Use of antenatal services and delivery care among women in rural western Kenya: a community based survey

In 2000, it was estimated that approximately 529 000 women died from complications related to pregnancy or delivery [1]. The majority of maternal deaths occur in developing countries. Causes of maternal deaths include complications of abortion, obstetric complications such as hemorrhage, dystocia, eclampsia, sepsis and infections such as tuberculosis and HIV-1 [2, 3]. Attendance at antenatal clinics (ANCs) and receipt of professional delivery care have been associated with a reduction in maternal deaths [4, 5]. The ANC system in developing countries has been adapted from developed countries without formal evaluations of the impact of interventions in developing country settings. [6]. More recently, studies have been conducted to identify procedures proven to improve reproductive health [7‐9]. These evaluations demonstrated that in low risk pregnancies (pregnancies with no abnormal obstetric or medical history), a decrease from 12 ANC visits, as had been previously recommended, to a less-costly 4 visit-schedule did not result in an increase in adverse maternal and perinatal events [10‐12]. As a consequence, the World Health Organization now recommends a 4-visit ANC schedule for low risk pregnancies [13]. Other interventions shown to be beneficial to mother and child include routine iron and folate supplementation in areas with a high prevalence of anemia, serologic screening for and treatment of syphilis, routine measurement of fundal height, malaria prevention, and tetanus immunization [9]. To fully benefit from these interventions, it is important that women begin attending ANC early in pregnancy.

In many developing countries, the majority of births occur without the help of a skilled assistant (defined as a midwife, nurse trained as midwife, or a doctor) at home or in other non-hospital settings [14]. Home deliveries in the absence of skilled professional attendants have been associated with adverse infant and maternal outcome [15, 16]. However, home deliveries without a skilled attendant are chosen or occur for a variety of reasons, including long distances or difficult access to a birth facility, costs of services and perceived lack of quality of care in a health facility [17‐19]. In an attempt to improve care during home deliveries and reduce maternal mortality, traditional birth attendants (TBAs) have been trained in modern delivery care, with varying reports of success [20‐22]. Presence of a professional attendant at each birth can lead to a marked reduction in maternal mortality and morbidity [15, 16]; professional health care during childbirth is one of the process indicators to assess progress towards the Millennium Development Goal of improving maternal health [23].

We conducted a community-based survey in rural western Kenya, where anemia, malaria and HIV rates are high, among a representative sample of women who had recently delivered, to identify the locations where they had sought care during their pregnancy and delivery and the range of services they had received. This information will help to improve maternal health care in this area.

The survey was conducted in the rural area of Asembo (Rarieda Division) and Gem (Wagai and Yala Divisions) in western Kenya. Asembo has a population of approximately 55,000 people living in 79 villages. Gem has 75,000 inhabitants living in 144 villages. This area historically experienced intense perennial malaria transmission before the initiation of a large community-based trial of insecticide treated bed nets (ITNs) that reduced malaria transmission considerably [24, 25]. Following this trial, a health and demographic surveillance system (HDSS) was established in the same area [26]. All births, deaths, pregnancies and in- and out migrations are recorded and updated every four months. Socio-economic information and educational level are obtained and updated annually. All households who participate in the HDSS receive ITNs; free retreatment is available in each village every 9 months.

Using a population-based sampling frame of women in the HDSS who had recently delivered, we selected a random sample of 730 women. We estimated that a sample size of 730 was required to calculate the proportion of those using antenatal or delivery care in both Asembo and Gem within 5 percentage points of the true proportion, assuming the true proportion was 70% and that 12% of women would not be available. Trained interviewers visited the selected women at home and administered a standardized questionnaire in the local language. Participants were asked if they had visited an ANC during their last pregnancy. If they answered "Yes," they were asked to specify their reasons for doing so; this question was repeated until no new reasons were reported. Answers were tick marked in predefined options or described under "other reasons." Interviewers were instructed not to probe with options. Socio-economic information was obtained from the HDSS. Information on household assets was used to derive a wealth index using the method developed by Filmer & Pritchet (2000) [27]. All households were categorized by quintile: a medium/low socio-economic status (SES) was defined as an SES in the bottom 3 quintiles of the wealth index.

