1041P - Key drivers of cost-effectiveness of anti-cancer drugs

Date 28 September 2014
Event ESMO 2014
Session Poster Display session
Topics Bioethics, Legal, and Economic Issues
Presenter Martin Hoyle
Citation Annals of Oncology (2014) 25 (suppl_4): iv357-iv360. 10.1093/annonc/mdu341
Authors M. Hoyle
  • Medical School, Veysey Building, University of Exeter, EX2 4SG - Exeter/GB

Abstract

Aim

Health economic models are often complex, with numerous parameters. Here, the intention is to give some insight in to the sensitivity of cost-effectiveness to various parameters that routinely appear in economic models.

Methods

Economics models written and reviewed at the author's institution in the UK from 2007 to 2014 were studied. These included one model for metastatic colorectal cancer, one for renal cell carcinoma, four for chronic lymphocytic leukaemia, three for chronic myeloid leukaemia, one for gastro-intestinal stromal tumours, and one for multiple myeloma. In all these analyses, the impact of various model parameters on the incremental cost-effectiveness ratio was inspected.

Results

Cost-effectiveness is generally very sensitive to: Mean overall survival, Adjustment for treatment switching where appropriate, Mean treatment duration, Drug cost per month, Utilities (quality of life) on and off treatment. Cost-effectiveness tends to be sensitive to: Cost of drug administration, Drug dose intensity, Medical management costs, Cost of subsequent treatments, Discount rate for chronic diseases, Vial wastage, Mean progression free survival. Cost-effectiveness is generally insensitive to: Costs of adverse events, Disutilities of adverse events, Cost of blood tests, Discount rates for acute diseases, Costs of death.

Conclusions

It is possible to identify parameters that usually strongly influence the cost-effectiveness of anti-cancer drugs.Where possible, it is recommended that information should be provided in published trials to inform these key parameters.

Disclosure

All authors have declared no conflicts of interest.