In the comparison between paclitaxel+cisplatin versus cisplatin-based regimens without paclitaxel, the cost-effectiveness ratio was not calculated because no survival difference was found in the analysis. In the comparison between high-dose treatments with hematopoietic rescue versus standard cisplatin-based regimens without paclitaxel, the survival gain was 2.34 (discounted) years per patient and the cost-effectiveness ratio was $25,641 (discounted) per discounted life year gained.
In the first sensitivity analysis, use of the less impressive survival data for transplanted patients resulted in a cost-effectiveness ratio of $47,619 per life year gained.
Despite the wide variations in the cost of hematopoietic rescue in the second sensitivity analysis, the cost per life year gained using high-dose treatments remained in the range of an acceptable cost-effectiveness profile (range: $8,574 to $51,282 per life year gained).
The third sensitivity analysis utilised an annual discount rate of 3% and produced a survival gain of 2.48 years per patient (MLS=5.22 years for transplantation and 2.74 years for platinum regimens) and a cost-effectiveness ratio of $24,193 per life year gained.
The fourth sensitivity analysis introduced adjustment for quality of life and compared high-dose treatments with hematopoietic rescue versus standard cisplatin-based regimens without paclitaxel. The gain was 1.88 QALYs per patient (mean lifetime value=3.54 QALYs for transplantation and 1.66 for platinum regimens) and the cost-utility ratio was $31,915 per QALY gained.