Background
Maternal death has declined substantially worldwide except in Sub-Saharan Africa [
1]. Of the 21 countries with the highest maternal mortality 15 are in sub-Saharan Africa, including Tanzania [
1]. In 2010, pregnancy and childbirth-related complications led to an estimated 454 maternal deaths per 100'000 live births in Tanzania [
2]. Most of these complications occur unpredictably during labour, delivery and the immediate postpartum period [
3]. Deaths could be averted with prompt and adequate diagnosis and care [
4]. However, 49% of all women in Tanzania still deliver at home without any skilled attendant [
2]. Moreover, according to the definition of the World Health Organisation (WHO) [
5] a quarter of all women in Tanzania begin childbearing as adolescents before reaching the age of 20 years [
2]. An estimated 70'000 adolescent mothers die each year worldwide because their bodies are not yet physically ready for motherhood and due to social disadvantages [
6,
7]. Pregnancy and childbirth thus constitutes the number one killer among 15-19 year old girls worldwide [
5].
Several studies have shown that women who started antenatal care (ANC) attendance early and attended frequently were more likely to be assisted during delivery by a skilled attendant compared to those who initiated ANC late and attended only few visits [
8‐
11]. Although ANC might not have the potential to predict and avert obstetric emergencies during pregnancy and childbirth, it exposes women to health education on risk factors and encourages them to deliver with a skilled attendant or in a health facility. Recent studies have suggested that women who knew about risk factors were more likely to utilize health facilities for delivery than those without knowledge [
10,
12]. Moreover, ANC provides the opportunity to detect and treat anomalies of pregnancy and to deliver preventive health services such as immunization against tetanus, prophylactic treatment of malaria and worms, and HIV testing and counselling (leading to Preventing Mother to Child Transmission of HIV, PMTCT) [
13]. To fully benefit from these interventions, it is important that women start ANC early on in their pregnancy. The revised Focused Antenatal Care (FANC) model of WHO [
13] as well as the Tanzanian FANC guidelines [
14] recommend at least four ANC visits for uncomplicated pregnancies with the first visit starting before 16 weeks of gestation [
13]. However, an analysis of Demographic and Health Surveys (DHS) from 45 developing countries showed that women in sub-Saharan Africa start antenatal care considerably later than women from other regions [
8]. Similarly, other studies reported late ANC enrolment after more than five months of gestation in sub-Saharan African countries [
15‐
18], including Tanzania [
2,
10,
19,
20]. A comparative analysis of the use of maternal health services in sub-Saharan Africa showed that adolescent mothers initiated ANC attendance even later and had poorer maternal health care than adult mothers [
21].
Quantitative studies on timing of ANC attendance from developing countries have been able to shed light on the influence of socio-demographic factors. Although there is mixed evidence, late booking of antenatal care has repeatedly been associated with young age [
21‐
24], premarital status [
21,
23] unwanted pregnancies [
16,
23,
25], high parity [
16,
21,
23,
26,
27], lack of formal education [
21‐
23,
27], low socio-economic status (SES) [
16,
23] and ethnicity [
16,
27]. Less is known about the influence of social and cultural determinants on prenatal care use among adult and adolescent pregnant women [
28]. Qualitative as well as quantitative studies have stressed the influence of social support from family members [
24,
29‐
31]. A study from Nepal for example reported the important role of mothers-in-law in deciding about ANC use of their pregnant daughters-in-law [
30]. Studies from Uganda showed that adolescents were more likely to experience violence from parents, to be rejected by their partner, expelled from school, and to be stigmatized [
29,
32], and therefore to hide their pregnancy [
32]. Late recognition of and uncertainty about the pregnancy [
33‐
35], as well as cultural beliefs and practices around pregnancy [
34‐
38], have been reported to influence women's timing of ANC attendance. Ethnographic studies from Mozambique and southern Tanzania illustrated for example that women at an early stage of pregnancy delayed ANC initiation purposely in order to protect the unborn from witchcraft and sorcery attacks of jealous neighbours and kin [
36,
37]. Other studies showed that women's ANC attendance is mediated by their experiences and the quality of care at earlier ANC visits [
39,
40]. These studies clearly indicate that beyond demographic and socio-economic factors, social and cultural factors as well as individual perception of pregnancy and care impact women's timing of ANC enrolment. Unfortunately, data are often not disaggregated by age, thus hiding particular vulnerabilities and issues [
6].
