This paper attempts to assess trends in inequities in selected health status and health services utilisation indicators in Malawi by using quintile ratios and concentration curves and indices. The analysis is based on data from the Demographic and Health Surveys of 1992, 2000 and 2004. This time period allows for analyzing trends in inequities of health indicators that often change gradually and over a longer period of time.
By and large, the findings indicate that in most of the selected indicators of health and healthcare, increases in pro-rich inequities have occurred. This is an undesirable trend in light of the government's explicit commitment to equity in health and healthcare and policy stances. Interventions intended to lessen inequities disfavouring the poor have not borne the expected results.
The quintile ratios for infant and under-five mortality rates indicate progressive inequities between the two extreme quintiles, i.e. wealth quintiles 1 and 5 during the period considered. This is also corroborated by the concentration curves in Figure 2, where the respective concentration curves for the year 2004 have moved further away from the line of equality. Thus, there was no improvement in inequities in these indicators and the improvement in the population averages was primarily due to marked improvements in the rates for the relatively wealthy segments of the population.
Although child mortality rates are influenced by a host of factors, many of which lie outside the health sector, they are often regarded as a proxy for overall disease conditions . Infant and under-five mortality rates are closely related to economic growth and distribution of economic and social resources. Studies have shown that countries whose IMR rates are relatively lower enjoy better economic growth rates than those otherwise . This significant correlation between child mortality rates and economic growth implies that, addressing inequities in infant and under-five mortality should be multi-sectoral and that beyond the biomedical solutions, there is a need to also address the underlying social determinants through concerted and complementary efforts of all sectors of the economy. This is also in line with the principles of the Primary Health Care strategy.
The main direct causes of mortality in under-five children are infectious diseases occurring because they were neither prevented (e.g. vaccine-preventable diseases) nor successfully treated (e.g. ARIs, diarrhoeal diseases) . Diarrhoea, ARIs, measles, malaria and malnutrition account for at least 70% of childhood diseases . The underlying causes are related to socio-economic factors. Thus, from the health sector's perspective, the immediate response to reducing infant and under-five mortality is improving access of the poor to preventive, curative and rehabilitative interventions that are geared towards addressing the major direct causes of childhood mortality. Improving coverage of the interventions through the Integrated Management of Childhood Illness (IMCI) programme may go a long way to bridge the inequity gaps, as 70% of the direct causes are related to the diseases and conditions covered in the IMCI strategy. In addition to improving access to health facilities, improving coverage of IMCI interventions also necessitates outreach services and an increase in community level activities . Widening inequities may imply that the poor's access to the appropriate preventive, curative and rehabilitative interventions has not improved or has even declined.
With respect to child malnutrition (stunting and underweight), there has been an increase in inequities between 1992 and 2004. After a significant increase in inequities in 2000 from the 1992 levels, there was a marginal but statistically insignificant decline in 2004. Thus, no change was observed in the inequity levels in child malnutrition between 2000 and 2004.
According to the WHO cutoffs used to identify nutrition problems of public health significance, the population averages of both stunting and underweight in Malawi fall under the categories of severe stunting (cutoff ≥ 40%) and moderate underweight (cutoff 20–29%). Although the rate of stunting is high even in the non-poor wealth quintile (Quintile 5), there is a marked difference in comparison to that of the poorest quintile (Quintile 1). Stunting, which is an indicator of chronic malnutrition poses adverse long-term consequences on economic productivity. Hence, strategies aimed at reducing poverty and income inequalities need to also tackle the problem of stunting in the overall population and in particular among the poorest of society.
Inequities in total fertility rate (TFR) have been increasing progressively over the given period of time despite a marginal decrease in the population average. The average TFR for Malawi is one of the highest in countries of the Southern African Development Community. Widening inequities suggest that the marginal decline in TFR observed is due to a decrease in TFR among the non-poor. This implies that health sector-specific interventions to curb high fertility rates (e.g. uptake of contraceptives) are not benefiting the poor due to a number of reasons including problems of access and cultural barriers. High TFR has far-reaching effects in that it adversely affects child survival and household welfare particularly among the poor. It is therefore necessary that policies aimed at improving household welfare need to boost coverage of the poor with the available effective interventions. Furthermore, barriers to accessing those interventions need to be identified and addressed appropriately.
