Since late 1980s, the reduction of maternal deaths has been the main focus of international health agendas. Several international summits and conferences have taken this into account, including the Millennium Summit in 2000. One of the eight objectives of the Summit was to improve maternal health (MDG 5) and to reduce the ratio maternal mortality ratio by three-quarters between 1990 and 2015  .
Despite several progress in accessibility and availability of maternal health services, this ratio is still high and estimated at 450 maternal deaths per 100,000 live births (NV) in sub-Saharan Africa   . In Senegal, there are 392 deaths per 100,000 NV against 723 deaths per 100,000 NV in the region of Kolda where the health care situation is unsatisfactory  .
According to the World Health Organization, 80% of maternal deaths are due to severe bleeding, infections, hypertensive disorders during pregnancy and dystocia. The 20% would be related to indirect factors including anemia  . Prevention of this anemia requires iron supplementation since the only source of iron, which is the diet, and sometimes does not sufficiently meet the needs of the pregnant woman    .
In 2011, in Senegal, the proportion of pregnant women who consumed iron for at least 90 days was 63%. Factors of non-adherence to iron supplementation in pregnant women would be age, place of residence, educational attainment, household well-being, achievement of PNCS and accessibility, and the existence of certain cultural practices such as geophagy     . These factors are not exhaustive and are not all described. Thus, the interest of this study is to help better know the determinants of iron consumption during pregnancy for at least 90 days in the Kolda region.
The aim of this work is to study the determinants of iron consumption for at least 90 days during pregnancy in the Kolda region.
2. Framework, Tools and Methods
This study was carried out in the region of Kolda, in the south of Senegal. It covers 13 718 km2, i.e. 7% of the national territory  . The population of the region of Kolda recorded in 2013 is 633 675 inhabitants with a density of 46 inhabitants per km2, a rate of urbanization of 26% and 74% of population living in rural areas  .
The antenatal care utilization rate (ANC) was 101.6%, the assisted delivery rate was 34.3% away from the target of 40%. Contraceptive prevalence was low at 6.9% versus a national average of 12% and the rate of caesarean section was 1%  . The maternal mortality rate is 723 deaths per 100,000 new birth against a national average of 392 per 100,000 new birth   .
2.2. Type of Study
This was a cross-sectional, analytical study of the determinants of at least 90 days of iron use during pregnancy among women in the Kolda region. The study population represented the female population in child-bearing age in the Kolda region.
All women who gave birth in the last 12 months prior to the study, between February 2013 and January 2014, were included, regardless of whether the delivery occurred in hospitals or at home. The refusal to participate and the absence of the woman at the time of collection constituted non-inclusion criteria.
The sample size was determined by the Schwartz formula, with an expected frequency of 63% and an accuracy of 5%. Given that a two-level random survey would be conducted, the sample size, representative of each of the three health districts in the Kolda region, was estimated at 1413 individuals. The first level of sampling consisted of the basic draw of census districts from the sampling frame used by the National Agency for Statistics and Demography. The second level of sampling corresponded to the systematic sampling of households within pre-selected census districts.
2.3. Ethical Considerations
Participation in this study was voluntary after obtaining the informed consent of the woman. Anonymity was respected and the data collected were secured in order to preserve their confidentiality. The approval of the National Ethics Committee for Health Research in Senegal (CNERS) was obtained before the start of the field activities, as well as the necessary administrative authorizations.
2.4. Data Collection Tool
A previously tested and corrected questionnaire was administered to the women included in the study. This questionnaire included several aspects such as the personal characteristics, the knowledge of these women on the health of the mother and the child and finally the attitudes of supplementation. The woman’s personal characteristics related to the age of the woman, her marital status, her professional activities and the economic level of the household (assessed through an assessment of incomes, assets and household size). The knowledge evaluated concerned prenatal consultations (num- ber, importance...), prevention of anaemias and iron supplementation (importance, molecule to be used...). Prenatal consultation and supplementation practices were also assessed through prenatal consultations, accessibility of the PNC service, adherence to supplementation.
