Response to letter by Robert M. Hamilton based on "Prophylactic radiofrequency ablation in asymptomatic patients with Wolff-Parkinson-White is not yet a good strategy: a decision analysis" by Chevalier et al.

Response to Letter by Robert M. Hamilton Based on “Prophylactic Radiofrequency Ablation in Asymptomatic Patients With Wolff– Parkinson–White Is Not Yet a Good Strategy: A Decision Analysis” by Chevalier et al We thank Dr Robert R. Hamilton for his insightful comments and are very pleased that he brings his own experience into the difficult debate on the indication of ablation of the asymptomatic patients with Wolff–Parkinson–White. We feel that some elementary notions about decision analysis have to be restated. There are 2 main arguments for using decision analysis models and to find out how the decision process for a given patient can be improved: a randomized prospective study comparing abstention versus ablation will never be done, and the risk:benefit ratio of both strategies has never been quantified. Interpretation of the results of our study will be facilitated with the following reminders: decision analysis is not a surrogate to a clinical trial, and our study aimed to allow patients and caregivers to participate in the decision, to generate hypothesis, and to highlight important points in the decision process. As mentioned in the article, the target population of our study consisted of 20to 40-year-old asymptomatic patients with Wolff– Parkinson–White without structural fatal heart disease or a family history of sudden cardiac death. The literature is scarce in the field of long-term prognosis of patients with Wolff–Parkinson–White, and the task of evaluating the published clinical data is daunting. In the present work, we took the expert opinion value when there were no data available. This is stated in the text. In addition, we used expert opinion when the probabilities of the literature were based on a single article. The articles in the literature were not necessarily the same people or the same method of recruitment or the same time patient monitoring. The experts interviewed are recognized for their competence, and their expertise seemed more reliable than a single study in the literature. Furthermore, data from the literature were used in the sensitivity analysis for the Tornado diagram, and the results do not change the conclusions that we found with the values of the experts. Because accessory pathway ablation is performed by physicians with different levels of competence and experience, we choose the real-world setting, the worst case scenario. Reducing the complication rate in our model still did not favor ablation. Finally, it is clearly stated in the article that our conclusions are specific to the statistical model and the population we tested. We end by quoting George E.P. Box who said “essentially all models are wrong some are useful.” Decision analysis is not an exception to the rule. The practical question is how wrong does it have to be, not to be useful. We think we have provided some hints that will be helpful for the physician in making recommendations for an asymptomatic patient with Wolff–Parkinson–White.