## Abstract

BACKGROUND: Obtaining laboratory studies to screen for a coagulation disorder prior to tonsillectomy and/or adenoidectomy (T&A) is a controversial procedure. The American Society of Pediatric Otolaryngology advises that such studies only be performed when indicated by history or physical exam. Nevertheless, many surgeons in clinical practice proceed with testing on all children scheduled for T&A. Several observational studies of pre-operative testing have been performed over the past four decades with conflicting results. The cost-effectiveness of this screening has not been studied in a rigorous, formal manner. We constructed a decision analysis model to address the costs and health outcome states of coagulation screening strategies for children undergoing T&A.

METHODS: A 14-day Markov model was constructed to test three pre-operative screening strategies:

1. test all children for coagulation disorders;

2. test only those children with a pertinent history or physical exam (PE) finding; and

3. perform no pre-operative testing on any child.

We performed a literature search to estimate the prevalence of abnormal pre-operative coagulation tests, the prevalence of true coagulation disorders among those patients, the rates of bleeding following T&A, and the mortality following T&A. From published sources, we obtained cost data on laboratory studies, hospital admissions (surgical and observational), hematology consultation, medications for prophylaxis and treatment of a bleeding disorder, and lost parental wages. We estimated quality of life (utility) data from published national sources. Parameters were varied widely in sensitivity analyses based upon the range of reported values. Using a societal perspective and a cycle length of one day, we compared the three strategies based on total costs and quality-adjusted life-years (QALYs).

RESULTS: Total costs for the three strategies were $3200 for testing all children,$3083 for testing only children with a history or PE finding, and \$3077 for not testing children. Total utilities followed a reversed pattern with 0.02579 QALYs for testing all, 0.02654 QALYs for testing some, and 0.02659 QALYs for not testing. These results indicate that the strategy of not testing children prior to surgery dominates the others; it is both less expensive and has a higher utility. Cost-effectiveness ratios were sensitive to variation in the overall prevalence of post-operative bleeding, the cost of post-operative care, and the cost of diagnostic coagulation testing after a post-operative hemorrhage. However, the strategy of not testing children pre-operatively was dominant throughout all sensitivity analyses.

DISCUSSION: Our results demonstrate that universally screening all children for coagulation disorders prior to T&A is not cost-effective. Furthermore, not performing any preoperative testing appears to be the most cost-effective strategy. This study does not prove the superiority of one strategy over another, though it does offer advantages over single institution observational studies. Given the short time duration of the model, absolute cost and utility differences between strategies are small. However, these differences are persistent in sensitivity analyses, indicating that the model is robust. The analysis provides evidence that suggests pre-T&A coagulation testing in children should be used judiciously at most. Detailed conversations between surgeon, patient, and family about the risks, benefits, and costs of such testing should determine its selective use.

Disclosures: No relevant conflicts of interest to declare.

## Author notes

Corresponding author