Differences in proportions were compared using the Chi-square test or Fisher's exact test when appropriate, and differences in means were compared using the Student's t-test (SPSS for Windows 11.0, SPSS Inc. Chicago, Illinois). A two-sided P-value < 0.05 was considered statistically significant. Odds ratios with 95% confidence intervals were calculated. Results were stratified by parity, age, level of education (< 8 years versus ≥ 8 years of education) area of residence (Asembo versus Gem), and history of child death or previous stillbirth. Infrequent usage of ANC was defined as having made < 3 ANC visits, the minimum number needed to benefit from intermittent preventive treatment with sulfadoxine-pyrimethamine for malaria. Primiparae were women who had delivered for the first time, multiparae were women who had delivered more than once. Grande multiparae were defined as women who had delivered 5 or more times, given that a high number of previous deliveries is an obstetric risk factor. Logistic regression was used to assess the effect of various factors on ANC attendance, place of delivery and type of attendant at delivery. The effect of the following factors was examined: maternal age, parity, marital status, socio-economic status, education level, area of residence (Gem versus Asembo), distance to the ANC and history of stillbirth or child death. Factors significant in the univariate analysis were included in the multivariate analysis.

The project was reviewed and approved by the institutional review boards of the Centers for Disease Control and Prevention (Atlanta, GA) and the Kenya Medical Research Institute (Nairobi, Kenya). All women who participated gave written informed consent after reading through the consent form with the interviewer; participants who could not write indicated their consent by a fingerprint, which was witnessed by the interviewer.

In this survey in a rural area in western Kenya, 9 out of 10 women reported at least one ANC visit during their last pregnancy; however, two-thirds of these women began attending the ANC in the third trimester, and only half of these women made the recommended number of 4 visits. Services provided by the various ANCs were not optimal, with a low coverage of intermittent preventive treatment with sulfadoxine-pyrimethamine, and supplementation of iron and folic acid. The proportion of women assisted during delivery by a professionally trained provider (17%) was similar to the proportion of women who delivered on their own (18%). Untrained persons and TBAs assisted in the majority of deliveries (65%).

The high ANC attendance (90%) in this rural area is encouraging, and is similar to a report for the whole of Kenya (88%) [29]. Having more than 8 years of education and being of higher socio-economic status were the most important factors associated with ANC attendance; these findings are similar to those from an evaluation of antenatal care attendance in developing countries [13]. Although the perceived expense of the ANC may hinder attendance, it is uncertain that free antenatal care would increase coverage substantially because transport costs, physical inability to travel long distances, and a perceived poor quality of care would remain barriers. Unlike Nigeria and Uganda, where the husband played an important role in determining their partner's ANC attendance, in our study most women made an independent decision to attend [17, 30].

There were fewer services offered by the ANCs attended by the women in this study compared to the whole of Nyanza Province; participants in this survey were half as likely to receive information on danger signs in pregnancy, or to receive a urine test (30% versus 14%, and 35% versus 16% respectively) than participants in the 2003 Demographic and Health Survey (DHS), and they were less likely to report a blood pressure measurement (67% versus 76%, P = 0.001) [29]. Women appreciated the information and advice received at the ANC. The main topics were care during pregnancy and care for the newborn. However, few women attended a health talk (14%), and other essential topics such as place of delivery, making an individual birth plan, family planning, malaria, and HIV/AIDS prevention received little attention. Information on the effect of health education on maternal and perinatal outcome in sub-Saharan Africa is limited; a study in Tanzania assessing the effect of group and individual counseling on the prevention of anemia did not see an additional effect of this approach above the intervention of ensuring an adequate supply of hematinics and malaria prophylaxis [31]. Other services with proven benefit had a low coverage in this survey as well; barely half of the women received hematinic supplements, one out of 5 women received one or more doses of sulfadoxine-pyrimethamine for malaria, and an antihelmintic treatment was received by 3%. This is in stark contrast with the needs as confirmed during a cross-sectional survey in July 2003 among 673 pregnant women in the same area; 36% of the women were malaria parasitemic (52% among primigravidae), 53% were anemic (hemoglobin < 11 g/dl) and 76% of the 391 women who had brought a stool sample were infected with a nematode (van Eijk et al., unpublished). In our study, we did not investigate the reasons why many ANC visits passed without provision of these services, e.g. if there was a lack of supply of drugs; this is a focus of future evaluations.