Data collection was carried out, over a period of 10 days, by three teams of ten investigators each, previously trained. The collection was done by face-to-face interviews between the interviewer and the woman at her home.
2.5. Data Capture and Analysis
The data thus collected were entered on Epi info version 3.5.3 and analyzed with software R.
A univariate analysis made it possible to describe the variables and to have their proportions for the qualitative variables or their averages and standard deviations for the quantitative variables.
The bivariate analysis made it possible to look for possible associations between the dependent variable (Iron consumption at least for 90 days) and the independent variables. Khi 2 and Fisher tests were used according to their applicability conditions with a significance threshold of 5%. Multivariate analysis was done in creating a simple logistic regression model from bivariate analysis by integrating all variables with p-values below 0.25 into the model  . The simple logistic regression was performed for the determination of factors related to iron consumption for at least 90 days with successive adjustment on personal characteristics, resources and needs taking into account the conditions of applicability, the threshold of significance remaining fixed at 5%.
Among the 1442 women surveyed, the mean age was 25.48 with a standard deviation of 6.5 years. The minimum age for women was 14, the maximum for 48, with a proportion of 73% aged between 20 and 34 years.
The majority of women (Table 1) were uneducated (65%), without income-gene- rating activities (52%), residing at a distance of less than 5 km from an antenatal clinic (55% On the importance of PNCS (63%) and iron supplementation (83%) and knew iron as an anemia medication (93%).
The study found that women did not perform at least the four recommended PNCS (68%), were supplemented with iron during the last pregnancy (91%) following a medical prescription (71%), and observed Supplementation for at least 90 days (56%) (Table 1).
Among women who had an income-generating activity, 52.4% had consumed iron for at least 90 days, compared with 48.9% for women who did not. However, this difference was not statistically significant (p = 0.20). In bivariate analysis (Table 2), iron consumption for at least 90 days was associated with personal characteristics (educational attainment, household well-being, distance from home to antenatal care, and woman’s obstetric experience), knowledge of women (receiving advice on the realization of PNC and iron supplementation, recognition of iron as an anemia treatment) and attitudes (achievement of PNCS and the age of the first PNC, prescribing method iron).
Multivariate analysis showed that the factors that increased iron consumption for at least 90 days were higher levels of education, higher levels of household welfare, an income-generating activity for women, the prescription of iron supplementation by health personnel, the precociousness of the first prenatal consultation (PNC 1) and the realization of at least four PNCS (Table 3).
The main limitation of this study is that the survey was conducted almost one year after delivery; which can be a source of memory bias. However the simplicity of the questions asked and the use of the notebooks of follow-up of the pregnancy made it possible to control it.
The results of our study show that 56% of women consumed iron for at least 90 days. The main determinants are personal characteristics (schooling, household well-being and income-generating activities) advice on iron intake and attitudes (method of prescribing iron, age of pregnancy at first NPC and number of NPCs performed).
Table 1. Distribution of women who gave birth between February 2013 and January 2014 in the Kolda region according to their personal characteristics, knowledge and attitudes (N = 1442).
Table 2. Characteristics, knowledge and attitudes of women who gave birth between February 2013 and January 2014 in the Kolda region and consumption of iron for at least 90 days.
Table 3. Modeling factors related to iron consumption for at least 90 days in women who gave birth between February 2013 and January 2014 in the region of Kolda.
Our study limits the fact that the survey was conducted almost one year after women gave birth; which can be a source of bias information including flawing memorization. However, the precision of the collection tool avoided this type of bias. The main strengths of this study are its accuracy because of the relatively high sample size (1442 women surveyed) and the method of analysis that allowed for correcting confusing factors.
According to the World Health Organization, all pregnant women in developing countries with a high prevalence of anemia must consume iron for at least 90 days to prevent pregnancy anemia and improve pregnancy outcomes, as well as maternal and child health    . This recommendation is not fully respected. Thus, the proportion of pregnant women who consume iron for at least 90 days during pregnancy varies from country to country and from region to region within the same country.
