Examining trends in inequality in the use of reproductive health care services in Ghana and Nigeria

Data

We used secondary data from the Demographic and Health Surveys (DHS). The surveys are conducted under an international program implemented by ICF International and funded by the USAID with contributions from UNICEF, UNFPA, WHO, and UNAIDS [40]. The DHS are cross-sectional and nationally representative surveys in low- and middle-income countries. The DHS adopt a multi-stage cluster design and samples selected for enumeration are ensured to be representative and comparative across countries. The DHS involves a two-stage cluster and systematic sampling design with households selected at random. In both countries, the sampling accounted for differences in population distribution regionally as well as for the urban-rural spread. Designated households were enumerated without allowance for a change or replacement to prevent bias. Respondents were selected on the basis of being female (for the female survey) or male (for the male survey), aged 15-49 and whether the respondent was a usual member of the household or having spent the night prior the survey in the household. These surveys employed standard Demographic Health Surveys (DHS) questionnaires and techniques for data collection [40]. All eligible women aged 15–49 were interviewed with the Women’s Questionnaire. Eligible women are all women aged 15–49 who stayed in a selected household the night before the interview, irrespective of whether they were usual residents in the household or not. The Women’s Questionnaire was used to collect respondent’s individual characteristics including age, marital status, occupation, residence as well as other information on topics including; reproductive history; contraceptive knowledge and use; antenatal, delivery and postnatal care; marriage; attitudes about family planning.

Analyses were performed using data from the women’s response file, from the full DHS dataset of Ghana (2003, 2008, and 2014) and Nigeria (2003, 2008, and 2013). We use data from women who have had at least one birth in the 5 years prior to the survey. A summary of indicators used and information on missing data is available in Additional file 1.

Measurement

Equity analysis

To measure household wealth inequality in access to family planning, antenatal care, and delivery care services in Ghana and Nigeria over time, we use the concentration curve and associated index. The concentration curve plots the cumulative percentage of use of various health care services on the vertical axis (y-axis) against the cumulative percentage of women ranked by their household wealth on the horizontal axis (x-axis), beginning with the poorest and ending with the richest households. The equality line runs diagonally across the figure when women, irrespective of economic status have the same access to health care service, that is, all values on the x-axis equals all values on the y-axis [43]. A curve that lies below the equality line indicates that access to the health care service is concentrated among wealthier households. If the curve lies above the line of equality it implies the presence of inequity, that is, use of the health care service is concentrated among poorer households.

where h is the healthcare outcome of interest, μ is the mean of h and r is the fractional rank of an individual in the household wealth distribution. Additional analyses performed test the null hypothesis of equality across groups to measure horizontal inequality, which is the hypothesis that the index is the same within a group. Comparison of the concentration indices within socioeconomic groups, including age, marital status, maternal occupation, location (rural or urban), and region of residence, was done using the homogeneity test as provided by O O'Donnell, S O'Neill, T Van Ourti and B Walsh [45].

We used sampling weights for all statistical analysis. Data analyses were performed using STATA version 15.1.

Concentration curve of household wealth-related inequality in use of reproductive health care services

In Figure 1, the top image indicates that the curves of family planning information via print and TV lie distinctively furthest away from the line of equality over the observed years in Ghana, suggesting that these services are to the disadvantage of women in poor households. The bottom image of Fig. 1 shows that all the curves describing the use of family planning services including the use of modern contraceptive lie below the line of equality observed in Nigeria. Lastly, the curve of unmet needs for family planning lies below the equality line and increased over the periods observed in Nigeria unlike in Ghana where the curve lies closely to the equality line all through (Additional files 2 and 3: Figures S1 and S2).

Figure 2 shows that in Ghana (top image), the concentration curve of antenatal care at private hospitals is below and furthest away from the equality line. Figure 2 (bottom image) suggests that all the examined antenatal care services are distributed to the disadvantage of women in poor households in Nigeria since the curve lies below the equity line. Other concentration curves depicting a woman’s use of antenatal care lie close to the equality line (Additional files 4 and 5: Figures S3 and S4).

Figures 3 (top and bottom images) show a clear picture of the curves for indicators associated with delivery care across the years in Ghana and Nigeria respectively. The curve of home births lies above the line of equality in both countries throughout the periods observed, indicating predominance among poor households (Additional files 6 and 7: Figures S5 and S6). Also, the curves depicting the use of private facility, Caesarean section, and assistance by a doctor during child delivery appears to follow a similar pattern and lie furthest away below the equity line throughout the years observed (Additional files 6 and 7: Figures S5 and S6).

Concentration indices of household wealth-related inequality in the use of reproductive health care services

Family planning services

We find that use of family planning information via TV or print media is unequally distributed in favor of women in wealthy households of Ghana, specifically, those with agricultural livelihood and women who live in Upper East region. This is unsurprising since wealth is correlated with education which facilitates access and assimilation of information [46]. This finding supports previous research that showed that higher socioeconomic status improves the use of family planning media messages [47]. Although studies show that few people get family planning information via media messages [48]. Another study found that access to family planning information via television or print medium is in disfavor of women in lower socioeconomic strata [49]. However, access to family planning information via media messages about reproductive health care services has mixed results in promoting access in Africa [23, 49, 50]. We also find that unwanted births, indicated by unmet needs for family planning, are concentrated in wealthy households of Nigeria and occur most among women who are not working or living in rural residences. This finding suggests that there is a high need for contraception among women in rural economically advantaged households. Studies show that wage-earning or economically self-sufficient women are more likely to seek contraception, though modern means of preventing unwanted births could be inaccessible in cultural and religious societies [48, 51]. Finally, regarding access to family planning services, while economic status does preclude women from making sole reproductive decisions it could, however, initiate a demand for contraception [52–54].

