Learning Objectives: After reading this article, the physician should be better able to:
1. Create a strategy to help distinguish signs and symptoms of appendicitis in children.
2. Interpret likelihood ratios and apply them to clinical practice.
3. Devise a treatment and disposition strategy for children with abdominal pain in the emergency department.
Article from the 2010 LLSA Reading List
Does This Child Have Appendicitis?
Bundy DG, et al
This article, part of the Rational Clinical Examination Series in the Journal of the American Medical Association, was written by experts from the pediatric and surgery departments of major academic hospitals. The authors share clinical tips for eliciting a relevant history, symptoms, and signs of appendicitis in children of different ages, and examine the literature using a meta-analytic approach to assess the value of these various clinical features in diagnosing pediatric appendicitis.
Abdominal pain is one of the most common presenting complaints in pediatrics, according to the article. Despite being the most frequent surgical diagnosis among children in the ED with abdominal pain, appendicitis remains a clinical diagnostic challenge. Up to 57 percent of children under 12 with appendicitis are initially misdiagnosed. (J Pediatr Surg 1993; 28:204.) Perforated appendicitis rates are as high as 20 percent in children between 10 and 17, and perforation rates approach 100 percent in children under 4. (Am J Epidemiol 1990; 132:910.)
The authors emphasize that caregivers are an important source of clinical information for preverbal children. Be cautious about accepting information from young verbal children because they tend to answer “yes” to every question. The authors suggest keeping younger patients on a parent's lap, and developing a good rapport with the child before beginning the formal physical examination. Clinicians should focus on children's facial expressions while reassuring them during abdominal palpation. For uncooperative or anxious patients, the authors suggest engaging patients in conversation and using a stethoscope to palpate. Keep in mind that voluntary guarding is often due to fear rather than actual pain, and false-positive findings can be mitigated by distraction techniques.
The authors also focus on a meta-analysis of the current literature on pediatric appendicitis examining the sensitivity, specificity, and likelihood ratios of its different presenting signs, symptoms, and laboratory values. After preliminary review of more than 2500 abstracts, 25 articles were selected for inclusion.
The authors emphasized that the presence of fever is the single most useful sign associated with appendicitis in children with abdominal pain. According to the article, assuming a pretest probability of 10% for appendicitis, fever (LR=3.4) increases the likelihood of appendicitis by almost threefold, while its absence (LR=0.32) decreases the likelihood by two-thirds. Interestingly, the presence of localized abdominal tenderness was not found to be helpful in predicting appendicitis (LR=1.2), but its absence appears useful in excluding appendicitis (LR= 0.45). Of all physical exam findings, the authors claim the presence of rebound tenderness (LR=3.0) nearly triples the likelihood of appendicitis.
A WBC less than 10,000/µl decreases the likelihood of appendicitis (LR=0.22), according to the article, but it is difficult to draw meaningful conclusions when the WBC is more than 15000/µl because the likelihood ratio ranges from 1.0 to 3.7. The authors also conclude that an Alvarado score greater than 7 increases the likelihood of appendicitis (LR=4.0), while a score of less than 7 is even better at reducing the likelihood of appendicitis (LR=0.20).
The article concludes by re-emphasizing the paramount importance of the clinical examination. They recommend a low threshold for diagnostic imaging and surgical consultation, particularly for preverbal and young children, in whom the diagnosis of appendicitis can be difficult. For those children at low risk for appendicitis, they recommend careful follow-up instructions when sending a child home.
After reading this article, clinicians should be able to create a strategy to help differentiate signs and symptoms of appendicitis in children, and devise a treatment and disposition strategy for children with abdominal pain in the emergency department.
Comments: This meta-analysis summarizes likelihood ratios of various signs and symptoms in an attempt to estimate the chance that a child with abdominal pain has appendicitis. A test result with a positive one (LR>1.0) increases this chance while a negative one (LR<1.0) decreases it. If a test has a likelihood ratio of 1.0, the test result has no impact on changing the chance of having a disease. But caution must be used before indiscriminately applying likelihood ratios to clinical practice without understanding their limitations. Take fever, for example. The article repeatedly emphasized that fever is the most powerful predictor of appendicitis in children with abdominal pain because it triples the likelihood of appendicitis when present (LR=3.4) and decreases the likelihood of appendicitis by two-thirds when absent. But is this really true?
