Background: The prevalence of iron deficiency anaemia remains high in pregnant women and the situation may be worse for pregnant adolescents. This study aimed to comparatively analyse the trends and determinants of anaemia between adult and teenage pregnant women in rural Ghana. Methods: A retrospective study design was employed. Data including primarily pregnancy history, haemoglobin levels and anaemia status were collected from the manual medical records of 1002 women stored in the repositories of two rural district hospitals in the Ashanti Region of Ghana over the years 2011–2015. Data was analysed using chi-square analysis, t-test, two-way ANOVA and binary logistic regression. Bivariate and multivariate analyses were also done to establish associations and predictors of anaemia. Results: An overall drop in mean haemoglobin from 11.1 g/dl in 2011 to 10.6 g/dl in 2015 was observed for adults, while an overall increase from 9.4 g/dl to 10.2 g/dl occurred in teenagers. Further, anaemia prevalence declined at the 36th week of pregnancy, and from 2011 to 2015, for both teenagers and adults. Among factors tested for association with haemoglobin levels using bivariate and multivariate analyses, gestational age alone was significant (P = 0.028). Between-subject effects determined using 2-way ANOVA indicated year of pregnancy alone, as well as in combination with age group (F = 3.1, P = 0.019) significantly affected haemoglobin levels. From binary regression analysis, BMI (OR 0.967, 95% CI:0.936–0.999, P = 0.042), gestational age (OR 1.058, 95% CI:1.013–1.106, P = 0.011) and pregnancy year (OR-2012(0.402); 2013(0.53); 2014(0.569); 2015(0.817), 95% CI: 2012(0.256–0.631); 2013(0.338–0.829); 2014(0.366–0.886); 2015(0.501–1.333), P = 0.001) were found to be significant predictors of anaemia at first antenatal visit. Also, the trimester of reporting for antenatal care, specifically the second trimester (OR-0.261, 95% CI: 0.072–0.951, P = 0.042) and the pregnancy year (OR-2012(0.235); 2013(0.206); 2014(0.530); 2015(0.222), 95% CI: 2012(0.101–0.545); 2013(0.081–0.522); 2014(0.197–1.428); 2015(0.049–1.018), P = 0.003) were more significant at the 36th pregnancy week. Conclusions: Although the trends observed were decreasing in both adults and teenagers in the years reported, anaemia levels remained high for each year in either group. Anaemia, therefore remains a major health problem, especially in the areas studied, and antenatal interventions need a second look to know what might make them more effective.
Keywords: Pregnancy; Anaemia; Iron deficiency; Supplementation
Anaemia is a condition which results from an inadequate number of erythrocytes and consequently a resultant decreased ability to carry oxygen to meet the physiological demands of the body [[
Iron deficiency anaemia is known to be the most common form of anaemia. Statistics indicate that globally, more individuals live with iron deficiency anaemia than any other medical condition [[
Iron and folic acid supplementation (60 mg of elemental iron and 400 μg folic acid for all pregnant women) was put in place by the WHO in 2001 to curtail anaemia in pregnant women [[
Moreover, a study by Glover-Amenyo et al., [[
This study thus sought to analyse the trends and associated factors of anaemia between adult and teenage pregnant women in two rural districts, Ahafo Ano South and Asante Akim South, in the Ashanti region of Ghana. These districts were interesting for such studies as the dietary patterns of the population, generally consisting of a high consumption of starchy foods- such as plantain, yam, maize, were representative of that of most rural parts of Ghana. Moreover, with these characteristics they served as the ideal sites for studying iron deficiency anaemia, which is largely linked to nutritional intake.
Seven (
The Ashanti region is the third largest region in Ghana after the Northern and the Brong Ahafo regions by land size. Asante Akim South District, with its capital, Juaso, is one of the administrative districts of Ashanti, with a population of about 150,000 people (according to the 2000 population census) (MOFA, 2018). The district is strategically located at the entry and exit point to the Ashanti Region and is mainly rural and agricultural in nature with over 60% of the farmers being tenant farmers. Ahafo Ano South district, also in the Ashanti region, has a population of 133,632 representing about 3.7% of the region's total population of 3,612,950 (MOFA, 2018). The district is particularly noted for a high growth rate due to high fertility rate, as well as its agricultural and mining activities.
The study was a retrospective study conducted between October, 2016 and May, 2017.
