The diagnosis of appendicitis can be difficult, occasionally taxing the diagnostic skills of even the most experienced surgeon. Likewise, the judgemental decisions in the management of patients with appendiceal inflammation or abscess can be difficult. The patient with appendicitis must first recognize that he has an episode of pain that is unique, and then present to a physician who recognizes the condition. Delays in diagnosis arise from errors on the part of either patient or physician, and all delays complicate the illness.

Patients with appendicitis generally present with a typical pattern of distress. The most common initial symptom is pain, typically diffuse, epigastric, or periumbilical in distribution and gnawing in character. Some patients have no pain, but feel unwell, with a severe upset stomach. Shortly after the initial pain, anorexia, nausea, and occasionally, vomiting develop. Anorexia is the most constant symptom of appendicitis. If it is absent, the diagnosis should be questioned. Vomiting, if present, occurs early in the attack, usually several hours after the initial pain. Vomiting occurring before the onset of pain, makes a diagnosis of appendicitis questionable, and a viral illness more likely.

The initial pain is due to the early obstruction, dilation, and infection of the appendix. As this visceral pain is mediated through visceral pain fibres, it is poorly localized. When the infection in the appendix becomes established and transmural, the serosa of the appendix and parietal peritoneum are involved, causing a localized pain from somatic pain fibres in the abdominal wall at the area of the appendix. Because the appendix is generally located one-third of the way from the anterior, superior iliac spine to the umbilicus (McBurney's point), the area of greatest pain is generally in the right lower quadrant.

Patients may present with localized pain in the right upper quadrant from a long appendix, in the left lower quadrant if malrotation is present, and in the anterior wall of the rectum if the appendix is located in the pelvis. The most common location of ‘atypical’ somatic pain is the right flank in patients with a retrocaecal appendix. Somatic pain reflects the location of the appendix; if the development of symptoms and signs suggest appendicitis, an atypical location of maximal pain does not rule out the diagnosis.

Many patients with appendicitis have pain on motion and pain present during the trip to hospital may abate when motion ceases. These events do not mean that their illness is resolving: pain on motion is caused by motion of the inflamed appendix against the peritoneum. The patient presenting after perforation of the appendix complains of generalized pain. The history is usually a sequence of generalized pain that became localized to the right lower quadrant and later became generalized.

Less common symptoms include diarrhoea, which may occur early or late in the course of appendicitis. Early in the course of appendicitis patients may have one or two loose bowel movements, or they may have an episode of massive evacuation of normal stool. This sequence represents a response to visceral pain and is usually limited to one or two episodes, rather than the persistent diarrhoea caused by viral or bacterial infection. Later in the course of appendicitis, diarrhoea may return because of irritation of the rectum by an inflamed pelvic appendix. This diarrhoea is mucoid and persistent; it is accompanied by tenesmus and can easily be mistaken for gastroenteritis if proper attention is not paid to the history. Testicular pain or retraction of the testes may occur at any time in the course of appendicitis, the appendix and testicle both being innervated by the tenth thoracic spinal segment.

The concept of recurrent appendicitis is gradually being accepted in the United States, and patients often describe previous episodes of pain that were the same as the present in all aspects except severity. Surveys suggest that this sequence occurs in about 25 per cent of patients.

Tenderness over the site of the appendix is the sine qua non of appendicitis. However, tenderness may be absent early in the course of the illness or unelicitable in obese individuals. Tenderness over the appendix is due to inflammation of the serosa of the appendix and the overlying parietal peritoneum; so early in appendicitis there may not be enough inflammation to cause this diagnostic finding, although it is unusual for patients to present this early in the course of their illness. More commonly, difficulty in eliciting tenderness over the appendix arises from its retrocaecal location. Patients with a retrocaecal appendix may experience some mild right-sided or right flank tenderness. In extremely obese people, localizing tenderness is difficult, simply because palpation of the abdomen is cushioned by subcutaneous fat. In addition, the entire panniculus of the abdomen may be caudal to the peritoneal cavity. When the abdomen is palpated in such a patient, the normal topographic anatomical landmarks must be ignored.

While much has been made of tenderness at McBurney's point, this is not as reliable as the presence of tenderness at some point in a patient with a suggestive history. Classically, the area of maximal tenderness will be one-third of the way from the anterior superior iliac spine to the umbilicus, but in fact it will be wherever the appendix is in the individual patient.

