Community-Based Maternal and Neonatal Health Services in Kolda and Sedhiou Districts of Senegal

Introduction: A recent assessment of the Millennium Development Goals has shown significant gaps in most developing countries, particularly with regard to the decline in maternal and infant mortality. Unaccess to health services is a major obstacle to reducing maternal and infant mortality. To support the Ministry of Health and Social Action in the implementation of its Strategic Plan for Reproductive Health (2012-2015), the NGO Micronutriment Initiative has developed a project based on Community-based Maternal and Neonatal Health services (CBMNH), which will be implemented at the level of the Kolda region. The general objective of this study is to carry out a situational analysis of maternal and neonatal health in the health districts of Kolda and Sedhiou to help the implementation and the evaluation of the project. Method: It was a descriptive and analytical cross-sectional survey. The study included women who gave birth between February 2013 and January 2014. The sampling was random at two degrees. The sample size was 471 women for each of the two health districts. The data collected during an individual interview focused on prenatal consultation (PNC), delivery, postnatal consultation (PONC) and exclusive breastfeeding (EBF). The quantitative analysis of the data consisted of the estimation of the main indicators, the comparison of the indicators between the health district of Sedhiou and the health intervention district of Kolda. Multivariate analysis identified factors associated with PNC, delivery, and EBF. Results: In total, 965 women were interviewed. The results of the study show that the average duration of PNC1 ranged from 3.41 months in Sedhiou to 3.82 months in Kolda. The proportion of women with full PNC was 38.8% and 54.9% respectively in Kolda and Sedhiou. The proportion of women who took iron-folic acid for at least 90 days ranged from 78.8% (Sédhiou) to 71.7% (Kolda). The delivery was carried out by trained staff in 61.0% and 57.4% respectively in Kolda and Sedhiou. The EBF was initiated in the first hour in 47.0% (Kolda) and 52.6% (Sédhiou). The results of the multivariate analysis showed that the early use of PNC services was related to low household income (0.65 [0.50-0.86]), proximity to the health facility (1.93 [1.34-2.78]) and multiparity (1.4 [1.05-1.87]). Completion of all PNC was associated with household income and early PNC1 (3.65 [2.58-5.18]). With regard to iron-folic acid intake, it was more common in women who achieved early PNC1 (2.19 [1.58-3.04]) and all PNC (3.58 [2.225.77]). The delivery by trained personnel was related to the proximity of the structure (2.43 [1.75-3.37]), but also to the preparation of the delivery. Women who knew the period of the MBF (1.37 [1.04-1.81]) and the protective role of the EBF (1.71 [1.28-2.27]) started their children early at the EBF. Conclusion: The results of the study show that the use of reproductive health services was linked to environmental, economic, cultural and structural factors. Improving the accessibility of these services requires joint action by the government, NGOs, community actors and the involvement of the population. This will be done through the construction, equipping and rehabilitation of health facilities, strengthening the knowledge of providers, community actors, women and support groups.


Introduction
The maternal and child health is a major concern in developing countries.This interest is justified by the importance of maternal and infant mortality.Among the indicators monitored by WHO, it is at the level of these two targets that we note the greatest difference between the high et low resource countries.According to the WHO, most maternal and infant deaths (99%) occur in developing countries, more than half of them in sub-Saharan Africa (WHO, 2012).A recent assessment of the Millennium Development Goals (MDG) has shown significant gaps in most developing countries [UN 2013].In Senegal, the maternal mortality rate is 392 deaths per 100,000 live births and the neonatal mortality rate is 29 deaths per 1,000 live births (National Agency of Statistics and Demography, 2012).
The factors related to maternal and infant mortality are well documented because they have been the subject of several studies (Khan et al., 2006).It is noted that deliveries sometimes occur in bad conditions of asepsis, complications are diagnosed late and evacuation to health facilities for better care is often very difficult.In Senegal, 26.8% of deliveries take place at home and the situation is more worrying in rural areas (39.2%) (Ministry of Health, Public Hygiene and Prevention, 2012).Under these conditions, newborns often benefit from precarious care.To all these factors is added the geographical and financial inaccessibility.Even when care is available, it is difficult to access because of the precarious economic situation of the populations.Households have few resources because almost half live below the poverty line.And when these resources are available, they are directed to other areas deemed to be priorities.The policy of the State of Senegal is to provide quality care as close as possible to the population through health structures and community-based services.According to WHO, about 80% of maternal deaths are due to severe bleeding, infections, high blood pressure and abortions (WHO, 2012); these deaths could be avoided through the implementation of key interventions offered as part of a continuum of care linking families and communities to health systems.
To accelerate progress in reproductive health, the Ministry of Health and Social Action of Senegal (MHSAS) has developed its Strategic Plan for Reproductive Health 2012-2015, whose objectives are focused on strengthening the community component of interventions to improve access to maternal and neonatal health (Ministry of Health, Public Hygiene and Prevention, 2012).It is within this framework that the Micronutriment Initiative (MI) is providing support to the MHSAS to help operationalize the implementation of this approach through a process that includes conducting a situational analysis at the national level.
The general objective of this study is to make a situational analysis of maternal and neonatal health in the districts of Kolda and Sedhiou before the implementation of the intervention.

