№  Слайд  Текст 
1 

Health inequality monitoring: with a special focus on low andmiddleincome countries Lecture 4: Simple measures of health inequality 
2 

How can health inequalities be measuredSimple measures make pairwise comparisons of health between two subgroups, such as the most and least wealthy main type of measurement used in inequality monitoring intuitive and easily understood 
3 

Difference and ratio measuresDifference shows the absolute inequality between two subgroups the mean value of a health indicator in one subgroup subtracted from the mean value of that health indicator in another subgroup Ratios show the relative inequality between two subgroups the mean value of a health indicator in one subgroup divided by the mean value of that health indicator in another subgroup When there are only two subgroups to compare, difference and ratio are the most straightforward ways to measure absolute and relative inequality 
4 

Absolute and relative inequalityAbsolute inequality reflects the magnitude of difference in health between two subgroups Absolute measures retain the same unit of measure as the health indicator For example, if health service coverage were 100% and 90% in two subgroups of one population, and 20% and 10% in subgroups of another population, both cases would report absolute inequality of 10 percentage points Relative inequality measures show proportional differences in health among subgroups For example, the relative inequality in a population with health service coverage of 100% and 50% in two subgroups would equal 2 (100/50 = 2); the relative inequality in a population with health service coverage of 2% and 1% in two subgroups would also equal 2 (2/1 = 2) 
5 

Applied examples: difference and ratioTable 1 Areabased inequality in antenatal care (at least four visits) in Colombia, DHS 1995, 2000, 2005 and 2010 Survey year Coverage in rural area (%) Coverage in urban area (%) Difference (urban – rural) (percentage points) Ratio (urban / rural) 1995 53.8 82.4 28.6 1.5 2000 64.7 84.9 20.2 1.3 2005 73.1 87.1 14.0 1.2 2010 80.5 90.3 9.8 1.1 Table 2 Sexbased inequality in underfive mortality rates in Egypt, DHS 1995, 2000, 2005 and 2008 Survey year Female (deaths per 1000 live births) Male (deaths per 1000 live births) Difference (male – female) (deaths per 1000 live births) Ratio (male / female) 1995 98.9 92.1 –6.8 0.9 2000 69.3 68.6 –0.7 1.0 2005 46.3 52.1 5.8 1.1 2008 27.7 38.4 10.7 1.4 
6 

Applied example: difference and ratioTable 3 Wealthbased inequality in births attended by skilled health personnel in the Philippines, DHS 1998, 2003 and 2008 1998 21.2 45.9 72.8 83.9 91.9 70.7 4.3 2003 25.1 51.4 72.4 84.4 92.3 67.2 3.7 2008 25.7 55.6 75.8 86.0 94.4 68.7 3.7 Survey year Quintile 1 (poorest) (%) Quintile 2 (%) Quintile 3 (%) Quintile 4 (%) Quintile 5 (richest) (%) Difference (quintile 5 – quintile 1) (percentage points) Ratio (quintile 5 / quintile 1) 
7 

Ordered and nonordered groupsOrdered groups have an inherent positioning and can be ranked For example, wealth, education level Nonordered groups, by contrast, are not based on criteria that can be logically ranked For example, region, ethnicity, religion, place of residence 
8 

Two subgroups and more than two subgroupsSome equity stratifiers naturally generate two subgroups For example, sex, urbanrural place of residence Other equity stratifiers may comprise multiple subgroups For example, wealth quintiles, region Many equity stratifiers could be classified either way For example, urbanrural place of residence could be expanded to include large cities, small cities, towns, villages, countryside, etc. Simple measures are appropriate to make pairwise comparisons of two subgroups; complex measures may be useful when there are more than two subgroups 
9 

Simple measures of inequality: multiple subgroupsWhen there are multiple subgroups pairwise comparisons may be made between: Subgroups with highest and lowest values of a health indicator Specific pairs of subgroups, based on a selected reference subgroup or subgroups For example, comparing each region with the capital region For example, comparing each wealth quintile to the richest quintile 
10 

Limitations of simple measures of inequality#1. Pairwise comparisons ignore all other subgroups that are not being compared 
11 

Limitations of simple measures of inequalitySurvey year Quintile 1 (poorest) (%) Quintile 2 (%) Quintile 3 (%) Quintile 4 (%) Quintile 5 (richest) (%) Difference (quintile 5 – quintile 1) (percentage points) 2003 20.6 31.9 43.3 73.0 90.4 69.8 2008 24.2 50.0 64.8 81.7 94.6 70.4 Table 4 Wealthbased inequality in births attended by skilled health personnel in Ghana, DHS 2003 and 2008 
12 

Limitations of simple measures of inequalityFigure 1 Births attended by skilled health personnel in Ghana, by wealth quintile, DHS 2003 and 2008 
13 

Limitations of simple measures of inequality#2. Pairwise comparisons do not take into consideration subgroup size 
14 

Limitations of simple measures of inequalityTable 5 Educationbased inequality in contraceptive prevalence (modern methods) in the Philippines, DHS 1993 and 2008 Figure 2 Contraceptive prevalence (modern methods) in the Philippines, by education level, DHS 1993 and 2008 Survey year None (%) Primary (%) Secondary or higher (%) Difference (secondary or higher – none) (percentage points) 1993 7.2 21.5 28.0 20.8 2008 8.7 30.3 35.8 27.1 
15 

Limitations of simple measures of inequalityFigure 3 Proportion of women of reproductive age in the Philippines, by education level, DHS 1993 and 2008 Source: Data provided by: International Center for Health Equity, Federal University of Pelotas, Brazil. 
16 

Limitations of simple measures of inequalityInterpretation challenges due to population shifts: Example: moreeducated subgroups may appear to be losing coverage of a health service over time, when in reality this could be the result of a population shift of uncovered persons from lesseducated subgroups into moreeducated subgroups Should report the relative size of the population subgroups alongside disaggregated mean values of the health indicator 
17 

Health inequality monitoring: with a special focus on low andmiddleincome countries Full text available online: http://apps.who.int/iris/bitstream/10665/85345/1/9789241548632_eng.pdf 
«Health inequality monitoring: with a special focus on low and middleincome countries» 