Exploratory studies carried out in the study area in 2007 as a preparation for this study confirmed several of the factors stressed in the literature. In semi-structured interviews [
41], health workers reported that women, and in particular women from the ethnic group of the Sukuma - semi-nomadic pastoralists who started to migrate into the region in the 1980s [
42] - initiated ANC attendance late and underutilized ANC due to lack of education and living in distant settlements. Data collected between 2007 and 2009 from the Health Management Information System (HMIS) of the health facilities within the study area [
41] indicated that the proportion of pregnant women who initiated ANC attendance after the fifth month of gestation rose from 53% to 56% between 2006 and 2008. Over this period, 18% of all ANC attendees were 19 years old or less. In an in-depth study with a small sample of recent adult and adolescent mothers (Gross 2007, unpublished data), adolescent women were found to visit the ANC clinic later and less frequently than adult women. Moreover, adolescent mothers differed from adult mothers in several ways: most of them were in their first pregnancy which was unplanned and prior to marriage, they still lived at their parents' home and they did not get any social or economic support from their partner or the child's father.
Based on the insights from the literature review and the exploratory studies, three main research questions arose that are addressed in this paper: First, do pregnant women - and in particular adolescent pregnant women - start ANC attendance late? Second, what factors are associated with early or late ANC attendance? And finally, do adolescent pregnant women differ from adult pregnant women in terms of social and economic support?
Discussion
This study showed that 71% of the pregnant women initiated ANC attendance after the recommended four months of pregnancy, at an average of 5.1 months (Table
2). This is consistent with the national average of 5 gestational months reported among facility users [
19]. A DSS household survey conducted in the study area around the same time found a similar average of 5.02 gestational months at women's first ANC visit [
47].
Adolescent pregnant women have been reported to most likely either not attend ANC or to attend late and infrequently [
5,
6,
21,
23,
31,
49] due to lack of knowledge, lack of power to take decisions, lack of money, or cultural factors including local concepts of illness [
5]. Contrary to the findings of these studies and our exploratory studies, we found no evidence of delayed attendance in adolescents (Table
2 and
3). In line with an early study from the US that reported lower prenatal care utilization among adolescents in their second pregnancy [
50] multiparous adolescents were found to start ANC attendance considerably later. Due to the study design of using exit interviews, we could only obtain information on women's timing of their first ANC visit and were unable to assess their overall utilization of ANC or even non-attendance. A study from Uganda comparing ANC attendance in adolescent and adult first time mothers found no difference in the timing of the first visit but a lower number of subsequent ANC visits in adolescents [
29]. Similarly, Magadi et al. [
21] found more variation by age with regard to frequency of ANC attendance than with timing. Little is known about adolescents ANC attendance in Tanzania, suggesting studies are needed to investigate their overall attendance.
Second, insights into factors influencing pregnant women's timing of ANC have been provided. Besides primiparity, having a history of a previous reproductive loss was found to be a strong predictor for an earlier ANC initiation in this study (Table
3). In accordance with other studies reporting that maternal care use varies across ethnic groups [
16,
47], the findings showed that the Sukuma ethnic group tended to have their first ANC visits later. Since Sukuma people live in very remote settlements of the study area, the effect is likely to be confounded by distance. GPS data is collected for each household within the DSS area, but unfortunately, it was not possible to merge this information with the demographic information collected during this study. Therefore neither data on distance between the homestead and the health facilities nor on women's socio-economic status were available for analysis. Some studies have reported an association between maternal secondary education and early timing of ANC initiation [
21,
51]. Contrary to these studies there was no evidence of such an effect in this study most presumably due to the overall low education level in the area, where few attend secondary school (see Table
1).
