Comparison of manual and automated blood cell counts in the diagnosis of pediatric appendicitis.

摘要

Children with appendicitis present with nonspecific symptoms such as abdominal pain, fever, and vomiting. Clinicians utilize clinical findings, and laboratory and imaging tests to determine the likelihood of appendicitis. A complete blood count (CBC) with manual differential is ordered to determine if the patient has an increased number of white blood cells (WBCs) or immature WBCs present in the peripheral blood (known as a left shift). Leukocytosis (WBC count >10,000/uL), left shifted differential, elevated band count, and neutrophilia >75% are used to risk-stratify patients with suspected appendicitis. Immature granulocyte percentage (IG%) is an alternative measurement of left shift. The IG% can be obtained from an automated differential, which is faster, more reproducible, and less subject to sampling error.;A cohort definition was used to compile data including patients who presented with a chief complaint of abdominal pain, and patients who received an ultrasound of the appendix in the Primary Children's Hospital emergency department (ED). Data collected included patient age, WBC count and differential, IG%, and pathology report. A diagnosis of acute appendicitis was determined by the pathology reports. The sensitivity, specificity, and area under receiver operating characteristic curves (AUC) were determined for total WBC count (>10,000/muL), band count, IG%, and neutrophil percentage (>75%).;The total WBC count (>10,000/muL) showed the best predictive value with a sensitivity of 85.3%, specificity of 63.4%, and AUC of 80.1%. A neutrophil percentage >75% was also predictive of appendicitis with a sensitivity of 70.5%, specificity of 66.6%, and AUC of 73.4%. Band count showed no predictive value with an AUC of 57.7%. IG% was slightly more useful with a sensitivity of 68.5, specificity of 59.7%, and AUC of 66.7%.;The parameters obtained from a CBC with automated differential count: WBC count, neutrophil percentage, and IG%; were each more successful in correctly identifying pediatric patients with appendicitis than band count, which was not a reliable indicator and showed no added benefit in diagnosis. Eliminating the band count, and hence the need for a manual differential, could improve turn-around-time for patients presenting with abdominal pain without reducing the utility of the CBC.