Surveillance intensity and costs after curative-intent treatment for melanoma patients: Wide variation or consensus?

Introduction: For most cancers, little consensus exists regarding the optimal follow-up strategy after curative-intent treatment. Published follow-up strategies were reviewed for patients with cutaneous melanoma after curative-intent treatment to permit comparisons of surveillance intensity and associated costs using charge data as a proxy. Methods: A Medline literature search for the 18-year period 1989-2006 and search of major textbooks was conducted to identify follow-up strategies. Nationwide Medicare-allowed charges for five years of surveillance were calculated from the Centers for Medicare and Medicaid Services' 2004 Part B Extract and Summary System data. Results: For five years of surveillance, follow-up intensity among the 115 identified strategies ranged from no regularly scheduled office visits or diagnostic tests to 20 visits, 10 chest radiographs, 10 abdominal ultrasounds, 10 pelvic ultrasounds, 4 computed tomography scans of the brain, 4 computed tomography scans of the chest, and 4 computed tomography scans of the abdomen. Excluding the strategy of no routine follow-up, Medicare-allowed charges for five years of follow-up by TNM stage ranged from $717 to $4,569 (outlier for this group) for TNM stage I, $977 to $3,003 for TNM stage II, and $902 to $4,414 for TNM stage III. By tumor thickness, average Medicare-allowed charges were $410 for patients with tumors < 0.75 mm and $1,470 for those with tumors > 0.75 mm. Using a 1.00 mm threshold, charges were $1,026 (< 1.00 mm) and $1,703 (> 1.00 mm). For the 1.5 and 3 mm thresholds, charges were $1,740 (< 1.50 mm), $2,438 (1.51-3.00 mm), and $2,592 (> 3.00 mm). Follow-up charges for 15 studies that recommended generally applicable strategies (no differentiation by TNM stage or tumor thickness) varied by only 1.9 fold, averaging $1,705. In comparison to the 18-year analysis, subset analysis of the last 10 years (1997-2006) of published recommendations identified no changes in charge ranges overall (39.5 fold), among the thickness-specific strategies grouped (33.0 fold), or among the stage-specific strategies grouped (5.4 fold). However, for the last five years (2002-2006), charge ranges dropped to 5.5 fold overall, 4.0 fold among tumor thickness-specific recommendations, and 3.9 fold among stage-specific recommendations. Conclusion: Conservatively, Medicare-allowed charges varied by 40-fold ($113 - $4569) among strategies that advocate follow-up, reflecting marked differences in surveillance intensity among strategies, particularly stage-specific strategies. The practice of evidence-based medicine cannot support such high-intensity follow-up. Only by focusing on the last five years of published recommendations is any hint of consensus detected.