Sources of Data
The study used Ghana’s Malaria Indicator Surveys (GMIS) of 2016 and 2019, whose main objective was to estimate key malaria indicators [23,24]. Both surveys had main objectives of; measuring extent of ownership and mosquito bed nets, coverage of intermittent preventive treatment to protect pregnant women, identifying practices and specific medications used for treating malaria among children under age 5, measuring of indicators of behaviour change communication messages, knowledge regarding malaria, and practices and assessing the measure the prevalence of malaria and severe anaemia among children age 6-59 months.
The samples for both GMIS’s 2016 and 2019 were designed to provide estimates of key malaria indicators for the country as a whole, for urban and rural areas separately, and for each of the 10 administrative regions (Western, Central, Greater Accra, Volta, Eastern, Ashanti, Brong Ahafo, Northern, Upper East, and Upper West) as defined in the Ghana 2010 Population and Housing Census (PHC) [23,24]. Both surveys used the 2010 Population and Housing Census, however in 2019, Ghana created six new regions, resulting in a total of 16 regions and 260 administrative districts but the new administrative boundaries were not available during the survey hence were not included. Therefore, both surveys were based on 10 regional boundaries defined according to the 2010 Population and Housing Census [23,24].
Socioeconomic status was assessed using the wealth index reported in both surveys. The wealth was based on household scores based on the number and kinds of consumer goods they own, ranging from a television to a bicycle or car, and housing characteristics such as source of drinking water, toilet facilities, and flooring materials [23,24]. The scores were derived using principal component analysis however, national wealth quintiles were compiled by assigning the household score to each usual (de jure) household member, ranking each person in the household population by their score, and then dividing the distribution into five equal categories (poorest, poorer, middle, richer, richest), each with 20% of the population. However, for this study the wealth index was categorised in to three categories (poor, middle and rich).
Wagstaff Normalised Concentration index
Household mosquito net ownership, mosquito net use among under 5’s and women, IPTp coverage, health insurance and awareness of malaria covered by national health insurance were binary variable, therefore the study adopted the Wagstaff normalized index (W(h)). The study opted to use the normalised formulae as, Wagstaff, (2005) argued that normalization of the health concentration index formula ensures remedying the bounds issue for a binary cardinal health variable. The Wagstaff normalized index (W(h)) can be expressed as:
C(h) is denoted by;
γii denotes socioeconomic position, with the best well-off individual ranked first and the least well-off individual ranked last
hi denotes real number which measures the health status
µn denotes the average health of the population
an, bn denotes defined lower and upper limits
n denotes a given individual in a population N
The Wagstaff health concentration index (W(h)) measures socioeconomic inequalities in health based upon information on the socioeconomic ranks and the health levels of all individuals in the population . The W(h) values range from -1 and +1 and tackle the bounds issue by stretching the index in such a way that it always has a uniform range. A positive value of W(h) indicates that health is distributed in favour of the rich (+1), and a negative one reflects it is distributed in favour of the poor (-1). The higher the absolute value of the index, the more extreme the pro-rich or pro-poor character of the distribution is supposed to be. Concentration curves were computed to present a graphical picture of the concentration indices. Concentration curves are derivatives of concentration indices and they rank the variable of interest by socioeconomic status .