Reconciling methodolatry and policy relevance in air pollution epidemiology: reflections from India

Abstract

Indians experience some of the highest exposures to air pollution both within and outside the home, and Indian cities feature prominently on the list of most polluted places on the planet. The most recent sub-national burden of disease estimates note that approximately 1.7 million deaths were associated with household and ambient air pollution exposure in 2019, the highest of any country in the world. However, the evidence base continues to remain relatively scant for a country the size of India. Recent efforts to strengthen the evidence base in India have involved the use of data from government sources. Examples of such sources in India include registry data and survey data from large-scale representative surveys conducted nationally. However, the use of these data in analyses for air pollution epidemiology presents fundamental challenges including: 1.Access and availability – data are either unavailable in a timely fashion to make meaningful inferences, or their public release involve several layers of complex bureaucracy 2.Quality – registry data are often incomplete and even when complete are replete with coding errors especially related to aspects such as cause of death 3.Periodicity – large-scale surveys are meant to be conducted with a pre-defined periodicity, but this has varied with gaps between iterations ranging from 4-10 years These challenges are layered with an additional complication for early-career environmental epidemiologists – going the "quick and dirty” route of using tried and tested methods that may be contextually relevant but dated or aligning your methods to appeal to a field in environmental epidemiology that has made rapid advances in recent years. This talk will focus on these challenges with examples from the author’s own perspective as an early-career researcher, and outline steps needed to strengthen the policy-relevant evidence base locally.