Many surgeons rely on the demonstration of local muscular rigidity to make a diagnosis of appendicitis. This muscular rigidity is produced by inflammation of the parietal peritoneum overlying the appendix, and thus takes longer to develop than local tenderness. Many patients with early appendicitis demonstrate little or no rigidity of the abdominal wall. Subtle rigidity can be demonstrated by palpation of the left lower quadrant while talking with the patient. The palpating hand is slowly moved toward the right lower quadrant (or area of maximal pain), which is gently palpated. This sequence helps to differentiate true rigidity from the voluntary spasm that occurs from nervousness. Similarly, Rovsing's sign, or pain in the right lower quadrant upon palpation of the left lower quadrant, is a sign of local peritonitis. Severe, well-established rigidity of the abdominal wall is a sign of well-established local peritonitis and of impending (or previous) perforation.

Another useful sign to establish the presence of local peritonitis is the shake test. Most surgeons perform this by grasping the iliac wings and shaking the pelvis from side to side. The patient complains of pain at the site of the appendix if local peritonitis is present. It is more helpful to kick or push the stretcher gently while watching the patient's face: a grimace is a sure sign of local peritonitis.

The elicitation of signs of local peritonitis may be difficult in patients with a retrocaecal or pelvic appendix. The psoas sign, pain caused by extending the thigh to stretch the psoas muscle, is generally positive in patients with a retrocaecal appendix and local peritonitis. If the appendix is adjacent to the obturator internus muscle, stretching of this muscle by external rotation of the hip elicits severe pain and spasm. With a true, inflamed pelvic appendix, the only signs of local peritonitis may be found on rectal examination in men or vaginal examination in women. It is often impossible to differentiate pelvic appendicitis from pelvic inflammatory disease by physical examination alone, but accurate history will generally solve the dilemma.

Patients who present late with appendicitis may have only generalized tenderness and rigidity. This is a sure sign of perforation, but without an accurate history, the diagnosis remains obscure.

Hyperaesthesia or dysaesthesia may occur in patients with non-perforated appendicitis. Although these are inconstant findings, when present they occur on the right side of the abdomen, in the distribution of the tenth, eleventh, and twelfth thoracic nerves. Paraesthesiae are best elicited by light scratching with a sharp sterile needle.

Fever is a late physical finding in appendicitis. Before perforation, body temperature is usually no more than 39 to 39.5°C, but with perforation may rise to 40 to 41°C. If fever has been present since the onset of the illness, consider other causes.

Laboratory examinations are rarely helpful in the diagnosis of appendicitis. Leucocytosis is common, usually in the range of 11000 to 17000 mm³. However, this occurs once appendicitis is well established, and the illness should generally be diagnosed before it develops. A leucocytosis over 20000 mm³ suggests perforation of the appendix or another diagnosis. Urinalysis is unhelpful in the diagnosis or exclusion of appendicitis. Mild pyuria may occur due to irritation of the bladder by a pelvic appendix or irritation of the ureter by a retrocaecal appendix. Thus, in patients with symptoms suggestive of either appendicitis or urinary tract infection, urinalysis is not diagnostic of either condition.

Radiographic techniques have been recommended in the evaluation of patients with possible appendicitis. Plain abdominal films demonstrate a faecalith in the area of the appendix in 10 per cent of patients with appendicitis. This finding is highly suggestive of appendicitis in a patient with a compatible history and physical examination. Plain abdominal films may reveal other abnormalities, such as localized ileus in the right lower quadrant, soft tissue density in the right lower quadrant, or free intraperitoneal air. These are so non-specific as to be of no value in the diagnosis of appendicitis.

Barium enema has long been recommended for the evaluation of possible appendicitis. Findings suggestive of appendicitis include spasm of the terminal ileum or caecum, external compression of the caecum, and non-filling or partial filling of the appendix. While several series suggest diagnostic accuracy rates of 90 per cent, review of these studies discloses that about one-quarter of patients with an abnormal barium enema have a normal appendix at the time of exploration. Probably the best case that can be made for barium enema is that if the entire appendix is filled, appendicitis is virtually ruled out.

