Why does the clinical diagnosis fail in suspected appendicitis?

OBJECTIVE: To identify systematic errors in surgeons' estimations of the importance of diagnostic variables in the decision to explore patients with suspected appendicitis. DESIGN: Prospective case series. SETTING: Two emergency departments, Sweden. PATIENTS: 496 patients with suspected appendicitis on admission, of whom 194 had a correct operation for appendicitis and 59 had a negative exploration. MAIN OUTCOME MEASURES: Predictors of a negative exploration expressed as the odds ratio (OR) for negative exploration. Variables influence on the decision to operate, expressed as the OR for operation, compared with the true diagnostic importance, expressed as the OR for appendicitis. RESULTS: Predictors of negative explorations were high ratings in variables describing pain and tenderness (patient's perceived pain, abdominal tenderness, rebound tenderness, guarding or rectal tenderness), weak or absent inflammatory response, female sex, long duration of symptoms and absence of vomiting, with OR of 1.8-3.0. Pain and tenderness had too strong an influence on the decision to operate whereas the lack of an inflammatory response, no vomiting, and long duration of symptoms were not given enough attention. There was no sex difference in the proportion of patients with non-surgical abdominal pain (NSAP) who were operated on, but NSAP was more common and appendicitis less common among women, leading to a larger proportion of negative appendicectomies among women. CONCLUSION: Negative explorations in patients with suspected appendicitis are related to systematic errors in the clinical diagnosis with too strong an emphasis on pain and tenderness, and too little attention paid to duration of symptoms and objective signs of inflammation. Rectal tenderness is not a sign of appendicitis. The risk of diagnostic errors is similar in men and women.

Repeated clinical and laboratory examinations in patients with an equivocal diagnosis of appendicitis.

In-hospital observation with repeated clinical examinations is commonly used in patients with an equivocal diagnosis of appendicitis. It is not known if repeated measurements of temperature and laboratory examinations have any diagnostic importance in this situation. The importance of repeated measurements of the body temperature, white blood cell (WBC) and differential cell counts, C-reactive protein concentration (CRP) and of the surgeon's repeated assessments was prospectively analyzed in 420 patients with an equivocal diagnosis of appendicitis at admission who were reexamined after a median of 6 hours of observation. The final diagnosis was appendicitis in 137 patients. After observation the inflammatory response was increasing among patients with appendicitis and decreasing among patients without appendicitis. The variables discriminating power for appendicitis consequently increased, from an area under the receiver operating characteristic (ROC) curve of 0.56 to 0.77 at admission, to 0.75 to 0.85 after observation. The ROC area of the surgeons' clinical assessment increased from 0.69 to 0.89. The WBC and differential cell counts were the best discriminators at the repeat examination. The change in the variables between the observations had weak discriminating power and had no additional importance in addition to the actual level at the repeat examination. To conclude, the diagnostic information of the temperature and laboratory examinations increased after observation. Repeated controls of the body temperature and laboratory examinations are therefore useful in the management of patients with equivocal signs of appendicitis, but the result of the examinations must be integrated with the clinical assessment.

Diagnostic value of disease history, clinical presentation, and inflammatory parameters of appendicitis.

The clinical diagnosis of appendicitis needs to be improved, as up to 40% of explorations for suspected appendicitis are unnecessary. The use of body temperature and laboratory examinations as diagnostic aids in the management of these patients is controversial. The diagnostic power of these variables compared to that of the disease history and clinical findings is not well studied. In this study we prospectively assessed and compared the diagnostic value of 21 elements of the history, clinical findings, body temperature, and laboratory examinations in 496 patients with suspected appendicitis. The diagnostic value of each variable was compared from the area under the receiver operating characteristic (ROC) curve and the likelihood ratios (LR). Logistic regression was used to analyze the diagnostic value of a combination of variables and to analyze independent relations. No single variable had sufficiently high discriminating or predicting power to be used as a true diagnostic test. The inflammatory variables (temperature, leukocyte and differential white blood cell (WBC) counts, C-reactive protein) had discriminating and predicting powers similar to those of the clinical findings (direct and rebound abdominal tenderness and guarding). Anorexia, nausea, and right-sided rectal tenderness had no diagnostic value. The leukocyte and differential WBC counts, C-reactive protein, rebound tenderness, guarding, and gender were independent predictors of appendicitis with a combined ROC area of 0. 93 for appendicitis. This showed that inflammatory variables contain important diagnostic information, especially with advanced appendicitis. They should therefore always be included in the diagnostic workup in patients with suspected appendicitis.