Table 1.

Overview and definition of variables in the HEAT Plus template.

Variable	Definitions and Notes
Mandatory variables
Setting	Setting name (e.g. a country like ‘Indonesia’, or a province like ‘Bali’)
Year	Year (e.g. ‘2016’)
Source	Data source (e.g. ‘DHS’)
Indicator_abbr	Indicator abbreviation (e.g. ‘anc’)
Indicator_name	Indicator name (e.g. ‘Antenatal care coverage’)
Dimension	Dimension of inequality (e.g. ‘Education’)
Subgroup	Population subgroup (e.g. ‘Primary school’)
Estimate	Subgroup estimate
Favourable_indicator	This dummy variable indicates the indicator type. It must be 1 for favourable indicators and 0 for non-favourable (adverse) indicators. Favourable indicators measure desirable health events that public health action promotes. They include health intervention indicators, such as antenatal care coverage, and desirable health outcome indicators, such as life expectancy. For these indicators, the ultimate goal is to achieve a maximum level, either in health intervention coverage or health outcome (e.g. complete coverage of antenatal care or the highest possible life expectancy). Adverse indicators measure undesirable health events that are to be reduced or eliminated through public health action. They include undesirable health outcome indicators, such as stunting prevalence in children aged less than five years or under-five mortality rate. Here, the ultimate goal is to achieve a minimum level (e.g. theoretically 0 deaths per 1000 live births).
Indicator_scale	This variable indicates the scale of the indicator, such as ‘100’ for indicators reported as percentages or ‘1000’ for indicators reported as rates per 1000 population.
Ordered_dimension	This dummy variable indicates the dimension type. It must be 0 for dimensions with two subgroups (binary dimensions). For dimensions with more than two subgroups, it must be 1 for ordered dimensions and 0 for non-ordered dimensions. Binary dimensions compare the situation in two population subgroups (e.g. males and females). Ordered dimensions have ordered subgroups that have an inherent positioning and can be ranked. For example, education has an inherent ordering in the sense that those with less education unequivocally have less of something compared to those with more education. Non-ordered dimensions have non-ordered subgroups that are not based on criteria that can be logically ranked. Subnational regions are an example of non-ordered groupings.
Subgroup_order	This variable indicates the order of subgroups for ordered dimensions. For ordered dimensions (i.e. if ordered_dimension = 1), this variable must be an increasing sequence of integers starting with the value 1 for the most-disadvantaged subgroup. For example, for education with three subgroups, the most-disadvantaged subgroup ‘No education’ will be assigned the value 1, ‘Primary school’ will be assigned the value 2 and the most-advantaged subgroup ‘Secondary school +’ will be assigned the value 3. For non-ordered dimensions and binary dimensions (i.e. if ordered_dimension = 0), this variable must be 0.
Reference_subgroup	This variable indicates the reference subgroup for non-ordered dimensions and binary dimensions. For ordered dimensions (i.e. if ordered_dimension = 1), this variable must be 0. For non-ordered dimensions and binary dimensions (i.e. if ordered_dimension = 0), a reference subgroup can be chosen. A reference subgroup can be chosen by assigning the value 1 to that subgroup and 0 to all other subgroups. For example, for subnational regions (with more than two subgroups), the capital city can be chosen as the reference subgroup. For place of residence (urban vs rural), urban can be chosen as the reference subgroup.
Recommended variables
se	Standard error of subgroup estimate
Population	The number of people affected or at risk within that subgroup (e.g. weighted sample size by subgroup in household surveys).
Setting_average	Setting average
iso3	ISO3 country code for country-level data (e.g. ‘IDN’ for Indonesia).
Optional variables
95ci_lb	95% confidence interval lower bound of subgroup estimate.
95ci_ub	95% confidence interval upper bound of subgroup estimate.
flag	Flag of subgroup estimate, indicating notes or observations relevant to the analysis. For example, if a subgroup estimate is based on a very small number of cases, this could be indicated in the flag.