Ultrasound and computed tomography (CT) have recently been used in the evaluation of patients with suspected appendicitis. Ultrasound is performed with high resolution linear array transducers and gentle but thorough compression of the abdomen, with the goal of eliminating any gas-filled bowel between the transducer and the appendix. Ultrasound criteria for the diagnosis of appendicitis are a non-compressible appendix, surrounded by a hypoechoic thickened wall more than 2 mm in diameter. In addition, the maximal diameter of the visualized appendix should exceed 6 mm. Using these criteria, sensitivity rates of 75 per cent and specificity rates of 100 per cent have been reported. The sensitivity is usually considerably lower in patients with perforated appendicitis, as the rigidity of the abdominal wall prevents its adequate compression. Realistically, a sensitivity of only 75 per cent is too low to be acceptable in a diagnostic test. Ultrasound has not, therefore, become widely used in the diagnosis of appendicitis, although a totally normal appendix visualized by ultrasound makes appendicitis unlikely. CT has been touted by radiologists but has not been widely accepted by surgeons for the diagnosis of appendicitis. CT findings suggestive of appendicitis include pericaecal increased density of pericolic fat, appendicoliths, a thickened appendix, and a periappendiceal abscess. Oral contrast media are necessary to enable adequate evaluation of the gastrointestinal tract. Reported sensitivity and specificity are approximately 80 per cent, making the test of limited value in establishing a diagnosis. Its real value is in the evaluation of patients with atypical pain that is not particularly suggestive of appendicitis, and in the evaluation of possible appendiceal abscess.

In evaluating a patient with abdominal pain suggestive of appendicitis, other possible causes of abdominal pain must be considered. From a management viewpoint, the various diagnoses can be divided into those that require surgery and those that do not. The former include perforated carcinoma of the right colon, perforated right colonic diverticulitis, perforated ulcer (with tracking of visceral contents along the right gutter to the right lower quadrant), sigmoid diverticulitis (especially with a mobile sigmoid colon), Meckel's diverticulitis, ectopic pregnancy, and ovarian torsion. Any of these disorders may be indistinguishable from appendicitis at presentation, but as each usually requires surgery for resolution, their differentiation from appendicitis is not crucial.

Surgery has limited value in the management of many other disorders which are almost indistinguishable from appendicitis. These include pelvic inflammatory disease, Mittleschmerz pain (from ruptured ovarian follicular cyst), ‘mesenteric adenitis’, viral gastroenteritis, bacterial gastroenteritis, acute Crohn's ileitis, and typhoid. Pelvic inflammatory disease is more likely than appendicitis to occur during the proliferative phase of the menstrual cycle. It has a longer duration of symptoms, higher fever, greater leucocytosis, and less well localized pain, with more pelvic pain and cervical motion tenderness than appendicitis. Mittleschmerz pain can usually be diagnosed by its occurrence at ovulation and by the fact that most patients have experienced painful ovulation before. Although fever is uncommon, tenderness and leucocytosis may occur. The diagnosis is usually made by exclusion—the patient is observed and the pain resolves. ‘Mesenteric adenitis’ is also a diagnosis of exclusion; the patient has all the typical features of appendicitis, but at exploration is found to have a normal appendix and some enlarged lymph nodes in the mesentery of the distal ileum. The cause of the illness is obscure. Viral and bacterial gastroenteritis are indistinguishable clinically, and the diagnosis of bacterial gastroenteritis is made only when the results of stool cultures become available (usually after the patient has been treated or the illness has resolved spontaneously). Viral and bacterial gastroenteritis may usually be differentiated from appendicitis by the massive diarrhoea present in typical gastroenteritis. Diarrhoea associated with appendicitis is rarely massive or prolonged. Most patients with gastroenteritis present with diffuse abdominal pain that rarely becomes well localized; if any tenderness develops, it is usually mild and generalized. It is most unusual to find a patient with true right lower quadrant tenderness due to gastroenteritis.

The symptoms and signs of appendicitis are altered when the appendix is retrocaecal. Pain is generally not as severe as that in patients with an abdominal or pelvic appendix, and pain rarely becomes well localized to the right lower quadrant. Many patients with a retrocaecal appendicitis have a history of generalized abdominal pain, fairly constant for several days, never localizing, and never really interfering with normal activities. Occasionally, the pain becomes localized to the right flank or to the right upper abdomen. Patients with retrocaecal appendicitis have tenderness in the area of the appendix, provided that the area can be palpated. In true retrocecal appendicitis, the area of tenderness is in the flank or costovertebral angle, simulating pyelonephritis. Occasionally the tenderness is subcostal, because the tip of the appendix is located at or near the hepatic flexure. In an obese patient tenderness can often not be elicited since the retrocaecal appendix is so well encased in retroperitoneal fat that it cannot be felt.