Quotation
In this study, the goal is to perform baseline measurements and compare indicators and baseline characteristics between intervention and control areas.The study was carried out in the Kolda district selected as an intervention zone.The control district was chosen in the region of Sedhiou.According to the characteristics of the three districts of this region, that of Sedhiou seems the most appropriate because of its geographical location, socio-cultural similarities, the situation of health indicators, and the isolation of certain areas.Thus, the intervention district was compared to the control district during the baseline measurement and secondarily to assess the effectiveness of the intervention.
In this present study, the aim is to do the situational analysis, that is to say the pre-test through a descriptive and analytical transversal study.

Study Population
It is represented by all women of reproductive age in health districts of Kolda and Sedhiou.

Sampling
A two-stage random survey was conducted.The first degree consisted of the census districts (CD) of villages or neighborhoods and the second degree by women who had a newborn in the 11 months preceding the survey.

a. The Statistical Unit
Women who gave birth during the 11 months preceding the survey.

b. Inclusion Criteria
Was included in the study, any woman who gave birth: • A live or lifeless child at birth following a delivery in the health districts of Kolda and Sedhiou; • During the 11 months preceding the survey.

c. The Criteria of Non-Inclusion
Was not included any woman: • Absent during two passes; • Or refusing to give consent and / or assent to the interview.

d. The Size of the Sample
It was estimated on the basis of a 95% confidence interval (= 0.05 and 1- = 0.9), calculated according to the method of Casagrande and Pike- [Casagrande JT, Pike MC, Smith PG. 1978], and a non-response rate estimated at 5%.

e. The Sampling Method
For the recruitment of statistical units, a two-stage random survey was conducted.
• First-degree sampling This sampling consists of selecting the census districts (CD) to be included in the study.
It was necessary to randomly draw the number of CDs for each SD from the updated list of all DRs in the districts of Kolda and Sedhiou.
The random selection of the DRs was done using the Emergency Nutrition Assessment (ENA) software for each district.

• Second degree sampling
This sampling concerns the selection of women in the CDs.The selection of the statistical units was made by a draw of the villages or districts composing each CD.Thus, at the level of each selected neighborhood or village, all the women fulfilling the selection criteria were chosen until the necessary number was reached.

at. Key variables
The data collection focused on the key variables of the CBMNH project of Kolda project.

• The prenatal consultation
She was appreciated on the number and the period of the first PNC (PNC1).This information is collected during the individual interview with the woman and the exploitation of the pregnancy logbook.NPC knowledge and practices were appreciated.

• Iron and folic acid
The use of iron and folic acid was appreciated by the individual interview and the exploitation of the pregnancy logbook.

• Childbirth performed by trained staff
Delivery by trained staff was appreciated during the individual interview with the woman, the collection of the place of delivery and the person who assisted the woman.

• Exclusive Breastfeeding (EBF)
Breastfeeding should be exclusive for up to six months.It must also be initiated just in the first hour of birth.These aspects were sought during the individual interview with the woman.Knowledge about breastfeeding has also been sought.

Data Analysis
The data was analyzed with the R Software Survey package or the Stata software.This was done for univariate, bivariate and multivariate analysis.The data analysis consisted of two parts: a descriptive part and an analytical part.

Descriptive part
Qualitative variables were described with 95% frequency and confidence interval and quantitative variables by the average, confidence interval, and standard deviation.

