declined during this study. Weight loss, dehydration, and metabolic abnormalities became less common. Physical examination yielded positive findings in only 23% of infants in the last 5 years of inquiry compared with 79% in the first 5 years. Presumably, palpating a pyloric tumor in healthier, less emaciated, vigorous infants was more difficult, decreasing the sensitivity of the physical examination. Macdessi⁵ also found a significant decline in palpable pyloric tumors over time, but an association with duration of illness was not established.

Imaging studies were routinely used to diagnose IHPS in the last 10 years of this study. Studies published before 1975 reported between 8% and 38% of infants with IHPS undergoing radiologic evaluation before surgical intervention.^1,2,7,16 However, diagnostic imaging was utilized increasingly in more recent series.^5,6

The referring physician ordered the majority of diagnostic tests performed before referral. Others have demonstrated a similar pattern.^6,9,14 Presumably, a pyloric tumor was not palpated by the referring physician and this prompted further evaluation of the patient's vomiting. Increased usage of imaging modalities to diagnose IHPS has raised the concern that physical examinations by inexperienced physicians has lead to the performance of unnecessary tests.^4,19,20 During the earlier periods of this study, 72% of UGIs ordered by referring physicians subsequently had a pyloric tumor palpated by the pediatric surgeon. However, in the last 5 years of study, only 21% of imaging studies ordered by referring physicians were performed in infants with a palpable pyloric tumor. As palpable pyloric tumors became less frequent, an imaging study was necessary for diagnosis.

PLOS is an indicator of the patient's clinical status. It can also be a measure of the uncertainty of the diagnosis. In the earlier years, when patients presented more frequently with metabolic derangements, rehydration and electrolyte replacement were more often necessary before surgery. PLOS decreased during the middle period of the review. However, over the last 10 years, PLOS increased again correlating with the more routine use of imaging studies and the time required to perform one or more imaging studies for diagnosis. An increase in PLOS and diagnostic studies being performed have coincided with these changes in the presentation and diagnosis of infants with IHPS. Nevertheless, the TLOS decreased over the study period.

One question this study raises is why the change in presentation of IHPS over time? This study demonstrates that over time, infants were not as ill when they presented. The most probable reason for this is the earlier usage of imaging studies in an outpatient setting to ensure the diagnosis of IHPS. Infants were diagnosed with IHPS before they had significant metabolic abnormalities, weight loss, or the pathonomonic finding of a pyloric tumor.

Is the early use of an imaging study necessary? Is it more cost-effective to obtain an UGI or U/S in an infant who has been vomiting for a few days to rule out IHPS or risk the infant becoming dehydrated from persistent vomiting and require hospitalization for electrolyte abnormalities? If we use PLOS as an indicator of the patient's clinical status to determine the cost of hospitalization for dehydration, an estimated .52 days are needed to correct metabolic abnormalities. (This calculation reflects the difference between the PLOS during the earliest time period of this study [1.2 days] when more patients had metabolic derangements, and the middle portion of the study [.68 days] when fewer metabolic derangements were encountered). The current charge of an UGI or U/S at our hospitals is less than a half-day of hospitalization.

We recommend the following algorithm to diagnose IHPS: when an infant presents with vomiting, attempts to palpate a pyloric tumor should be made. If the primary physician is unable to palpate a pyloric tumor but has a high index of suspicion, a pediatric surgeon should examine the patient. If a pyloric tumor is palpated, the patient should undergo PM. If no tumor is palpated or a pediatric surgeon is not available, an imaging study should be performed.