The combination of abdominal pain, more or less localized to the right side of the abdomen, with some aspect of nausea or anorexia, and no other suggestive diagnosis should raise the suspicion of retrocaecal appendicitis. The presence of fever or leucocytosis makes the diagnosis of retrocaecal appendicitis even more likely. Unfortunately, most such patients are dismissed by physicians as having gastroenteritis, a most unusual occurrence in the absence of significant diarrhoea. Eventually, such patients present with a retrocaecal abscess due to perforation of the appendix (Fig. 1) 1131.

Elderly patients tend to present late in the course of appendicitis and with less well-defined symptoms; the incidence of perforation is therefore higher. Most elderly patients with appendicitis have a history of several days of poorly defined abdominal pain, anorexia, and fever. Rarely is the pain well localized to McBurney's point; rather, it is generally described as being in the right side of the abdomen. Most old patients have fever and abdominal tenderness at the time of presentation with appendicitis. Tenderness is either in the right lower quadrant, the right flank, or is diffuse from the effects of free perforation. The psoas sign and obturator sign are unhelpful in old patients because almost every manoeuvre is equally uncomfortable.

The diagnosis of appendicitis in pregnancy is difficult because the appendix is displaced by the gravid uterus. Early in the course of pregnancy the appendix remains in its normal position, and diagnosis is routine. By the middle of the second trimester of pregnancy, however, the appendix becomes displaced superiorly, attaining a position in the right upper flank or epigastrium. Appendicitis may be easily mistaken for pyelonephritis or cholecystitis.

The abdominal wall is lifted from the appendix by the gravid uterus, and muscular laxity occurs: the abdominal findings associated with peritoneal irritation by the inflamed appendix may therefore be fewer than one might expect in the non-pregnant individual. Leucocytosis is a normal physiological response of pregnancy (up to 12 500 leucocytes/mm³) and cannot be relied upon to help confirm the diagnosis of appendicitis. White blood cell counts as high as 25 000 leucocytes/mm³ are not unusual in pregnant women with appendicitis.

The management of non-perforated appendicitis is surgical removal. If the diagnosis is clear, an incision should be made over the point of maximal tenderness, generally at McBurney's point. Many surgeons favour a true McBurney's incision, with an oblique skin incision at McBurney's point and splitting of the external and internal oblique muscles in the line of their fibres; others prefer a transverse skin incision with muscle splitting. The skin incision need be only 3 to 6 cm long (depending on the patient's build), and can be easily extended by cutting the rectus fascia and retracting the rectus muscle medially. The taenia of the colon are followed to the base of the appendix, and blunt dissection is used to free the appendix from its surrounding inflammatory tissue. The mesoappendix is divided between clamps and ligated with an absorbable suture. The base of the appendix is divided and is also ligated with absorbable suture material. The base of the appendix may be inverted using either a purse-string suture or a ‘Z-stitch’ (Fig. 2) 1132, although there are no firm data to suggest that inversion of the stump produces better healing than simple ligation. The muscle layers of the abdominal wall are closed with absorbable suture; the skin is closed with a monofilament suture that should be placed deep enough to eliminate dead space in the subcutaneous tissue (Fig. 3) 1133.

For patients in whom the diagnosis of appendicitis is suspected, but not definite, several options are available. The most attractive of these is simply to admit the patient to the hospital for further observation. Over the course of 12 to 24 h one should be able to characterize the course of the illness well enough to determine whether an exploration for possible appendicitis, further tests (for example, computerized tomography), or discharge is in order. Although there is some risk of perforation of the appendix during this period of observation, if the patient is re-examined frequently, one should be able to avoid this complication. The risk of perforation is low in the first 24 h of symptomatic appendicitis; it is therefore safe to observe patients early in the course of their illness, when physical findings may not be well developed.

Another option is laparoscopy. This examination is commonly used in patients with symptoms and signs suggestive of both appendicitis and pelvic inflammatory disease. To exclude the diagnosis of appendicitis by laparoscopy, the entire appendix must be seen and must be normal. If the appendix is retrocaecal, mobilization of the right colon and retraction of the appendix will be necessary. In a patient with an unclear diagnosis, laparoscopy may be performed as both a diagnostic and a therapeutic manoeuvre. The technique for laparoscopic appendectomy is straight forward. A high flow rate laparoscopic insufflator, video camera and monitor, and adequate assistance are required. The abdomen is filled with carbon dioxide, a large trocar (10, 11, or 12 mm) is placed infra-umbilically and the laparoscope with camera is placed through it. A large trocar is placed through the left rectus sheath, avoiding the inferior epigastric vessels, halfway between the pubis and the umbilicus. An additional (5 mm) trocar is placed suprapubically. Through these trocars, the appendix is freed from surrounding tissues. Once free, it is usually helpful to place an additional small trocar in the right upper quadrant to hold the appendix up toward the abdominal wall. The mesoappendix is dissected and bleeding is controlled with clips or suture. The base of the appendix may be controlled with pretied suture loops, sutures, or endoscopic staples. The laparoscope is moved to the left-sided large trocar, and the appendix is removed through the umbilical trocar by widening the fascial opening in the abdominal wall.

