Home

Increasing the coverage of key maternal, newborn, and child health interventions is essential if Millennium Development Goals (MDG) 4 and 5 are to be reached. We have assessed equity and trends in coverage rates of a key set of interventions through a summary index, to provide overall insight into past performance and progress perspectives.

Data from household surveys from 54 countries in the Countdown to 2015 for Maternal, Newborn and Child Survival initiative during 1990-2006 were used to compute an aggregate coverage index based on four intervention areas: family planning, maternal and newborn care, immunisation, and treatment of sick children.

The four areas were given equal weight in the computation of the index. Standard measures were applied to assess current levels and trends in the coverage gap measure by wealth quintile. The overall size of the coverage gap ranged from less than 20% in Tajikistan and Peru to over 70% in Ethiopia and Chad, with a mean of 43% for the most recent surveys in the 54 countries.

Large intracountry differences were noted, with a country mean coverage gap of 54% for the poorest quintiles of the population and 29% for the wealthiest Differences between the poorest and the wealthiest were largest for the maternal and newborn health intervention area and smallest for immunisation. In 40 countries with more than one survey, the coverage gap had decreased by an average of 0.9 percentage points per year since the early 1990s.

Declines greater than 2 percentage points per year were seen in only three countries after 1995: Cambodia, Mozambique, and Nepal. Country inequity patterns were remarkably persistant over time, with only gradual changes from top inequity (disproportionately smaller gap for the wealthiest) in countries with coverage gaps exceeding 40%, to linear patterns and bottom inequity (disproportionately greater gap for the poorest) in surveys with gaps below 40%.

Despite most Countdown countries having made gradual progress since 1990, coverage gaps for key interventions remain wide and, in most such countries, the pace of decline needs to be more than doubled to reach levels of coverage of these and other interventions needed in the context of MDG 4 and 5. In general, in-country patterns of inequality are consistant and change only gradually if at all, which has implications for the targeting of interventions.

In the absence of targeting, health interventions tend to be adopted initially by the wealthiest, and later trickle down to the rest of the population who often emulate the behaviour of the elite groups.18 The top inequity pattern seen in countries with the largest gaps, therefore, might represent an unavoidable phase in the scaling-up process. The challenge in these countries is how to reduce the gap rapidly for all strata of the population, rather than improve equity by reducing uptake by the wealthiest people.19

The widespread availability of data for soaoeconomic inequities in maternal and child health is a recent occurrence. Policy makers and health managers in low-income and middle-income countries need to become aware of the magnitude of inequities in their countries, of which services or interventions are least equitable, of how inequities are evolving over time, and of what population subgroups are being most affected. Mainstreaming equity considerations into health policies and programmes can help to speed up achievement of national and international health goals.

The strengths of the coverage gap index is that it is easy to compute, it is robust (as proven by data from over 100 surveys), and it can be adapted at country level by inclusion of different interventions, including adult health interventions such as antiretioviral treatment coverage or mammography. With a broader set of interventions, the coverage gap index becomes increasingly suitable as a measure of health system strength. It can be used for equity and subnational analyses as long as survey sample sizes are adequate.

In countries with several surveys the coverage gap patterns by wealth quintile were remarkably persistent over time. In some MICS, patterns were not consistent with previous or subsequent DHS surveys, or showed ; irregular patterns by wealth quintile. This irregularity might be due to a poorer quality of data in some MICS.