The sampling frame covered the entire available antenatal attendance records of pregnant women who visited the selected health centres (above described) between 2011 and 2015. The sample consisted of those with complete records in the ANC units and record keeping sections of the health centres at the time of the study. Missing or incomplete records were excluded from the study.
Data collection was done in two forms. Firstly, all complied data were collected from the yearly ANC summaries of the electronic medical records of the health centres in each of the districts for the review years, 2011 to 2015. The data retrieved included total anaemia cases per year and total Hb level checked per year.
In the second phase of data collection, cluster random sampling was used. Extrapolating from the available number of complete records, individual data of a total of 1000 pregnant women (both teenagers and adults), specifically 501 per district were randomly collected from ANC records of the antenatal care units of the main district health centers, i.e. Juaso District Hospital (Asante Akim South) and Mankraso Government Hospital (Ahafo Ano South), for each of the review years. Randomly collected data was later segmented according to age during analysis. The main district centres were the focus in the second phase of data collection due to the inability to obtain the manual records of the sub-district health centres. This involved the use of a comprehensive data collection form designed to capture the Hb levels of subjects, their ages, their blood pressures, their HIV statuses, sickling statuses, weight, height, parity, gestational period, and other special conditions observed. Data was retrieved for years, 2011 to 2015.
Microsoft Excel was used in the random sampling of the patient registration numbers during the second level of data collection. GraphPad Prism vs 8, was used in chi-square analysis, t-test analysis and calculation of p-values. Comparison of mean Hb between teenagers and adults and within each group at different years and stages of pregnancy were conducted using t-test, assuming normal distribution. Normal distribution was tested using normality and lognormality tests (Shapiro- Wilk test) with GraphPad Prism vs 8. Using SPSS vs, 20, bivariate and multivariate analyses, two-way ANOVA and binary logistic regression were carried out to determine the factors associated with Hb levels in the participants, effects of multiple factors on Hb levels and the predictors of anaemia at antenatal registration and at 36th week of pregnancy respectively. Anaemia was classified based on Hb levels. Hb levels below 11 g/dl were classified as anaemic (WHO, 2004).
From the chi-square analysis at 95% confidence interval (Table 1), the mean age for the teenagers was 17.57 years and that of the adults was 28.5 years. The adults had significantly higher gravida (p < 0.001), mean Hb at registration (p = 0.028), body weight (p < 0.001) and BMI (p < 0.001) than the teenagers. Mean Hb at the 36th week of pregnancy, systolic blood pressure and gestational age between the teenage and adult pregnant women were however not significantly different.
Comparison of mean demographics, Hb and pregnancy history between teenage and adult groups
N Mean Std. Deviation F Sig. Age (years) Teenager 119 17.57 1.37 381.07 < 0.001 Adult 879 28.53 6.10 Total 998 27.22 6.75 Height (cm) Teenager 112 156.30 15.23 2.71 0.1 Adult 842 158.06 9.88 Total 954 157.85 10.65 Gravida Teenager 101 0.37 0.75 105.57 < 0.001 Adult 843 2.26 1.84 Total 944 2.06 1.85 Gestational Age (weeks) Teenager 121 17.29 7.88 0.08 0.784 Adult 879 17.05 9.12 Total 1000 17.08 8.97 Reporting Trimester Teenager 122 1.85 0.64 0.02 0.885 Adult 880 1.84 0.76 Total 1002 1.84 0.74 Hb at registration (g/dl) Teenager 121 9.79 1.97 4.85 0.028 Adult 874 10.73 4.62 Total 995 10.61 4.40 Hb at 36th week (g/dl) Teenager 25 10.53 2.50 0.02 0.893 Adult 213 10.58 1.65 Total 238 10.58 1.75 Weight Teenager 122 55.30 8.78 34.33 < 0.001 Adult 874 61.32 10.86 Total 996 60.59 10.81 Systolic Blood Pressure (mmHg) Teenager 118 104.51 16.44 1.05 0.306 Adult 829 106.01 14.70 Total 947 105.82 14.92 BMI (kg/m2) Teenager 112 22.43 3.60 21.22 < 0.001 Adult 836 24.45 4.45 Total 948 24.21 4.40
