Statistics from Altmetric.com
If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.
Dr D G Bundy, Johns Hopkins University School of Medicine, Baltimore, MD, USA;
how accurate are symptoms, signs, and laboratory tests for diagnosing appendicitis in children with abdominal pain?
Medline (to March 2007), Cochrane Library, textbooks, and reference lists.
Study selection and assessment:
fully published studies in English that evaluated the accuracy of medical history, physical examination, or basic laboratory tests for diagnosing appendicitis in children, using independent, blinded comparison with a gold standard—surgical pathological findings, clinical observation, or follow-up (level of evidence 1–3). 25 studies were included: 1 level 1 study (n = 246) of unselected children presenting to the emergency department with undifferentiated abdominal pain (appendicitis prevalence 10%) and 24 level 3 studies (n = 5590) of selected patients with suspected appendicitis (appendicitis prevalence 25–89%).
sensitivity, specificity, and positive and negative likelihood ratios (LRs).
The table shows the results.
The most useful single feature for diagnosing appendicitis in children with abdominal pain was fever in unselected children and rebound tenderness in selected children.
Bundy DG, Byerley JS, Liles EA, et al. Does this child have appendicitis? JAMA 2007;298:438–51.
Clinical impact ratings: GP/FP/Primary care 5/7; Paediatrics 6/7; Surgery—general 6/7; Paediatric emergency medicine 5/7; Emergency medicine 5/7
Acute appendicitis is the most common serious cause of abdominal pain among children presenting to emergency departments, and atypical presentations make diagnosis challenging, particularly in the very young. The review by Bundy et al illuminates the usefulness of symptoms, signs, and laboratory tests in assisting practitioners facing this dilemma.
Clinical assessment first identifies children whose presentation is dominated by abdominal pain and proceeds to define a subgroup for whom appendicitis is the lead surgical diagnosis. Finally, diagnostics, such as computed tomography or ultrasonography, extended observation, and serial physical examinations, identify children who are candidates for admission to hospital or surgery. The prevalence of appendicitis across these successive subgroups is approximately 10%, 30%, and ⩾85%, respectively. Clinicians are most in need of diagnostic aids when diagnostic uncertainty is highest, before the decision to scan, admit, or operate. Information about the accuracy of diagnostic assessments drawn from populations in the later stages of the process, as in most of the studies found by Bundy et al, may be misleading.1
Only 1 emergency department–based study identified by Bundy et al involved children with undifferentiated abdominal pain. This study showed that the presence or absence of fever and vomiting are relatively reliable discriminators. However, focal abdominal tenderness was a poor indicator in this and the other studies. Likewise, laboratory tests seem to be relatively useless across most of their ranges. Studies done in patients with a high probability of appendicitis suggest that the Alvarado score may help to objectify the diagnostic process. A score ⩾7 multiplies the odds of appendicitis by 4, equivalent to the increase of likelihood from 10% to 30% required to identify a child with abdominal pain for whom definitive assessment or action is appropriate.
Practitioners should use caution in applying the results of this review to younger children or to patients in primary care settings where the likelihood of appendicitis is much lower.
Funding: no external funding.