Post-tonsillectomy bleeding: a meta-analysis.

OBJECTIVE/HYPOTHESIS To reconcile conflicting reports and opinions of the value of preoperative coagulation studies for patients undergoing tonsillectomy. STUDY DESIGN Meta-analysis: Articles were identified by MEDLINE search, references from review articles, textbook chapter, and retrieved reports. Independent observers selected prospective trials of patients undergoing tonsillectomy or adenoidectomy and tonsillectomy. Retrospective studies meeting other inclusion and exclusion criteria were included for sensitivity analyses of results. METHODS Data were abstracted from studies for an end point of bleeding with normal and abnormal coagulation tests. Four prospective studies met all inclusion and exclusion criteria. These four studies were used in the data synthesis. An additional eight retrospective studies met all other criteria and were used in the sensitivity analysis. RESULTS Pooled analysis of 3384 patients revealed a rate of 3.3% (95% confidence interval [CI], 2.5%-4.1%) for post-tonsillectomy bleeding in patients with normal coagulation studies. A rate of 8.7% (95% CI, 1.5%-15.9%) was obtained for bleeding in patients with abnormal coagulation studies. No significant rate difference in post-tonsillectomy bleeding was demonstrated. CONCLUSION There is no difference in the rate of post-tonsillectomy bleeding in patients with abnormal coagulation studies as compared with patients with normal coagulation studies obtained preoperatively.