The mean ages of the teenage and the adult groups were 17.57 and 28.53 years respectively. Based on chi-square analysis of individual data of a total of approximately 1000 individuals, significant differences were observed with respect to gravida, Hb levels at registration, weight and BMI between teenage and adult groups. The ratio of teenagers to adults per 1000 records collected was about 1:9. Hb- Haemoglobin; BMI- Body mass index
Figures 1, 2 and 3 illustrate respectively, 5-year mean Hb at registration, 5-year mean Hb at 36th week and 5-year mean BMI trends between the pregnant teenagers and adults. The results (Fig. 1) showed a fluctuation over the 5 years but an overall drop of mean Hb from 11.1 g/dl in 2011 to 10.6 g/dl in 2015 for the adult population while for the teenagers, there an overall increase from 9.4 g/dl to 10.2 g/dl in the same period. For the results of the five-year trend of mean Hb level at the 36th week of pregnancy (Fig. 2), the trend for the adult pregnant women showed very little variation (stayed around 10 g/dl) in comparison to the trend for teenagers, over the same period. Specifically, for teenagers, the mean Hb level peaked at 2015 (13.5 g/dl) and 2012 (12 g/dl) and had its least value in 2011 (8.9 g/dl). For BMI, most values fell within the normal range, that is, 18.5 to 24.9 kg/m
Graph: Fig. 1 Five- year trend of mean Hb at antenatal registration for teenage and adult pregnant women. A decline from 11.1 to 10.6 is seen from 2011 to 2015 for the adult group while an increase from 9.4 to 10.2 is seen for the teenage group. Hb- haemoglobin; blue- teenage; red- adult. P value< 0.05 (significant)
Graph: Fig. 2 Five- year trend of mean 36th week Hb for teenage and adult pregnant women. The mean Hb of the adult group hovered around 10 g/dl -11 g/dl whle for teenagers, a maximum of 13.2 g/dl was recorded in 2015, after a series of rise and decline from 8.9 g/dl in 2011. Hb- haemoglobin; blue- teenagers; red- adults. P value> 0.05 (not significant)
Graph: Fig. 3 Five- year trend of mean BMI for teenage and adult pregnant women. The mean BMI for the teenage group was lower than that of the adult group across the years (21.7–23.1 vs 23.7–25.1 kg/m2). However, in both cases, BMI generally fell within the normal range (18.5–24.9 kg/m2). BMI- body mass index; blue- teenage; red- adult. P value< 0.05 (significant)
For the five-year anaemia trend for both pregnant teenagers and adults presented in Fig. 4, anaemia prevalence was generally higher in the teenage group than the adults throughout the five-year period, although anaemia prevalence dropped from about 80% to about 70% in the teenagers and from 70 to 60% in the adult group. Additionally, in the adult group, the least anaemia prevalence was recorded in 2012 (50%).
Graph: Fig. 4 Five-year trend of anaemia prevalence for teenage and adult groups. For both teenage and adult groups, anaemia prevalence decreased from 2011 to 2015 (79.20–65.40 vs 72.20–58.60%). However, in the adult group, a slight increase was seen from 2012 to 2013 but this was followed by a gradual decline till 2015. blue- teenage; red- adult. P value< 0.05 (significant)
Table 2 shows the comparison of mean Hb at antenatal registration and at the 36th week of gestation for the combined group (teenage+ adult) and for each group (teenage or adult). Using t-test analysis at 95% confidence interval, results indicated no significant difference in the mean Hb levels at ANC registration and by 36 weeks of pregnancy in each case (that is teenagers only, adults only and the combined group). Further analysis in the teenage and adult groups showed a significant difference in the mean Hb levels of the two groups at ANC registration (p = 0.028); but not at 36 weeks (p = 0.893). The correlation coefficients obtained showed a weak but positive correlation between Hb at antenatal registration and at 36 weeks for the teenage group (R = 0.423, p = 0.039), adult group (R = 0.179, p = 0.001) and both groups (R = 0.2, p = 0.001).