Consistent with a report from a neighboring area, only 37% of the woman started attending ANC in the first or second trimester [32]. Low ANC attendance in the first trimesters has also been reported in other African and developing countries [19, 29, 30, 33, 34]. Late ANC attendance may preclude women from benefiting fully from preventive strategies, such as iron and folic acid supplementation, treatment of helminthic infections, and intermittent preventive treatment with sulfadoxine-pyrimethamine for malaria in pregnancy. Late attendance was associated with a perceived lack of services at the ANC. As summarized in table 2, provision of services in this area was generally poor and worse than that reported in the national ANC service profile in the most recent DHS [29]. To encourage earlier ANC attendance, service delivery must be improved.

Despite the fact that 90% of the women reported attending antenatal care, fewer than 2 in 10 gave birth in a health facility. This is lower than the national estimate of 40% for Kenya and 39% for Nyanza Province [29], although higher than that reported in some other African countries [14]. Observed contributing factors for a home delivery included the fast progression of labor, distance, difficulty of (night) travel, and cost. Distance was a barrier for facility delivery but not for ANC attendance. This may be because not every ANC facility has 24-hour maternity service, so the distance to reach a health facility with maternity service may be greater than for antenatal care. In addition, given the urgent nature of deliveries, there may be less time to cover the distance. Other work has confirmed the importance of distance on access to maternity care [18, 34].

We were surprised at the high proportion of women who gave birth without any assistance. Our estimate of 18% was nearly twice that observed in the most recent Kenya DHS (National estimate 8.0%, Nyanza Province 9.8%) and in a study in Mbeere District, Kenya (6.5%) [19, 29]. Unassisted delivery was particularly common among grande multiparae, a high-risk group in obstetrics [19, 29]. The lack of any attendant makes it difficult to seek assistance in the event of life-threatening complications. Women should be strongly encouraged to deliver with assistance. The promotion of a delivery plan may be a good step towards sensitizing women on this issue.

TBAs frequently form the backbone of maternity services in rural areas, and in our survey they attended about one-third of deliveries. It has been suggested that training of TBAs could reduce maternal and perinatal mortality, but recent data have not supported this strategy [20, 35, 36]. Some participants preferred TBAs because of greater flexibility in payment.

Limitations of this survey include the fact that the sample of women who could not be found were younger than those enrolled, and young women were more likely to visit the ANC; however, even if all excluded women would have visited the ANC, the estimate of the prevalence of having attended ANC at least once would not have changed substantially. We did not cross check the information provided by the women during the interviews with ANC card data; thus, services actually provided could have been underreported. Lastly, only women who survived their last delivery were able to participate in this survey.

The Safe Motherhood Initiative promotes antenatal care, and skilled assistance (defined as a midwife, nurse trained as midwife, or a doctor) during childbirth [37, 38]. Based on this definition, there is much to be done to achieve 'Safe Motherhood' in rural areas of western Kenya: a trained provider attended fewer than 1 in 5 deliveries. The discrepancy between the national and provincial estimates of unassisted deliveries or deliveries without skilled assistance and our findings illustrate that these estimates may underestimate the situation in the rural areas. Safe motherhood strategies promote counseling of clients on danger signs and the development of an individual birth plan; pregnant women and their relatives are encouraged to arrange transport, money and a companion before onset of labor [28]. This survey indicates that the situation in western Kenya needs urgent improvement. The Kenya Ministry of Health is working on expanding focused antenatal care and prevention of malaria in pregnancy in rural areas [28]. This survey can provide baseline data for this rural area against which efforts to improve services in the ANCs can be evaluated.