Almost all health systems seek to offer some form of publicly financed healthcare insurance, and governments must therefore choose the size of the benefit package and the types of treatments to cover. Conventionally, the usual approach of economists has been to recommend choices on the basis of cost effectiveness of treatments, using metrics such as the ‘cost per quality adjusted life year’. However, this approach is based on the assumption of health maximization subject to a budget constraint and ignores the potential impact of any additional concern with protecting individuals from the financial consequences of a health shock. Furthermore, it does not take account of the possible availability of complementary privately funded health care. This paper develops a model in which risk-averse individuals care about health but also place a value on protection from the financial consequences of rare but costly events. The paper shows how conventional cost-effectiveness analysis can readily be augmented to take account of financial protection objectives. The results depend on whether or not there exists a market in complementary privately funded health care. They have important implications for the methodology adopted by health technology assessment agencies and for the broader design of publicly funded health systems. Copyright © 2012 John Wiley & Sons, Ltd.