Methods to include environmental impacts in a cost-effectiveness analysis: a case study of PET/CT surveillance imaging after treatment for stage III melanoma

Objectives: To demonstrate methods to include carbon emissions in a cost-effectiveness analysis of different positron emission tomography and computed tomography surveillance imaging schedules for detecting metastasis in patients treated for stage III melanoma.

Methods: Decision-analytic modeling was used to compare 12-monthly imaging with no routinely scheduled imaging. The health outcome was the proportion of patients without a false-negative result for a distant metastasis. Health system costs were included in 2023 Australian dollars. A 5% discount per year was applied. Carbon emissions were obtained from existing studies. Four methods were demonstrated: (1) reporting emissions in parallel, (2) the incremental carbon footprint cost ratio, (3) the incremental carbon footprint effectiveness ratio, and (4) integrating emissions as a cost in the incremental cost-effectiveness ratio (ICER).

Results: In the base case, 12-monthly imaging was more expensive and more effective than no scheduled imaging, with an ICER of $71 757 per false-negative avoided. Carbon emissions were higher for 12-monthly imaging (1497 kg CO₂-e) than no routine imaging (1037 kg CO₂-e). The incremental carbon footprint cost ratio was 0.2 kg CO₂-e per additional dollar spent. The incremental carbon footprint effectiveness ratio was 17 331 kg CO₂-e per false-negative avoided. Valuing emissions using New South Wales carbon values increased the ICER modestly to $73 919 per false-negative avoided. Using the mortality-accounting social cost of carbon increased the ICER to $107 297 per false-negative avoided.

Conclusions: Parallel methods are difficult to interpret without agreed value thresholds. Integrated methods meaningfully altered the results of the base-case analysis when higher monetary values of carbon emissions were used.