The results reported here are those of a cross-sectional analytical study carried out in the region of Kolda in 1442 women who gave birth between February 2013 and January 2014. They reveal that 91% of women had consumed iron during their pregnancy, but the proportion who consumed it for at least 90 days was only 51%. This low proportion has also been found in several developing countries such as Cambodia, Mali, Nepal and Indonesia, where they are respectively 47%, 28%, 20.7% and 40%     . However, relatively high rates have also been observed in other developing countries such as Nigeria  and India  where the authors reported rates of 65.9% and 64.7%, respectively. In Europe, studies of pregnant English women indicated that 70% of these women adhered to iron supplementation   .
These prevalence differences across the world may be related to the sociodemographic characteristics of the study population or to whether or not prenatal care is available; However, Lacerte et al. in a study conducted in Cambodia stated that this is rather the consequence of the ineffectiveness of the implementation of prevention programs in place  .
Several factors have been identified as associated with iron consumption for at least 90 days in pregnant women. Schooling, especially at the secondary or higher level, can positively influence this consumption. Indeed, in our study 74.1% of women who had this level of education had respected the appropriate duration of iron consumption and were 2.49 times more likely to respect it. Corresponding findings were found in Nepal  with an OR of 3.1, Nigeria  with an OR of 5.5 and Pakistan  with OR of 2.36. These results are similar to previous studies in the DHS where the authors found that women with a high level of education were likely to consume iron properly    . This may be explained by the fact that educated women have better access to health information, increasing their concern about their health and that of their future babies, but also have easier access to health information. Health services that provide iron.
Indeed, it is often through awareness that women receive information about health. Women of reproductive age who have attended health awareness sessions acquire knowledge that pushing them to adhere to the consumption of iron. In our study, nearly 60% of women met this criterion with ORa of 1.62. Previous studies in Cambodia  and Sweden  have also reached similar conclusions.
According to Khanal et al., counseling about the benefits of consuming iron through- out pregnancy and postpartum may influence women to consume it  . This hypothesis is verified in our study where women who received these tips were 2.66 times more likely to consume iron for at least 90 days. These results are similar to those found by several authors, such as in Senegal  , India  , Ethiopia  and Viet Nam  . This suggests that IEC interventions and especially the quality of information provided to women can play an important role in accessing health care facilities and hence in the consumption of iron for at least 90 days.
The results of our study showed that the higher the level of economic well-being of the household, the more likely a woman is to consume iron properly. Superposable results have been found in the literature       . On the other hand, studies in Denmark and Ethiopia have shown contrary results   . This may be explained by the fact that women from wealthy backgrounds may consider themselves to be less vulnerable to anemia and therefore have poor compliance.
Iron supplementation is an integral part of the refocused PNC and it is during these PNCS that midwives reiterate and reinforce the interest in iron use throughout pregnancy among women. The high percentages found in women who had early PNC (70.7%) and those with 4 or more PNCS (67%) were justified and were similar to those reported in several studies   .
We know that the prescription of iron is the proof of the contact between the pregnant woman and the health professional. At present, the association between prescription and iron consumption is not documented in the literature, but we have shown by showing that the chance of consuming iron for at least 90 days was multiplied by 1.64 Women with a prescription in iron than in women in self-medication.
The use of iron for at least 90 days during pregnancy remains an effective means of controlling serious deficiency anemia and thus reducing maternal and infant morbidity and mortality.
The results of this study reveal that this consumption is still low (51%) in the Kolda region, where it is influenced by a number of factors, including women’s education, awareness of women in health, counseling on the benefits of iron intake during pregnancy, the level of well-being of the household and the woman’s attitudes (number of PNCS performed, age of pregnancy at the first PNC and mode of prescription of iron).
Knowledge of these determinants may contribute to the development of good practice guidelines for reproductive health interventions to improve the prevalence of iron use for at least 90 days during pregnancy.
Acknowledgment to the Micronutrient Initiative that funded this study and to all Kolda women who agreed to participate in this survey.
No conflict of interest to declare.