A remarkable achievement has been scored in low BMI (body mass index) of mothers, an indicator of maternal undernutrition. Pro-rich inequity that was observed during the earlier years (i.e. 1992 and 2000) was reversed in 2004. Hence there are no inequities in this indicator; maternal undernutrition does not vary systematically with socio-economic status. The DHS data also indicate that overweight and obesity are less of a problem among women from poor households .
The BMI, which is an indicator of chronic energy deficiency among adults, is less of a biomedical problem than it is socio-economic. It is influenced by a host of factors including household socio-economic status, household feeding patterns and seasonal factors . It can therefore be discerned that improvement in those influencing factors among the poor was registered over the years, thus bridging the inequity gap. Reduction in the rate of low BMI in women is beneficial, as low pre-pregnancy BMI is an established risk factor for low birth weight , which in turn affects child survival negatively. It is therefore essential to identify the measures that effectively resulted in abolishing pro-rich inequities so as to replicate them in other related areas and avert any future relapses of inequity in BMI.
Inequities in the prevalence of diarrhoea and ARI among under-five children have also increased over the years significantly. These two conditions are among the major killers of children in sub-Saharan Africa and amenable to low-cost preventive and curative interventions. The fact that pro-rich inequities have widened may imply that environmental conditions (including biological, physical and social environments) that are necessary for the propagation of these diseases among the poor have been deteriorating. Many of the enabling factors for diarrhoeal diseases and ARIs are related to household and community-level socio-economic conditions. Therefore, preventing the disproportionately higher burden of diarrhoea among the poor needs a multi-sectoral strategy beyond the bounds of the health sector (e.g. provision of safe water supply; sanitation, decent housing etc).
The population average for immunization coverage in 2004 has declined by about 17 percentage points from the levels in 1992. Besides, the inequities in immunization coverage seem to have widened over the years implying that the immunization coverage among the poor has continuously declined. It is a well established fact that effective and equitable health systems are a pre-requisite for achieving the MDGs and other health goals . Therefore, the current trend is likely to slow down or even reverse the achievement of the Millennium Development Goal aimed at reducing child mortality.
With respect to Diarrhoea and ARI interventions it has to be noted that an equitable condition demands that those with a higher burden of illness receive more of the treatment according to their need. Hence, the concentration curves should lie above the diagonal (line of equality). Equal use is not equitable in this case. As discussed earlier, diarrhoeal diseases and ARIs are among the major causes of morbidity and mortality among under-five children. It is therefore, necessary to identify the barriers to the utilization of these interventions by the poor so that the poor make use of these interventions more than the non-poor who have less need for it. The current situation of inequity may potentially affect progress towards the aforementioned MDG.
Although there is no inequity in antenatal care, delivery by medically trained personnel favours the non-poor. Moreover, delivery in public facilities is inequitable and to the advantage of the non-poor. This implies that the poor get less of the benefits of publicly financed/subsidized services, contrary to the government's policy objectives. Not unexpectedly, child delivery at home has a pro-poor orientation, which implies that the poor deliver at home proportionately more than the non-poor. The fact that government services are utilized more by the non-poor implies that the poor have a constrained access to child delivery services. This may be related to physical distance, low perceived quality or cultural barriers to name but a few. The definitive contributing factors should be identified by means of further studies. By and large, this trend is likely to jeopardize the pace of reducing maternal mortality and thereby achieving the MDG 5 target, that is reducing maternal mortality.
The inverse equity hypothesis proposed by victora et al  states that new interventions will initially benefit those of higher socio-economic status and only later do they reach the poor. This results in initial increase in inequity ratios for coverage, morbidity and mortality . Policy makers should, therefore, take this phenomenon into account and counteract the widening of inequities through appropriate service delivery strategies. Increasing coverage in poor communities through targeting of those interventions that mainly benefit the poor as well as universal coverage of interventions that address conditions that significantly affect the poor is needed .
Overall pro-rich inequities in health and healthcare are widespread in Malawi and in some cases are widening despite the concerted efforts of government and its development partners. Improvements in population averages of the indicators should not be taken at face value, as the widening disparities imply that the MDG targets may be achieved by the non-poor, but the poor segments of society might not be able to reach them. The fact that the non-poor benefit more from the publicly provided services, which are highly subsidized, is also a point of concern that calls for effective means of targeting the scarce resources. Initiatives such as the sector-wide approach (SWAp)  and the design of essential healthcare package are not inherently equitable if not complemented with policies and strategies that uphold the principles of equity. It is therefore, important to assess interventions/initiatives not only in terms of their efficiency, but also their impact on equity through an appropriate equity gauge .