Analytical part
It allowed the comparison of the main indicators between the intervention district and the control district.
• The bivariate analysis was done using the following statistical tests: the Khi² for the comparison of the proportions, the Student or the ANOVA for the comparison of the averages.When distributions were not normal, Mann Withney and Kruskall Wallis non-parametric tests and Fisher's exact test were used.This bivariate analysis also identified the factors associated with the key variables.
• A multivariate analysis was used to account for confounders.
This multivariate analysis allowed to model the key variables of the study that constituted the dependent variables.The explanatory variables related to personal characteristics, knowledge and practices in sexual and reproductive health.The model was developed using the approach described by Hosmer and Lemeshow (Hosmer & Lemeshow, 1989).All explanatory variables whose p are less than 0.25 were considered for the final model.The variables that were not selected (except those whose p are greater than 0.8) were introduced one by one and the model comparison was made by the likelihood ratio.Then, by a descending procedure, the least significant variables were removed one by one, and the comparison was made by the likelihood ratio test.Confusion was sought by a decrease of more than 10% in the beta coefficient (Greenland, 1989).

Ethical Considerations
Newsletters were sent to the various administrative authorities.At each district level, the administrative and local authorities have been informed.Informed consent was sought from all persons 18 years of age and older participating in the study.For persons under the age of 18, the consent of the parents or guardian has been requested, together with their consent.
They have been informed beforehand of the personal health questions that will be asked.

Discussion
This study allowed us to measure the main basic indicators of the SCSMN project, to compare them and finally to model the factors determining the PNC, the delivery and the early initiation to the EBF at Kolda district level.At first, factors influencing PNC, childbirth and then those influencing EBF were analyzed.

Prenatal Consultation
The PNC is a special moment of contact between the provider and the pregnant woman.It allows the detection of health problems and their follow-up in the mother and the fetus.It also allows the prevention of childhood diseases by vaccinating the mother (VAT for example).It should be an opportunity to give many tips for the implementation of health programs and should allow to retain the patient in relation to health facilities, including the place of delivery.In Senegal, four PNCs are proposed, the first of which in the first trimester of pregnancy.In this study we focused on early PNC, complete PNC, and iron-folic acid consumption during pregnancy.
The results of our study show that 51.7% of women achieved PNC in the first three months in Kolda against 61.8% in Sédhiou.The average duration of the PNC is lower in Sédhiou than in Kolda.The results of the multivariate analysis show that early PNC is related to personal characteristics, PNC knowledge, and advice received about the importance of PNC.
Several studies have shown the role of the economic level in the use of health services for prenatal care (Magadi et al., 2000;Koné-Péfoyo & Rivard, 2006;Faye, Manga et al., 2011).It allows the satisfaction of needs and access to structures in case of suffering.Wilkinson and Marmot in the report on the social determinants of health have shown the role of income on the health status of populations.This situation is found both among the poor and the rich.Solving health problems involves reducing inequalities.This requires the involvement of all actors, including political ones.The inaccessibility of structures is not only financial, but also geographical (Faye, Manga et al., 2011].In rural areas, women are often obliged to travel more than five kilometers to reach the nearest health facility due to poor road infrastructure and inadequate transportation. The achievement of early PNC is linked to the knowledge of the period and the number of PNCs needed during pregnancy.Particular emphasis should be placed on sensitizing women in health facilities for a better knowledge of PNC but also at home by Bajenu Gox and Relays, particularly in primigravida who do not yet have a great experience of pregnancy.Indeed, there are still some buildings surrounding pregnancy (Ndiaye et al., 2005).This makes the statement late to avoid evil spirits (Niang, 2003).To be effective the PNC must be early but also complete.The proportion of women who completed the complete PNC in Kolda is 38.8% in our study.This proportion is even lower if we rely on the mother's health record (18%).This difference can be explained by under-reporting of providers.The results of our study are similar to those found in EDS IV with a national average of 40% (Ndiaye, Salif, & Mohamed Ayad, 2006) but are lower than the national average in EDS V (50%) [Ministry of Health (Senegal)].However, this proportion was 41% in rural areas.Faye A in a study conducted in Gossas found that 35% of women had achieved 4 PNC (A.Faye, M. Faye, et al., 2010).It should also be noted that, of the women who have done a PNC1, most do not perform the four PNCs.Similar results have been found in EDS IV and V.This raises the problem of continuity of services.A special effort should be directed towards interventions aimed at getting women to continue PNCs at the level of structures but also at community level through greater involvement of support groups.Bajenu Gox too can play a big role at this level by accompanying women in this process.The results show that women who had received information about PNC were more likely to complete four PNCs hence, the importance of emphasizing women's awareness and information in order to improve their knowledge.In fact, knowledge improves as well PNC1, PNC coverage and iron-folic acid intake for at least 90 days.
Iron-acid consumption for 90 days is a recommendation of the Reproductive Health Directorate.The results of the multivariate analysis show that accessibility is a major determinant in the consumption of iron-folic acid during pregnancy.Community actors can play an important role in reducing these difficulties by facilitating access at the community level.However, the results of the study show that they are very weakly involved in the distribution of iron.

