1. ¹Department of Biomedicine, Sahlgrenska Academy at Göteborg University, Göteborg SE-405 30, Sweden
2. ²Department of Surgery, Lundberg Laboratory for Cancer Research, Sahlgrenska Academy at Göteborg University, Göteborg SE-413 45, Sweden
3. ³Department of Mathematical Sciences, Chalmers University of Technology, Göteborg SE-412 96, Sweden

Correspondence: Professor A Odén, E-mail: anders.oden@mbox301.swipnet.se

Received 26 February 2007; Revised 19 April 2007; Accepted 23 April 2007; Published online 22 May 2007.

Abstract

A key task for health policymakers is to optimise the outcome of health care interventions. The pricing of a new generation of cancer drugs, in combination with limited health care resources, has highlighted the need for improved methodology to estimate outcomes of different treatment options. Here we introduce new general methodology, which for the first time employs continuous hazard functions for analysis of survival data. Access to continuous hazard functions allows more precise estimations of survival outcomes for different treatment options. We illustrate the methodology by calculating outcomes for adjuvant treatment of gastrointestinal stromal tumours with imatinib mesylate, which selectively inhibits the activity of a cancer-causing enzyme and is a hallmark representative for the new generation of cancer drugs. The calculations reveal that optimal drug pricing can generate all win situations that improve drug availability to patients, make the most of public expenditure on drugs and increase pharmaceutical company gross profits. The use of continuous hazard functions for analysis of survival data may reduce uncertainty in health care resource allocation, and the methodology can be used for drug price negotiations and to investigate health care intervention thresholds. Health policy makers, pharmaceutical industry, reimbursement authorities and insurance companies, as well as clinicians and patient organisations, should find the methodology useful.