OBJECTIVES

: Systemic anti-cancer therapies (SACTs) are commonly dosed according to patient body surface area (BSA) or weight. Recent National Health Service England (NHSE) guidance encourages hospitals to use dose banding, wherein individualised doses are rounded up or down to predetermined standard doses. Dose banding is expected to lead to reduced wastage and administration time. This study aimed to estimate per-patient annual drug and administration costs for Stage IV melanoma treatments for which dose-banding guidance is available, and compare these costs to when dose banding is not considered.

METHODS

: A log-normal distribution was fitted to Health Survey for England weight and BSA data to calculate the proportion of patients requiring each number of vials. Dose-banding guidance is available for dacarbazine, ipilimumab and pembrolizumab. Three costing methods were explored and compared: (1) non-banded dosing assuming individual preparation, (2) banded dosing assuming individual preparation, and (3) exact dose (i.e. no drug wastage due to preparation for the population). List prices were used in all analyses. Administration efficiency savings were incorporated, based on published NHSE evidence.

RESULTS

: For ipilimumab and pembrolizumab, dose banding resulted in annual respective drug costs savings of 6.4% (-£4,581) and 9.1% (-£3,143) versus no banding. For dacarbazine, dose banding led to a 0.3% increase in annual drug costs (+£0.89). Assuming no wastage led to savings of 3.9%-15.6% across the regimens considered. Estimated annual cost savings due to reduced administration time ranged from £329 to £528, dependent on administration frequency.

CONCLUSIONS

: While assuming no wastage may be unrealistic within routine NHSE practice, the results of this analysis demonstrate the potential cost savings possible with dose banding, due to similar or lower drug costs and reduced administration time. Economic evaluations incorporating the costs of SACTs should include scenarios with dose banding where relevant.