Nature and Burden of the Condition
Abdominal pain is a common presenting complaint for patients seeking care at emergency departments, with approximately 3.4 million expected cases per year in the United States. 1 Appendicitis is a frequent cause of abdominal pain, caused by acute inflammation of the appendix, and occurs in approximately 8 to 10 percent of the population (over a lifetime). 2 , 3 Appendicitis has its highest incidence between the ages of 10 and 30 years. The ratio of incidence in men and women is 3:2 through the mid-20s and then equalizes after age 30. Appendicitis is the most common abdominal surgical emergency, with over 250,000 appendectomies performed annually in the U.S. The risk of acute appendicitis in pregnant women is similar to that of the general population, making appendicitis the most common non-obstetric emergency during pregnancy. 4 – 7 Untreated, appendicitis can lead to perforation of the appendix, which typically occurs within 24 to 36 hours of the onset of symptoms. Perforation of the appendix can cause intra-abdominal infection, sepsis, intraperitoneal abscesses, and rarely death. 4 In order to avoid the sequelae of perforated appendicitis, a low percentage of “negative” appendectomies (i.e., removing a normal, non-inflammed appendix in patients mistakenly diagnosed with appendicitis) is generally accepted from a surgical standpoint.
Diagnosis of Suspected Acute Appendicitis
Guidelines suggest that when a diagnosis of acute appendicitis can be made on clinical grounds surgical consultation should be sought without delay for additional diagnostic testing. 8 Clinical symptoms and signs suggestive of appendicitis include a history of central abdominal pain migrating to the right lower quadrant (RLQ), anorexia, fever, and nausea/vomiting. On examination, RLQ tenderness, along with “classical” signs of peritoneal irritation (e.g., rebound tenderness, guarding, rigidity, referred pain), may be present. Other signs (e.g., the psoas or obturator signs) may help the clinician localize the inflamed appendix. 9 – 11 The performance of clinical symptoms and signs for identifying acute appendicitis seems to be variable across studies, many patients present atypically, and few clinical findings appear to have adequate sensitivity and specificity when used in isolation. 10 , 11
For patients with right lower quadrant (RLQ) pain, when the diagnosis cannot be made on clinical grounds alone, laboratory or imaging tests are often used to attempt to establish a diagnosis and guide treatment. Laboratory evaluations potentially useful for the diagnosis of appendicitis include white blood cell count, granulocyte count, the proportion of polymorphonuclear blood cells, and C-reactive protein concentration. 10 – 12 Imaging tests, such as ultrasound (US), computed tomography (CT) with and without contrast, and magnetic resonance imaging (MRI), are also used extensively for the diagnosis of appendicitis. 13 – 19 Imaging tests can be used alone or in combination. For example, US is sometimes used as a triage test to separate patients in whom sonography alone is adequate to establish a diagnosis from those who require further imaging with CT. 8 Different factors may affect the performance of alternative tests and their impact on clinical outcomes. For example, US examination is considered to be operator dependent 20 and is technically challenging in obese patients or women in late pregnancy. CT scanning can be performed with or without the use of contrast agents, and contrast can be administered orally, rectally, intravenously, or via combinations of these routes. 8 It has been suggested that low body mass index (BMI), a marker for lack of sufficient mesenteric fat (which helps visualize periappendiceal fat stranding, a radiological sign of appendicitis), may affect the relative test performance of CT performed with or without contrast (contrast being more useful in individuals with low BMI and children). 8
Clinical symptoms and signs, along with the results of laboratory or imaging tests, can be combined into multivariable diagnostic scores (sometimes referred to as “clinical prediction rules”), multivariable that synthesize the findings of different investigations to determine the most likely diagnosis. 21 In adults, the most commonly used multivariable score for appendicitis is the Alvarado score, 22 which separates patients into 3 groups of increasing probability of appendicitis (the score is based on 8 items: pain migration, anorexia, nausea, tenderness in RLQ, rebound pain, elevated temperature, leukocytosis, and shift of white blood cell count to the left). 23 Although the Alvarado score is also used in pediatric populations, 24 , 25 the Pediatric Appendicitis Score has been developed and validated for use in children. 26 It is based on 9 items (migration of pain, anorexia, nausea/vomiting, fever, cough/percussion tenderness, hopping tenderness, RLQ tenderness, leukocytosis, polymorphonuclear neutrophilia) and classifies children into two groups (high vs. low probability of appendicitis). 26
Diagnostic laparoscopy is also used for the evaluation of patients with RLQ pain and suspected acute appendicitis, primarily when a diagnosis cannot be established via other means. Although diagnostic laparoscopy is generally considered safe, studies have reported variable rates of morbidity and mortality from the procedure. 27
In general the diagnostic tests discussed in this section are widely available in the U.S. Clinical symptoms and signs can be evaluated relatively easily and inexpensively. Evidence from the National Hospital Ambulatory Medical Care Survey suggested that CT and complete blood counts are obtained in the majority of patients presenting to the emergency department with abdominal pain. The survey also showed that over time (between 1992 and 2006) the use of CT for both adults and children has been increasing. Over the same period, the use of the complete blood count has increased in adults but decreased in children. 28 , 29 Various other sources suggest that the use of US and MRI is increasing in populations where exposure to ionizing radiation is a particular concern (e.g., children and pregnant women). 30 – 36
Importance of Accurate Diagnosis and Impact on Outcomes
As with all diagnostic tests, the modalities used in the diagnostic investigation of patients with RLQ pain/suspected appendicitis affect clinical outcomes indirectly, through their impact on clinicians’ diagnostic thinking and decisionmaking. 37 More accurate and timely diagnosis of appendicitis can minimize the time to the indicated intervention (e.g., surgery), thus reducing the time patients are in pain and improving clinical outcomes (e.g., reducing the rate of perforated appendicitis and its attendant complications). 38 Conversely, time-consuming or unnecessary diagnostic workup (an important, but hard to operationalize outcome) may delay the indicated treatment and increase the risk of complications or result in false positive results and more “negative” appendectomies. Furthermore, diagnostic testing can impact resource utilization for the management of patients with acute abdominal pain. For example, examination with CT may reduce length of stay by avoiding prolonged observation in cases where a diagnosis cannot be established clinically or by eliminating the need for additional diagnostic testing. 18 In some cases, CT can also facilitate direct therapeutic intervention. For example, in patients with perforated appendicitis complicated by an abscess, the radiologist can not only detect but also treat the abscess by percutaneous drainage, thus avoiding the need for immediate operative intervention.