Hb levels per age group at ANC registration and at the 36th week of gestation
Groups Mean N Std. Deviation Mean difference within group P value Within group correlation coefficient P value Combined group (Teenage+ Adult) Hb at registration 10.20 231 1.8 0.26 0.070 0.22 0.001 Hb at 36th week 10.50 231 1.7 Teenage Hb at registration* 10.03 24 2.0 0.46 0.376 0.42 0.039 Hb at 36th week^ 10.50 24 2.5 Adult Hb at registration* 10.30 207 1.8 0.23 0.123 0.18 0.001 Hb at 36th week^ 10.55 207 1.6
*p-value- analysis of the mean difference between teenage only and adult only groups for Hb levels at antenatal registration = 0.028 ^p-value- analysis of the mean difference between teenage only and adult only groups for Hb levels at the 36th week of gestation = 0.893 No mean difference was found between teenagers and adults at ANC registration and at the 36th week of gestation, when combined and analysed separately. A significant difference was found between the Hb levels of teenagers and adults at ANC registration. Further analysis of the correlation between Hb levels at ANC registration and the 36th week of gestation showed a weak but positive correlation in both teenagers and adults, when analysed together and separately. Hb- Haemoglobin; ANC- Antenatal care
Further, bivariate and multivariate analyses were conducted to analyse the effect of various variables (factors) on Hb. The results of a multivariate analysis, conducted at 95% confidence interval (Table 3) showed that amongst various factors that were tested (BMI, District, Gestational age, age group, year, sickling status, age group and year, age group and sickling status, year and sickling status, age group and year and sickling status), gestational age alone was significantly associated with the Hb levels.
Bivariate and multivariate analyses of the factors associated with Hb levels in the participants
Effect Value F Hypothesis df Error df Sig. Factor Pillai's Trace 0.001 .112b 1 197 0.738 Factor * BMI Pillai's Trace 0.001 .111b 1 197 0.74 Factor * District Pillai's Trace 0.019 3.715b 1 197 0.055 Factor * Gestational age Pillai's Trace 0.024 4.910b 1 197 0.028 Factor * Age group Pillai's Trace 0.002 .319b 1 197 0.573 Factor * Year Pillai's Trace 0.013 .670b 4 197 0.613 Factor * Sickling status Pillai's Trace 0 .025b 1 197 0.874 Factor * Age group Pillai's Trace 0.033 1.695b 4 197 0.153 Factor * Age group * Sickling status Pillai's Trace 0 .b 0 0 . Factor * Year * Sickling status Pillai's Trace 0.012 .580b 4 197 0.678 Factor * Age group * Year * Sickling status Pillai's Trace 0 .b 0 0 .
The association between multiple factors (BMI, district, gestational age, age group, year, sickling status, age group and year, age group and sickling status, year and sickling status, age group and year and sickling status) and the Hb levels of subjects (factor) was tested by bivariate and multivariate models. Gestational age alone was found to be significantly associated with the Hb levels of subjects (pregnant women of all age groups). BMI- body mass index; Hb- haemoglobin
The results of the tests of between-subject effects, conducted in 2-way ANOVA (95% confidence interval), displayed in Table 4, indicated that the year of pregnancy and age group together had a significant effect on Hb levels of subjects. In addition to this, the year of pregnancy alone also had a significant effect on Hb levels (p-value = 0.001).
Tests of between-subject effect of multiple factors on Hb levels
Tests of Between-Subjects Effects Measure: Hb Transformed Variable: Average Source Type III Sum of Squares df Mean Square F Sig. Intercept 589.003 1 589.003 178.552 < 0.001 BMI 5.24 1 5.24 1.588 0.209 District 5.459 1 5.459 1.655 0.200 Gestational Age 0.373 1 0.373 0.113 0.737 Age group 0.947 1 0.947 0.287 0.593 Year 62.294 4 15.573 4.721 0.001 Sickling 3.871 1 3.871 1.173 0.280 Age group * Year 39.659 4 9.915 3.006 0.019 Age group * Sickling 0 0 – – – Year * Sickling 19.566 4 4.892 1.483 0.209 Age group * Year * Sickling 0 0 – – – Error 649.857 197 3.299
Amongst a variety of factors tested -BMI, district, gestational age, age group, year, sickling status, age group and year, age group and sickling status, year and sickling status, age group and year and sickling status- the year of pregnancy alone and in combination with age group, were significantly in terms of determining Hb levels of subjects. Hb- Haemoglobin; BMI- Body mass index
The results of binary logistic regression analyses to determine the predictors of anaemia at antenatal registration (Table 5) and at week 36 of gestation (Table 6) showed the significant predictors of anaemia at antenatal registration to be BMI (p-value = 0.042), gestational age (p-value = 0.011) and pregnancy year. Specifically, as compared to others, 2011, 2012 and 2015 were associated with lowered odds of anaemia. On the other hand, the significant predictors of anaemia at 36th week of gestation, were the trimester of first reporting to the hospital, with the second trimester being the most significant, as well as the year of pregnancy, such that using 2011 as reference, later years were less associated with anaemia. Anaemia was defined using Hb levels of less than 11 g/dl. Hb levels of 11 g/dl and above were defined as non-anaemic (WHO, 2004).