Delivery
Home birth by unskilled people in poor asepsis conditions is one of the leading causes of maternal death, unlike in developed countries where delivery is done by trained staff.Home delivery ranged from 39% in Kolda while in Sedhiou this proportion was 41.3%.In the EDS-MICS 2010-2011, 56% of deliveries took place at home in Kolda.The national average, which was 38% in 2005, decreased to 28.8% in 2010-2011 (National Agency for Statistics and Demography, 2012).
Unlike developed countries where it is well planned with the presence of qualified personnel, home delivery in Africa is very problematic.The results of our study show that it is mainly related to the preparation of childbirth.Similar results were found by Faye and al in a study conducted in five regions of Senegal (Faye et al., 2010).The preparation of childbirth helps to better involve women in the management of pregnancy.Indeed, the identification of the health facility where the childbirth is to take place and the person who must perform it, allows the woman to have more information about the pregnancy especially during the delivery period.The surprise effect of childbirth is often responsible for home delivery (Bolam et al., 1998).Often, women do not know the time of delivery.Particular emphasis should be placed on the preparation of delivery during PNCs.But this assumes that health care providers are equipped enough to do so, which is not often the case.These aspects should be taken into account during their training.The involvement of Bajenu Gox can be an alternative.However, the results of the study show a low involvement of the latter.In this project, special emphasis should be placed on greater involvement of the Bajenu Gox.
The preparation for childbirth can be hindered by the woman's position in society.Her limited decision-making power in the household does not give her the ability to negotiate and make certain decisions (Nikièma, Haddad, & Potvin, 2008).She is also often very badly seen in these societies because considered as an interference with the divine will.It is clear that to get around these cultural barriers, it is important to involve husbands, religious and customary leaders.
The situation is all the more alarming that the government has implemented a policy of free of charge cesarean delivery (Witter, Armar-Klemesu, & Dieng, 2008).The Free of charge responds to logic of social justice that aims to prevent the poorest from being marginalized and falling into permanent exclusion (De La Rocque, 1996).
In most countries where these policies have been put in place, they have promoted access to care for the poor (Ensor & Ronoh, 2005), thereby reducing social inequalities in health (Nabyonga, Desmet, & Karamagi, 2005).However, some writings show that they do not always benefit the poorest (Asante et al., 2007).In some contexts, they would have even benefited the non-poor more than the poor (Leighton & Diop, 1995).Indeed, In order to benefit from this free service, you must have access to health services.
In our study, distance is the main factor of geographic accessibility (Pison, 2000).Similar results have been found in Burkina Faso (Haddad, Noutgara, & Ridde, 2004), Kenya, Rwanda, Sudan (Pearson & Shoo, 2005) and Uganda (Mbonye, 2001).The accessibility of health facilities poses real problems in developing countries, particularly in Africa.In rural areas, one in three women must travel more than five kilometers to reach the

Table 1 .
Distribution of women by health districts

Table 2 .
Main characteristics of the mothers of the 2 health districts

Table 3 .
] Main household possessions in the 2 health districts

Table 4 .
Distribution of women according to the knowledge and advice received in the 2 health districts

Table 5 .
Main indicators of the CBMNH project

Table 6 .
Main indicators of the CBMNH project

Table 9 .
Factors associated with iron consumption within 90 days Factors Determining Iron-Folic Acid Intake for at Least 90 Days During Pregnancy Table13.Factors determining iron-folic acid intake for at least 90 days during pregnancy