This thesis is about the economic evaluation of interventions for air pollution control. It is structured around three major components. First, a critical analysis of current methods for health impact quantification in the environmental health literature. Second, the development of a Markov model of the health impacts of long-term exposure to air pollution, using the quality-adjusted life year as health metric, in order to evaluate the cost-effectiveness of reducing air pollution in England. Third, evidence synthesis and COPD incidence estimation by severity stage, in order to parameterise the model developed.
I demonstrate that the current approach to quantifying the health benefits from air pollution reduction leads to substantially biased estimates. By ignoring interactions between morbidity and mortality impacts, including differential susceptibility to risk by health status, it overestimates the change in morbidity cases and underestimates life expectancy effects. I also show that current European guidelines for uncertainty analysis in assessments of air pollution control interventions underestimate decision uncertainty and may misguide air quality strategies.
My Markov model fully captures, for the first time, the lifetime impact of air pollution exposure on individuals' quality and length of life, and identifies the joint health care budget impact of a reduction in chronic morbidity and premature death. Air quality improvement has important health implications. In London, investing up to £500 million to reduce fine particulate concentrations by 1 ug/m3 (i.e. by 7%) is highly likely to be cost-effective, whether the investment is funded by the NHS or through taxation. If this improvement were to cost more than that, however, funding through taxation is more likely to be cost-effective than funding via the NHS, since consumer willingness to pay for a QALY is higher than the estimated NHS expenditure required to deliver one QALY. The optimal level of pollution reduction, as well as the decision about whether and for how long to delay investments, is therefore expected to depend on the source of financing.
