Abstract

Background: Many epidemiological studies require a measure of socioeconomic position. The monetary measure preferred by economists is consumption expenditure; the wealth index has been proposed as a reliable, simple alternative to expenditure and is extensively used.

Methods: A systematic review was conducted of the agreement between wealth indices and consumption expenditure, summarising the agreement and exploring factors affecting agreement.

Results: Seventeen studies using 36 datasets met the inclusion criteria. Of these, 22 demonstrated weak agreement, 10 moderate agreement, and four strong agreement. There was some evidence that agreement is higher: in middle-income settings; in urban areas; for wealth indices with a greater number of indicators; and for wealth indices including a wider range of indicators.

Conclusions: The wealth index is mostly a poor proxy for consumption expenditure.