The use of intensive radiological assessments in routine surveillance after treatment for head and neck cancer: An economic evaluation

Background: There is uncertainty around the optimal surveillance of head and neck cancer patients after the primary curative treatment. This study aims at assessing the cost-effectiveness of a post-treatment programme of frequent radiological assessments (maximal approach) compared with a symptom-driven surveillance (minimal approach). Materials and methods: A decision-analytic Markov model is developed to assess the cost utility of two alternative follow-up programmes with a lifetime horizon. The two interventions differ in the number of radiological assessments (i.e. magnetic resonance imaging, computed tomography and positron-emission tomography) performed over a 5-year period. Clinical and utility parameters are derived from published and unpublished literature and expert opinion. The cost analysis is conducted from the perspective of a major Italian region's health care system. Cost-effectiveness results are expressed as incremental cost per life year gained (LYG) and per quality-adjusted life year (QALY) and checked against a cost-effectiveness threshold of (sic)25,000-40,000 per QALY. One-way, two-way and probabilistic sensitivity analyses are carried out. Results: In the base-case analysis, an intensive programme of radiological investigations leads to 0.10 additional QALYs (0.15 LYG) and an increase in costs of (sic)1903 per patient compared with those of a minimal option, resulting in an incremental cost of (sic)19,951/QALY gained ((sic)13,123/LYG). In probabilistic sensitivity analysis, 72% of the results lie below the (sic)40,000 threshold (55% below (sic)25,000). Conclusions: An intensive post-treatment follow-up with scheduled radiological assessments over time might be cost-effective compared with symptom-driven surveillance in head and neck cancer patients. Further research is needed to check these results in empirical studies or real-world settings. (C) 2018 Elsevier Ltd. All rights reserved.