Binary regression to determine the predictors of anaemia at antenatal registration
B P value Exp (B) 95% Confidence Interval for Exp(B) Lower Upper BMI −0.034 0.042 0.967 0.936 0.999 Gestational Age 0.057 0.011 1.058 1.013 1.106 Year 0.001 2012 −0.911 0 0.402 0.256 0.631 2013 −0.635 0.005 0.53 0.338 0.829 2014 −0.564 0.013 0.569 0.366 0.886 2015 −0.202 0.419 0.817 0.501 1.333
Significant predictors of anaemia at antenatal registration were BMI, gestational age and pregnancy year. Hb < 11 g/dl = anaemic; Hb levels ≥11 g/dl = non-anaemic (WHO, 2004). Hb- haemoglobin; BMI- Body mass index. B = slope. Exp (B)- odds ratio
Binary regression showing predictors of anaemia at the 36th week of gestation
VaVariable B P value Exp (B) 95% Confidence Interval for Exp (B) Lower Upper Trimester (1) .125 Trimester (2) −1.342 .042 .261 .072 .951 Trimester (3) −2.330 .063 .097 .008 1.137 Year .003 2012 −1.448 .001 .235 .101 .545 2013 −1.579 .001 .206 .081 .522 2014 −.634 .210 .530 .197 1.428 2015 −1.504 .053 .222 .049 1.018
At the 36th week of gestation, significant predictors were ANC reporting in the second trimester and the year of pregnancy. Hb < 11 g/dl = anaemic. Hb levels ≥11 g/dl = non-anaemic (WHO, 2004). Hb- Haemoglobin; BMI- Body mass index. B = slope. Exp (B)- odds ratio
The aim of the study was to analyse the trends and associated factors of anaemia between adult and teenage pregnant women in two rural districts, Ahafo Ano South and Asante Akim South, in the Ashanti region of Ghana. A study of the demographics indicated that the average age difference between teenagers and adults was about 10 years. This further showed that the teenagers had carried fewer pregnancies than the adults (lower gravidae index). They also had lower mean body weight, BMI and mean Hb at first antenatal services. These implied that the teenagers were more likely at risk of a poorer health status. Previous reports also indicated an increase in anaemia occurrence and severity with a higher gravida index [[
As part of antenatal services offered to these pregnant women, Hb levels are assessed on their first attendance to antenatal care and at 36 weeks of gestation. A comparison of the mean Hb levels between each year within each age group and between the two over the five-year period reviewed showed that between 2011 and 2015, mean Hb at registration reduced in adults by 0.3 g/dl while the teenagers increased by 0.8 g/dl. At 36 weeks of pregnancy however, the trend for adult pregnant women showed very little variation (stayed around 10 g/dl). Also, the Hb levels of the teenagers showed similarity to that of the adults, at 36 weeks of pregnancy, thus suggesting an improvement in the Hb levels of the teenage group. The reasons for this are not obvious, but it is possible that antenatal interventions were more effective in teenage group. Future studies especially, randomised controlled trials, are required to better understand the reasons behind earlier mentioned observations.