Patients seen late in the course of appendicitis often have a palpable mass in the right lower quadrant. These account for about 3 per cent of patients, and traditionally they have been managed with antibiotics and bed rest, reserving surgery for those who do not respond to such conservative therapy. Non-operative therapy is successful in 80 to 90 per cent of patients; causes of failure include sepsis, unresolved abscess, and small bowel obstruction from adhesions to the inflammatory mass. The rate of failure of non-operative therapy is higher in patients with a well defined abscess at presentation than in those with diffuse inflammation, and abscess can be reliably diagnosed using CT. If an abscess is found, operative drainage or CT-directed drainage is indicated. Percutaneous, radiologically guided drainage is successful in 85 per cent of patients and avoids the need for operative intervention in patients with active sepsis. Most patients treated in this way can be discharged from the hospital in 7 to 10 days.

Traditionally, all patients with an appendiceal mass underwent appendectomy 6 to 8 weeks following their acute illness. However only 20 per cent of patients develop another episode of appendicitis following treatment of an appendiceal abscess or diffuse inflammation, prompting some surgeons to abandon interval appendectomy. The morbidity associated with interval appendectomy is so low, however, that there is a case for it being performed in patients for whom surgery does not pose a risk because of coexistent medical illnesses. If an interval appendectomy is not performed, patients beyond young adulthood should undergo a barium enema examination or colonoscopy to exclude the possibility of a locally perforated right colon carcinoma.

All patients with suspected appendicitis should receive broad-spectrum antibiotics preoperatively. The duration of antibiotic therapy is the subject of some controversy, but it should be governed by the severity of the appendicitis. In the patient with early, acute appendicitis, without purulence or gangrenous changes, 24 to 48 h of antibiotic therapy should suffice. In those with perforated appendicitis, appendiceal abscess, or gangrene of the appendix, 7 to 10 days of antibiotic therapy seems prudent. The choice of antibiotics seems to be less important in the prevention of wound infection than the timing of their administration. Effective regimens include cefoxitin alone, ampicillin-sulbactam, ampicillin-clavulanic acid, cefotetan, cefoperazone, an aminoglycoside with metronidazole, aminoglycoside and clindamycin, and imipenem.

Balthazar EJ, Megibow AJ, Hulnick D. CT of appendicitis. Am J Roentgenol 1989; 153: 687–91.

Bauer T, Vennits B, Holm B. Antibiotic prophylxasis in acute nonperforated appendicitis. Ann Surg 1989; 144: 338–40.

Berry J, Malt RA. Appendicitis near its centenary. Ann Surg 1984; 200: 567–75.

Burns RP, Russell WL. Appendicitis in mature patients. Ann Surg 1985; 201: 695–704.

Crabbe MM, Norwood SH, Robertson HD. Recurrent and chronic appendicitis. Surg Gynecol Obstet 1986; 163: 11–13.

Ferzli GE, Ozuner G, Davidson PG. Barium enema in the diagnosis of acute appendicitis. Surg Gynecol Obstet 1990; 171: 40–2.

Fisher KS, Ross DS. Guidelines for therapeutic decisions in incidental appendectomy. Surg Gynecol Obstet 1990; 171: 95–8.

Gotz F, Pier A, Bacher C. Modified laparoscopic appendectomy in surgery. Surg Endosc 1990; 4: 6–9.

Horowitz MD, Gomez GA, Santiesteban R. Acute appendicitis during pregnancy. Arch Surg 1985; 120: 1362–7.

Puylaert JBCM, Rutgers PH, Lalisang RI. A prospective study of ultrasonography in the diagnosis of appendicitis. N Engl J Med 1987; 317: 666–9.

Silen W. Cope's Early Diagnosis of the Acute Abdomen. New York: Oxford University Press, 1991.