It depends. The answer is yes only if your pretest probability of appendicitis in your patient is 10%, as the authors assumed when performing this calculation. The reason is different pretest probabilities will give rise to different post-test probabilities even if the likelihood ratio remains the same. The pretest probability of disease is actually one of the most important factors to consider when evaluating the need for further testing, an important concept not emphasized in this article. The 25 studies used in this meta-analysis evaluated different patient populations with self-reported prevalence (pretest probability) of appendicitis ranging from 10 percent to 92 percent.
Perhaps the best way to demonstrate the variable value of a test based on pretest probability of disease is by showing a few examples. Consider a child with a 90% pretest probability of having appendicitis. Without detailing calculations, if temperature is taken into account, fever (LR=3.4) increases the post-test probability to 97%, representing only a seven percent increase, which is far from a threefold jump. In the same child, if fever is absent (LR=0.32), post-test probability decreases to 74%, a 16 percent drop, but far from decreasing chances by two-thirds. In this clinical case, because the pretest probability of appendicitis is so high, temperature is much less likely to affect clinical decision-making and management.
Consider a teenage girl presenting with urinary symptoms, left flank pain, and abdominal pain. You conclude this child is likely to have pyelonephritis, and has a very low chance of having appendicitis (say 2%). Now considering temperature, if fever is present (LR=3.4), post-test probability increases to 6%, three times more likely, but only an absolute change of four percent. If fever is absent (LR=0.32), the post-test probability drops to about 1%. Yet despite a threefold increase or a 50 percent decrease in the risk of appendicitis, temperature again in this case is likely to have minimal impact on your clinical decision-making. As a rule, the most meaningful signs, symptoms, and clinical findings have very positive (LR>10) or very negative (LR<0.1) likelihood ratios. (JAMA 1994;271:389.)
Another important point to consider is that the article does not take into account that the prevalence of appendicitis varies tremendously based on age. For emergency physicians, age-stratified pre- and post-test probability are extremely important because more deadly diseases should be considered in children with abdominal pain depending on their age. In children between 3 months and 3 years, intussusception must also be excluded. In children under a year, especially with bilious vomiting, malrotation should be considered before appendicitis. In older children, missed testicular torsion may be more detrimental than missed appendicitis because ischemia may be irreversible by the time the child returns. Ectopic pregnancy in teenage girls with abdominal pain must be considered, and a urine pregnancy test should be checked even if sexual activity is denied.
Abdominal pain in children can be diagnostically challenging. The younger a child is, the more difficult this process can become. If a patient presents classically and more deadly etiologies have been excluded, consider surgical consultation even without imaging. If the patient is playful and looks well (except in the case of intussusception when the telescoping is gone) and the history and examination are not consistent with appendicitis, involve caretakers in the decision-making process, and make sure they are reliable and understand the return precautions you give them.
CME Participation Instructions
To earn CME credit, you must read the article in Emergency Medicine News, and complete the evaluation questions and quiz, answering at least 80 percent of the questions correctly. Mail the completed quiz with your check for $12 payable to Lippincott Continuing Medical Education Institute, Inc., Two Commerce Square, 2001 Market St., Third Fl., Philadelphia, PA 19103. Only the first entry will be considered for credit, and must be received by Lippincott Continuing Medical Education Institute by June 30, 2011. Acknowledgment will be sent to you within six to eight weeks of participation.
Lippincott Continuing Medical Education Institute is accredited by the Accreditation Council for Continuing Medical Education to provide medical education to physicians. Lippincott Continuing Medical Education Institute designates this educational activity for a maximum of 1 AMA PRA Category 1 Credit.™ Physicians should only claim credit commensurate with the extent of their participation in the activities.
Comments about this article? Write to EMN at firstname.lastname@example.org.
Click and Connect! Access the links in this article by reading it on www.EM-News.com.
© 2010 Lippincott Williams & Wilkins, Inc.