Statistical codes for health inequality analysis

Household health surveys are the main data source for health inequality monitoring because they include both data on health indicators and inequality dimensions. Yet, there are sampling design complexities (e.g. clustering, weighting, stratification) that must be taken into consideration when analysing survey data.

The statistical codes shared here demonstrate how complex survey sampling design may be taken into account in the calculation of a) estimates for health indicators disaggregated by inequality dimensions (e.g. economic status, education and urban-rural areas) and b) population subgroup sizes for each inequality dimension. These two pieces of information (a and b), may be used to calculate summary measures of inequality.

Statistical codes are provided for commonly used selected statistical packages using a sample dataset.

Notes:

The health indicator and inequality dimensions are illustrative. Further, the aim here is not to demonstrate how an indicator may be defined or measured, or to advocate for any specific inequality dimension. Rather, the codes demonstrate how for a pre-defined and measured indicator (i.e. births attended by skilled personnel), with pre-selected and measured inequality dimensions, the calculation may be undertaken.

For more information about disaggregated data, see the following eLearning courses: 

Summary measures of health inequality summarize the magnitude of inequality in a given health indicator using a single number - which facilitates the comparison of inequalities over time and across different settings and indicators. Many summary measures exist, each with different applicability depending on the characteristics of the underlying data and the dimensions of inequality that are the focus of analysis. Statistical codes for the calculation of 21 summary measures of inequality are provided.

Simple summary measures – difference and ratio – make pairwise comparisons between two population subgroups. Complex summary measures of inequality take into account all population subgroups to assess inequality. They may also account for the population size of each subgroup and make use of a reference group. Considerations for the selection of summary measures include: 

• Whether the dimension of inequality is ordered (i.e., has an inherent ordering, such as economic status) or non-ordered (i.e., subgroups cannot logically be ranked, such as subnational region).
• Whether the dimension of inequality has two or more than two subgroups. 
• The optimum level that is to be achieved for the health indicator. 
• Whether inequality is measured in absolute or relative terms. 
• Whether subgroups should be weighted according to their population size or not. 
• The selection of a reference point for non-ordered dimensions of inequality. 

For more information about summary measures of health inequality see: