Discussion

Ideally, a scoring system would work as a tool alongside clinical acumen to increase the accuracy of decision-making, while reducing the need to expose patients to harmful imaging and/or increased time before surgical intervention to prevent appendicular perforation. Previous systematic reviews have documented poor specificity of the Alvarado in pediatric populations.10–12 Similarly, in comparative studies, the AIR method was superior in terms of specificity, but not necessarily sensitivity in the adult population.13 This is the first study to compare the AIR and Alvarado scoring systems exclusively within a pediatric population who already had a clinical diagnosis of appendicitis and a decision to operate. The aim was to find out whether scoring systems alone can reduce the negative appendicectomy rate. In contrast with other studies,14–16 we did not find AIR to be superior to Alvarado in specificity in our population.

The low-risk classification classes of both AIR and Alvarado systems were similarly uninformative about appendicitis rates. However, both systems performed better at medium and high-risk levels. Adopting either or both scoring systems as a supplemental decision-making tool must take this into account with risk assessment. If in doubt, then a score of medium or high risk in either or both of the scoring systems should favor early surgical intervention. There is no indication that there is clinically important distinction between medium and low-risk classes. In AIR, only five cases were assigned to the high-risk category. In the Alvarado system, although negative appendicectomy rates were lower in high versus medium-risk classes, the advantage in decrease in negative rate may be outweighed by increase in danger due to perforation. A low-risk classification within either scheme will result in >50% negative appendicectomy rates.

Supplementary imaging is a common diagnostic tool guiding decision to operate. However, imaging as a matter of course has two main disadvantages. Ultrasound is variously advocated. Pershad et al found ultrasound was the most cost-effective diagnostic approach in children with suspected appendicitis,17 while others have reported ultrasound imaging as inappropriate given its delay on treatment.18 In this study we found that time to theater was considerably extended with ultrasound (2.03 vs 0.75 days). In reality, however, management strategies are rarely based on negative ultrasound, offsetting some of its benefit except as a confirmatory tool. However, the sonographer may also find other signs such as free fluid and thickened bowel loops, but their sensitivity and specificity in diagnosing appendicitis is uncertain.19 CT imaging has been shown to have significantly higher sensitivity than ultrasound in the pediatric population,20 but increases risk of radiation exposure. Our negative appendicectomy rates were higher than reported in the literature. This may be due to non-reliance and reluctance on imaging such as CT within our hospital. Adoption of a scoring system could rationalize the need for selective radiological assessment in cases where the insight into equivocal cases in the low-risk category outweighs the risk of the procedure and unnecessary surgery. Radiological scans in medium and high-risk categories would be largely redundant.

The AIR score was found to more confidently identify those patients with a high probability of appendicitis in whom supplemental imaging is unlikely to change management and thus an early decision to operate should be made. This is of benefit as imaging is also shown to increase time to theater.

Our study has some limitations and caveats. Ours was a retrospective study of all children who were operated for suspected acute appendicitis. The preselected population created an inherent systematic bias. Given some degree of clinical acumen had already been used to rule out an unknown percentage of cases, this study does not attempt to generalize the performance data across a pediatric population presenting with non-specific abdominal pain, for which appendicitis may be a differential diagnosis. Our study population is clearly of those within the age range for whom appendicitis has been diagnosed.

Due to the nature of retrospective data collection, there were instances where scores had to be inferred, based on the accuracy of the documentation, or classified as ‘not present/negative’ if there was no documentation. Rebound tenderness when documented was graded as 1 point but no grade given when not documented. ‘Moderate’ rebound tenderness was given 2 points whereas descriptions such as ‘severe’ rebound tenderness were given 3 points. One of the benefits of the AIR score over the Alvarado score became apparent. The Alvarado score requires the child to be able to recognize and describe feelings such as nausea and verbalize migratory pain and recognize anorexia. The AIR relies more on laboratory findings and objective clinical findings. Symptomology and signs of appendicitis are relative to the progression of the inflammatory response and time of presentation. In our data collection, we decided to look at results on admission, therefore, a selection of patients may have benefited from a scoring tool used on admission and repeatedly during observation prior to discharge.

Given the clinical overlap between acute appendicitis and non-surgical causes of abdominal pain, a scoring system that stratifies risk of acute appendicitis certainly does have a place. In high scoring children, the decision should be made to operate with no need for imaging to confirm or disprove diagnosis. There is perhaps a requirement for a modified AIR score for use in the pediatric population which could draw on observations of serum CRP and percentage polymorphonuclear leucocytes being directly related to disease stratification while remaining clear of criteria difficult for children to appreciate.