Predictive value of abnormally raised serum bilirubin in acute appendicitis: a cohort study.

Appendicitis is a common clinical diagnosis aided by biochemical, haematological and radiological investigations. The role of some investigations, such as bilirubin, is controversial but could indicate complicated appendicitis. Accurate diagnosis enables prioritisation of patients on operating lists and a possible reduction in unnecessary investigations. Methods: This is a retrospective cohort study of 1347 patients who underwent appendicectomy. Statistical analysis of serum bilirubin levels was performed according to histological classification of appendicitis. Results: Mean serum bilirubin levels; perforated/gangrenous appendicitis 20.5 mg/L (SD 12.6), inflamed appendicitis mean 17.5 mg/L (SD CI 11.1), normal appendix mean 12.6 mg/L (7.0). Kruskal–Wallis indicated bilirubin levels were significantly different (H = 128.87, df = 4, p < .001) between histological groups, and a post hoc analysis with Bonferroni adjustment showed perforated/gangrenous to be significantly higher than all other groups (p < .001). Binary logistic regression combining White Cell Count (WCC) level, C-Reactive Protein (CRP) and Bilirubin levels gave a sensitivity and specificity of .69 with AUROC = .766 (std error .015) for gangrenous/perforated. Assessment according to clinical relevance showed only 30.4% of patients with an abnormally raised bilirubin had gangrenous/perforated appendicitis. Conclusions: Serum bilirubin does not independently predict perforation/gangrenous appendicitis. Statistical analysis showed differences in mean bilirubin between histological groups however this did not relate to clinical significance as bilirubin levels were still within normal clinical limits. Diagnosis of complicated appendicitis should be made on clinical grounds, with utilization of biochemical/haematological investigations, but there should not be independent reliance on investigations such as bilirubin.