LETTER TO THE EDITOR In Reply: Cost Effectiveness of Staging Procedures in Patients with Stage I/II Melanoma TO THE EDITORS: Van Akkooi and colleagues 1 suggest that ultrasound- guided fine-needle aspiration cytology (FNAC) is cost effective compared with sentinel node biopsy for the staging of melanoma. Unfortunately there are a number of issues that mean their cost-effectiveness analysis is overly simplistic and therefore potentially misleading. First and most importantly, the estimates of effective- ness and test performance of ultrasound-guided FNAC are based on a single-center observational cohort study. 1 Noncomparative single-center data of test performance have uncertain generalizability beyond the center where the data were collected. The test sensitivity reported by van Akkooi et al. was 65%, but other centers have reported much lower sensitivity values (39, 34, and 24%). 2–4 Better estimates of test performance for use in economic evalua- tions come from meta-analyses or at least a range of cohort studies from multiple centers that represent patients with different clinical presentations, including primary lesions of varying thickness and anatomical location. Second, there has been no attempt to consider the longer-term outcomes of staging using ultrasound-guided FNAC. These include clinical outcomes such as disease recurrence, impact on survival or quality of life, as well as potential adverse events such as pain, or anxiety caused by inconclusive results. These outcomes should be prospec- tively assessed from the viewpoint of the patient. Third, the methods by which the vastly different esti- mates of procedural costs for sentinel node biopsy (€1254.83) and ultrasound-guided FNAC (€168.61) were obtained are not explicitly described or referenced. The costs did not appear to be based on individual patient costs (as ours were for the sentinel node biopsy study 5 ), and there was no indication of the resources that were actually used (e.g., staffing, equipment, maintenance and over- heads), nor was there any statement of perspective, (i.e., did the figures represent total costs to the patient, the hospital, or to society?). Direct and indirect medical costs such as the fees associated with ultrasound and its sub- sequent follow-up should be measured for an appropriate time horizon. Finally, the economic evaluation performs no sensitivity analyses performs no sensitivity analyses to assess whether the results are robust to changes in variables. Additional analyses should be performed using a range of estimates that might include best and worse case scenarios repre- senting the diversity in clinical presentation, costs, and test performance. 6 Because these components are not reported we cannot assess the veracity of the claim that ultrasound- guided FNAC is an effective and cost-saving staging pro- cedure compared with sentinel node biopsy. We agree that it is important to consider the costs and effects of new staging procedures in melanoma, as has been done by van Akkooi and colleagues. However, a compre- hensive economic evaluation is required if readers are to adequately assess the applicability of the cost-effectiveness results, and therefore the procedure, to their own clinical situation. A future evaluation should include: a clear description of the alternatives and the economic perspec- tive, a full assessment of outcomes including benefits and risks, a transparent assessment of costs relevant to the specified perspective, an overall measure of cost effec- tiveness (e.g., € per life year saved), and sensitivity analyses to help characterize the uncertainty around the results. 6 Rachael L. Morton, MScMed (Clin Epi) 1 , Kirsten Howard, PhD 1 , and John F. Thompson, MD 2,3 1 Sydney School of Public Health, The University of Sydney, Sydney, NSW, Australia; 2 Melanoma Institute Australia, Royal Prince Alfred and Mater Hospitals, Sydney, NSW, Australia; 3 Discipline of Surgery, The University of Sydney, Sydney, NSW, Australia e-mail: rachaelm@health.usyd.edu.au Published Online: 5 December 2009 Ó Society of Surgical Oncology 2009 REFERENCES 1. Voit CA, van Akkooi ACJ, Scha ¨fer-Hesterberg G, Schoengen A, Schmitz PI, Sterry W, et al. Rotterdam criteria for sentinel node (SN) tumor burden and the accuracy of ultrasound (US)-guided Ann Surg Oncol (2010) 17:663–664 DOI 10.1245/s10434-009-0823-8