In general, the range of values obtained for anaemia prevalence over the years (that is 50–80%) was comparatively higher than that recorded in previous studies in other parts of Africa (42.7%) (Tunkyi and Moodley, 2015), as well as WHO estimates for anaemia prevalence (56%) in low and middle-income countries [[
Generally, mean Hb levels of both teenagers and adults were found to be higher by 36th week of pregnancy than at antenatal registration. Also, positive correlations were observed between Hb at antenatal registration and at 36th week gestation, indicating that pre/ early pregnancy Hb levels, largely determined anaemia occurrence during latter stages of pregnancy. Further, the above observation also implied that the effectiveness of iron-folate supplementation and other interventions to advance blood Hb levels during pregnancy was largely dependent on pre-pregnancy Hb levels. Thus, implying the need for a focus on pre-pregnancy measures to prevent anaemia in pregnancy. Measures such as encouraging food fortification and diversification which have been shown to reduce anaemia in women of child-bearing age, should be encouraged [[
As mentioned earlier, the Hb levels of the pregnant women were higher at the 36th week, suggesting an improvement over the course of pregnancy. However, according to findings, the average age of antenatal registration was 17 weeks of gestation, implying that pregnancy cases were presented to the health centres within the second trimester. Effective supplementation covers the course of the pregnancy (folic acid in first trimester, iron+ folic acid in second and third trimesters) and thus, although an improvement was observed, late reporting of the pregnant women to the health centers may have reduced the overall efficiency of the supplementation [[
In our study sample, BMI fell within the normal range (i.e 18.5 kg/m
Analyses of the factors associated with anaemia indicated that both the year of pregnancy and the age group of participants were co-effective for determining Hb levels, which might be predicted by socio-economic factors such as economic conditions and food availability [[
Also, from binary regression analysis, significant predictors of anaemia at antenatal registration were different from the predictors at the 36th week of gestation. BMI, gestational age at first antenatal service and year of pregnancy predicted Hb levels at antenatal registration. At the 36th week of gestation, only the year of pregnancy and the trimester of receiving antenatal services, specifically the second trimester, were significant predictors of Hb levels. Such information might be useful in developing a model for studying and monitoring anaemia prevalence and trends.
The devastating effect of anaemia on pregnancy outcomes has been widely recognized globally, and thus is a major setback in eradicating maternal mortality, with reference to the UN 2015 sustainable development goals. Although some studies have reported that anaemia may protect against stillbirth and preeclampsia in pregnancy [[
In conclusion, anaemia prevalence was high at both antenatal registration and the 36th week of gestation. Also, pregnant teenagers and adults were likely to visit antenatal services and start supplementation as late as 17 weeks into pregnancy further reducing the effectiveness of the supplementation programme. The teenage group was found as more anaemic, thus calling for interventions to reduce teenage pregnancies. Policies surrounding education, advocacy and mobilization should be put in place to engage teenagers especially on the medical complications associated with teenage pregnancy. The predictors of low Hb suggest general improvement over the years have led to improved Hb status in the population. Teenagers weight status or BMI should be considered since that also predicts their Hb at antenatal registration.
No external funds were received for this work.
Staff of the various district health centres who contributed to the success of this work in one way or the other.
RAA and CA were the main supervisors of the project; involved in the concept and design of the project, supervision of data collection, data analysis and interpretation, manuscript review and publication approval. MKMA and JJK were students on the project and were also involved in data collection and drafting of the manuscript for publication. All authors have read and approved of the manuscript.
All data generated or analysed during this study are included in this published article [and its supplementary information files [Additional file 1: COMPILED DATA (HEALTH CENTRES)_amended; Additional file 2: RAW DATA_ALL PARTICIPANTS_amended].
Ethical approval for the study was obtained from the Committee on Human Research Publications and Ethics, School of Medical Sciences, Kwame Nkrumah University of Science and Technology and Komfo Anokye Teaching Hospital, with reference, CHRPE/AP/222/17. For each of the health centres in each district, letters of approval were issued from the respective District Health Directorates (that is Asante Akim South Health Directorate and Ahafo Ano South Health Directorate) set in place by the Ghana Health Service, for data collection in all named health centres.
Not applicable.
The authors declare that they have no competing interests.
Graph: Additional file 1. A compilation of the raw data collected from all health centres included in this study. (XLSX 14 kb)
Graph: Additional file 2. A compilation of the raw data collected from the ANC records of the participants included in this study. (XLSX 78 kb)
• ANC
- Antenatal care
• ANOVA
- Analysis of Variance
• BMI
- Body Mass Index
• GHS
- Ghana Health Service
• Hb
- Haemoglobin
• HIV
- Human Immunodeficiency Virus
• IRIDA
- Iron Refractory Iron-deficiency Anaemia
• IUFD
- Intrauterine Fetal Death
• IUGR
- Intrauterine Growth Restriction
• MOFA
- Ministry of Food and Agriculture
• SPSS
- Statistical Package for Social Sciences
• WHO
- World Health Organization
Supplementary information accompanies this paper at 10.1186/s12889-019-7603-6.
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
By Millicenta K. M. Ampiah; Jerry J. Kovey; Charles Apprey and Reginald A. Annan
Reported by Author; Author; Author; Author