Special Considerations for the Diagnosis of RLQ Pain/Acute Appendicitis
The diagnostic workup of acute appendicitis is complex because patients with acute abdominal pain of different etiologies can present with similar symptoms. Diagnosis is particularly challenging in children, women of reproductive age, pregnant women, and frail or elderly patients. 8 , 39 , 40
Acute appendicitis in children is often diagnosed after perforation has occurred. 41 – 43 Children have a thinner appendiceal wall and less developed omentum (the largest peritoneal fold), and thus may not readily wall off a perforation. In addition, many common childhood illnesses have symptoms similar to those of early acute appendicitis (e.g., fever, nausea, and vomiting), making the differential diagnosis more challenging. Young children may have difficulty communicating about their discomfort or describing their symptoms, making the clinical examination less informative and leading to diagnostic delays. 11 In addition, the use of modalities that involve ionizing radiation (e.g., CT) possibly entails greater radiation-related risks for children. 8
Women of Reproductive Age
A large proportion of women of reproductive age with appendicitis are misdiagnosed. 44 Establishing a diagnosis in women of reproductive age with RLQ pain/suspected acute appendicitis can be particularly challenging because symptoms of acute appendicitis can mimic those of gynecologic disease (e.g., pelvic inflammatory disease, ectopic pregnancy, etc.).
Diagnosis of suspected acute appendicitis in pregnant women can also be challenging because some symptoms of appendicitis (nausea and vomiting) are common in normal pregnancies and because enlargement of the uterus can alter the location of the appendix, which often moves higher and to the back. 45 Anatomic changes induced by pregnancy make the clinical examination of pregnant patients with abdominal pain more challenging and result in technical difficulties when using US. 36 , 46 , 47 Tests involving ionizing radiation (e.g., CT) are also generally avoided during pregnancy to prevent exposure of the fetus to radiation. Finally, obtaining a white blood cell count is generally not helpful in the diagnosis of acute appendicitis in pregnant women because leukocytosis is common during pregnancy. From a decisionmaking perspective, the management of suspected appendicitis in pregnant women is complicated by the need to balance the potential benefits and harms of testing for both the mother and the fetus.
Frail and Elderly Individuals
The elderly typically present with appendicitis in more advanced stage, when compared to younger patients. 48 Older patients may delay seeking care, and definitive diagnosis is sometimes delayed further because competing etiologies for abdominal pain (e.g., malignancy or diverticulitis) are considered more likely. Therefore, the performance of diagnostic tests may be modified by patient age (e.g., US has been reported to have higher diagnostic performance in older patients) and by the more advanced disease stage that is common in this age group. Elderly and frail individuals with appendicitis have a higher complication rate and a higher risk of mortality, compared to younger/less-frail patients.
Rationale for Evidence Review
Accurate testing of patients with RLQ pain, or less typical presentations consistent with acute appendicitis, to identify those who need treatment can improve clinical outcomes and reduce resource utilization. Our review of guidelines and published systematic reviews indicated a lack of specific guidance for selecting diagnostic modalities, particularly in patient subgroups in whom the diagnosis is known to be particularly challenging (e.g., children, women of reproductive age, and pregnant women). Existing systematic reviews have not adequately investigated the comparative effectiveness of alternative diagnostic approaches (typically they assess a single diagnostic modality), have focused almost exclusively on test performance outcomes (without providing evidence on the impact of tests on intermediate or patient-relevant outcomes), and have not addressed factors that may modify the test performance (such as patients’ age and sex, setting of care, or aspects of the test itself, e.g., the use of oral contrast, its administration via different routes, etc.). No review has comprehensively examined all tests of interest or focused on comparisons between alternative strategies.
This review addresses the following Key Questions:
Key Question 1.
What is the performance of alternative diagnostic tests, alone or in combination, for patients with right lower quadrant (RLQ) pain and suspected acute appendicitis?
What is the performance and comparative performance of alternative diagnostic tests in the following patient populations: children, adults, nonpregnant women of reproductive age, the elderly (age ?65 years)?
What factors modify the test performance and comparative test performance of available diagnostic tests in these populations?
Key Question 2.
What is the comparative effectiveness of alternative diagnostic tests, alone or in combination, for patients with RLQ pain and suspected acute appendicitis?
For the populations listed under Key Question 1a, what is the effect of alternative testing strategies on diagnostic thinking, therapeutic decisionmaking, clinical outcomes, and resource utilization?
What factors modify the comparative effectiveness of testing for patients with RLQ pain and suspected acute appendicitis?
Key Question 3.
What are the harms of diagnostic tests per se, and what are the treatment-related harms of test-directed treatment for tests used to diagnose RLQ pain and suspected acute appendicitis?