Antenatal care services

We find that, in both Ghana and Nigeria, women in poor households have increasingly become disadvantaged as inequality in use of antenatal care at private facilities increased in favor of their counterparts in wealthy households. A study carried out in Ghana using DHS data from years 1988 – 2008 found a similar increase. However, this study did not disaggregate antenatal care by type of provider [36]. It is reported that wealthier women are better able to overcome barriers of informal payments of cash or kind and are less likely to encounter negative health workers attitudes often seen in private health care facilities [55, 56]. Other studies have shown that wealth-related inequalities in the use of antenatal care have increased in the past years [36, 57]. Results from our study further suggest that the use of health worker’s assistance during pregnancy outside a facility in both countries became less equitable; women in poor households in urban areas or with professional/sales occupation use such assistance more frequently. Professional occupation and urban area residents are generally thought to have better access to good quality reproductive health care services, given their knowledge and the service availability accessible to these women to draw from [58–61]. Our finding deviates from other studies which suggest that women in these groups access better antenatal care services. To explain the variation in the use of health worker’s assistance during pregnancy outside a facility, studies have also shown that transport and health facilities waiting time may facilitate the use of such assistance or deter the use of modern antenatal care services among women in these categories [60, 62–65]. Specifically, the observed change in home antenatal care from prominence among poor to richer households in Nigeria is unexpected. AF Fagbamigbe and ES Idemudia [66] also noted non-use of antenatal care among the wealthier women during pregnancy and suggest that not only poverty but also other factors like personality and view on the quality of services are relevant. It is also plausible that these are a response to increased pressure on resources in government or other maternal care formations [64]. Unfortunately, there is no information on the quality of care in the DHS data.

The finding that antenatal care in government facilities in both countries is pro-poor and consistently changing to the advantage of women in poor households was observed in different years. No effect was observed among women in the agrarian sector, however. One study report that the use of antenatal care improved among women in Ghana, and, though economic challenges are being surmounted, it may be delayed among women in agricultural occupations [36]. Other studies report pro-wealthy inequality changes in antenatal care use among women in Ghana and Nigeria between 2003 and 2008 [19, 57]. In addition, urban-rural inequality in the use of antenatal care at government facilities was observed in the later years in Nigeria but diminished in Ghana. In Nigeria, we find no evidence that rural women in poor households seek antenatal health care services at government facilities. Other studies found unequal use of antenatal care services to the detriment of women in rural households [62, 66]. A study of Nigeria’s midwives service scheme found insignificant success in rural areas attributable to pro-wealthy resource distribution [24]. Nonetheless, the observed diminished urban-rural differential to benefit women in poor households in Ghana has been partially credited to improvements in infrastructure and maternal health care services [36].

Delivery care services

Our study finds that childbirth at home persists among women in poor households although overall inequality magnitude appears to have declined in Ghana while it has increased in Nigeria, there are substantial geographical variations. It appears that by 2014 inequality became notable among women in all seven regions of Ghana. In a 2005 research on the free delivery care policy in Central and Volta regions of Ghana, an increase in facility delivery and a decline in home delivery was reported [26]. Another study found that coverage of the doorstep community-based health planning and services program in Ghana was substantial in mainly the Upper East region [23]. A separate study carried out among Nigerian women in 2004 did not find a substantial increase in institutional delivery facilities in Nigeria despite the midwives service scheme [24]. Evidence of substantial pro-rich inequality between the Northern and Southern regions was observed in Nigeria, while delivery at government health facilities favored women in the Northern regions. Delivery at a private health facility is more inequitable among the richer households in the Southern regions. We find persisting rural-urban disparities associated with childbirth at a government health facility in Nigeria, but not in the later years in Ghana, that is 2008 and 2014. Other research did not find evidence of rural-urban differences in the shift from home to health facilities in Ghana and Nigeria [67]. The observed inequalities among women who have childbirth at home suggest that implemented health policies such as the community-based health planning and services initiative in Ghana, a free delivery scheme in Ghana, the midwives service scheme in Nigeria and insurance schemes in both counties have not substantially reduced inequality in home birth among women. This is also confirmed in other studies [16, 36, 57].

The finding that Caesarean section in both Ghana and Nigeria is pro-wealthy is not surprising. Our findings are in line with the findings of previous studies that Caesarean section remains under-provided for women in poor households in both countries [16]. The trend suggests that coverage gap in both countries remains relatively high. In addition, the equity trend observed in Nigeria suggests top inequity, indicating that the increase in the CI magnitude related to Caesarean section is extremely high.

Study limitation

The cross-sectional design of the study implies that we can show associations without concluding about causal relationships. Other methodological limitations of the study include recall bias since the survey collects events over a five-year period. A limitation regarding the country comparisons concerns the fact that inequalities were investigated based on the position of women in the distribution of household wealth in their own country. We recognize that a woman who is poor by Ghana standards may be better off in Nigeria. Also, we recognize that measures of coverage gap [68], which we have not analyzed, are as equally important as the equity gap evidenced in this paper. Coverage gap refers to the difference between the targeted and actual use of essential health care services by the population, while equity gap indicates the distribution of the services use across the wealth-based population groups [68]. Thus, the underutilization of reproductive care is not directly addressed in our equity study. However, the study has some strengths as well. In particular, the merging of important indicators of maternal care improved the ability to capture the different categories of reproductive care. In addition, we use a generalized concentration index as a measure of inequality, which is not sensitive to outcome measures because it quantifies the absolute differences in health between income groups. Finally, the measurement of the magnitudes of inequalities over different years gives indications about the changing horizontal inequalities which are country and time specific.