Am Fam Physician. 2008;77(6):828-830
This is part I of a two-part piece on the diagnosis of appendicitis. Part II, “Laboratory and Imaging Tests,” will appear in the April 15, 2008, issue of AFP.
Clinical Question
What are the most useful elements of the patient history and physical examination in the diagnosis of appendicitis?
Evidence Summary
Appendicitis is a relatively uncommon, but potentially serious, cause of abdominal pain in the primary care setting. An accurate diagnosis is important to prevent unnecessary surgery and avoid complications.
The probability of appendicitis depends on patient age, setting, and symptoms. A retrospective study, including three family practice centers, identified 556 adults with abdominal pain, of whom six (1.1 percent) were diagnosed with appendicitis.1 Studies of children with abdominal pain, largely in the emergency department setting, found that 10 to 25 percent of patients had appendicitis.2 In studies of patients with abdominal pain who underwent computed tomography and ultrasonography, 31 percent of children and 40 percent of adults had appendicitis.3 The percentage of patients with appendicitis ranges from 60 to 90 percent in patients who undergo surgery. Perforation rates range from 4 to 28 percent.4
Individual signs and symptoms are of some value in the evaluation of patients with suspected appendicitis. In many patients, signs and symptoms are useful at ruling in appendicitis when findings are positive or abnormal, but the absence of signs and symptoms does not necessarily reduce the risk of appendicitis (Table 12,5). In adults, right lower quadrant pain and migration of pain from the umbilicus area to the right lower quadrant are the symptoms that best predict appendicitis, whereas the absence of pain before vomiting greatly reduces the likelihood of appendicitis.5 The accuracy of history and physical examination findings is somewhat different in children. Vomiting, rectal tenderness, rebound tenderness, and fever are more helpful (greater positive likelihood ratio) in children than in adults, whereas right lower quadrant tenderness is somewhat less helpful.2 The usefulness of the finding of pain before vomiting has not been evaluated in children.
| Clinical finding | Likelihood ratio* | |
|---|---|---|
| Adults | Children | |
| Helpful for ruling in appendicitis | ||
| Right lower quadrant pain | 8.4 | — |
| Migration (periumbilical to right lower quadrant) | 3.6 | 1.9 to 3.1 |
| Initial clinical impression of the surgeon | 3.5 | 3.0 to 9.0† |
| Psoas sign | 3.2 | 2.5 |
| Fever | 3.2 | 3.4 |
| Pain before vomiting | 2.7 | — |
| Rebound tenderness | 2.0 | 3.0 |
| Rectal tenderness | — | 2.3 |
| Vomiting | — | 2.2 |
| Helpful for ruling out appendicitis | ||
| Absence of pain before vomiting | 0.02 | — |
| Absence of right lower quadrant pain | 0.18 | — |
| Absence of vomiting | — | 0.33 |
| Absence of fever | 0.42 | 0.32 |
| Absence of rebound tenderness | — | 0.28 |
The Alvarado score (also known as the MANTRELS [Migration of pain, Anorexia, Nausea/vomiting, Tenderness in the right lower quadrant, Rebound pain, Elevation of temperature, Leukocytosis, Shift to the left] score; Table 22,6) has been prospectively validated in several populations of children6–10 and adults.11 Variations include the modified Alvarado score, which excludes the left shift of the white blood cell (WBC) count,12 and the Pediatric Appendicitis Score, which substitutes right lower quadrant pain with cough, hopping, or percussion for rebound tenderness.7 However, these modifications have not been shown to perform better than the original Alvarado score.
| Clinical Finding | Points |
|---|---|
| Migration of pain to the right lower quadrant | 1 |
| Anorexia | 1 |
| Nausea/vomiting | 1 |
| Tenderness in the right lower quadrant | 2 |
| Rebound pain | 1 |
| Elevated temperature (≥ 99.1º F [37.3º C]) | 1 |
| Leukocytosis (≥ 10,000 WBCs per mm3 [10 × 109 per L]) | 2 |
| Shift of WBC count to the left (> 75 percent neutrophils) | 1 |
| Total: | _______ |
Another score, which includes nine clinical variables and does not require a WBC count, accurately predicted the likelihood of appendicitis in the initial validation study of 109 children.13 However, this and several other proposed diagnostic scores have not been prospectively validated in large, clinically relevant populations.
The Ohmann score (Table 3) includes seven clinical variables and a WBC count.14 The score was developed in a group of 870 patients at German and Austrian hospitals and was validated four months later in a second group of patients at the same hospitals. In the prospective validation, the Ohmann score successfully identified patients at low, moderate, and high risk of appendicitis.14
| Clinical Finding | Points |
|---|---|
| Tenderness in the right lower quadrant | 4.5 |
| Rebound tenderness | 2.5 |
| No difficulty with micturition | 2.0 |
| Steady pain | 2.0 |
| Leukocytosis (≥ 10,000 white blood cells per mm3 [10 × 109 per L]) | 1.5 |
| Age less than 50 years | 1.5 |
| Migration of pain to the right lower quadrant | 1.0 |
| Abdominal rigidity | 1.0 |
| Total: | _______ |
The Alvarado and Ohmann scores alone are not accurate enough to diagnose or exclude appendicitis. However, they provide a useful starting point by identifying children and adults at low, moderate, and high risk of appendicitis. Most patients at low risk can be observed without further diagnostic study; patients at moderate risk may benefit from further diagnostic testing, including imaging studies; and patients at high risk should receive urgent surgical evaluation. Part II of this series addresses diagnostic testing in the context of the Alvarado and Ohmann scores.