OBJECTIVES To demonstrate the impact of QALY weightings based on the burden-of-illness (BoI) of pancreatic cancer on the incremental cost-effectiveness ratio (ICER) of nab-paclitaxel plus gemcitabine (NPG) versus gemcitabine (G). METHODS A markov model using data from the  MPACT trial plus resource use data and costs from NHS Scotland have been submitted to the Scottish Medicines Consortium. The base case ICER was £52,885/QALY based on a cost of £8,232 and a QALY gain of 0.156 (SMC DAD). QALY weightings up to a maximum of 2.5 distributed across six modifier factors, including BoI, have been proposed (NICE consultation on Value Based Assessment), with BoI measured according to  proportional QALY shortfall associated with the condition. The estimated 98% loss of healthy life (proportional QALY shortfall) in pancreatic cancer (Hutchings 2014) represents an almost complete loss of life, and thus a very high BoI. A BoI weighting of 2.5 (maximum weighting allocated entirely to BoI, or BoI FULL) and an alternative BoI weighting of 1.417 (maximum weighting shared equally between six modifiers, so 1/6^thof 2.5,  or BoI PARTIAL) were therefore applied to the QALY gain of NPG versus G.  RESULTS The BoI FULL weighting gives an adjusted QALY gain for NPG versus G of 0.39 and a corresponding ICER of £21,108/QALY. The BoI PARTIAL weighting gives an adjusted QALY gain for NPG versus G of 0.221 and a corresponding ICER of £37,249/QALY.  CONCLUSIONS Various ways of accounting for disease severity can be considered and made workable by HTAs, including QALY weightings according to proportional QALY shortfall. The adjusted QALY gain and corresponding ICERs of NPG versus G in pancreatic cancer show that the value of medicines for life-threatening ‘end-of-life’ conditions with a high relative shortfall can be reflected by an appropriate system of QALY weightings.