Women were well aware about their timing of ANC attendance, suggesting that confusion about the recommended starting time was not a problem. Few women (22%) could name more than four ANC services, but contrary to expectations, neither knowledge about correct ANC timing nor good knowledge of ANC services were associated with early ANC attendance (Table
3). Knowledge about available services might thus not imply that women are aware of the services' benefits. This matches with the surprisingly large number of women (53%) who indicated that they had attended ANC early because everyone does so, because of nurses' advice or because they feared the consequences of non-compliance with nurses' rules. In the exploratory study women indicated that their principal reason for attending the ANC clinic was to obtain an ANC card which was perceived as a necessary 'entry ticket' for services during delivery and illness rather than any conviction that ANC was good for their own or their child's health. The important pull factor of the ANC card has previously been reported by studies from Tanzania [
20,
52], South Africa [
35], Malawi [
34] and Uganda [
40,
53]. Attending ANC to obtain an ANC card thus might be one precautionary measure for women to conform to nurses' rules and to avoid harassment or informal payment requirements [
53,
54]. The extremely high rate of overall ANC attendance with 99.6% [
47] and an average of 3.1 visits to ANC clinics over the course of their pregnancy (personal communication: M. Alexander) reported from the area, suggests that by creating informal rules nurses successfully force women to attend the ANC clinics, however not necessarily at an early point of time. This is of course no reason to excuse nor to foster negative attitudes on the side of the nurses towards their clients, but rather calls for better health education in health facilities, outreach services and in the community. Trained focal persons such as TBAs, religious leaders and other opinion leaders working as community volunteers in close collaboration with existing community structures and health services have been found to be effective promoters of obstetric care but also of early and frequent utilization of ANC in Southern Tanzania [
55].
Perceived quality of care was generally high among the participants compared to a study from Kenya where almost a third of women complained about incomplete and inadequate services [
31]. Considering that ANC services were of similarly low quality in the study area [
56], women's high appraisal of the quality of ANC services rather reflects their low expectations of health care services. The fact that the interviewees were recruited at the health facility and interviewed in the proximity of the health facilities also potentially affected women's answers. Among those who were not satisfied with the services provided, perceived poor quality was, however, a strong predictor for late ANC attendance (Table
3). The findings indicate that quality of care, including patient-provider-relationship, plays a critical role in determining a woman's utilization of health care services, and needs to be improved but also better understood. In particular patient-provider-relationships should be further investigated through observational in-depth studies.
Late recognition of pregnancy was found to be a strong predictor of delayed ANC attendance in this study (Table
3). Similarly, late recognition of pregnancy and subsequent delay of ANC attendance has also been reported among South African women who received long acting hormonal contraceptives in the form of injections [
33]. Although pregnancy tests seem to be available at drug shops in the study area at a price of between 500-1000 TSh (~0.30-0.60 USD) they are not widely used (personal communication: I. Mayumana). More than a quarter of participating women said they waited for the quickening before initiating ANC attendance. However, due to the limitation of quantitative methods to investigate topics that need more in-depth inquiry and trust for people to discuss them openly, this study was not able to explain whether women only waited to ensure pregnancy or also due to other reasons. Studies from Tanzania and other sub-Saharan countries have shown that late disclosure of the pregnancy due to local practices or beliefs such as witchcraft is common and has a negative influence on the timing of ANC attendance [
34,
36,
37,
55].
The study provides evidence for the negative influence of lacking social and financial support on women's timing of their first ANC visit and the key role of the husband or partner. The results legitimize the attempts of the Tanzanian Ministry of Health and Social Welfare to encourage greater male involvement in maternal health issues [
57‐
59] in the sense that they are better informed about maternal health risks and live up to the expectations of support towards their children and their mothers. It is important, however, that this effort does not stop at policy level but reaches down to the health facility and community level. The community-based intervention conducted in Southern Tanzania found that in particular the equal inclusion of male community volunteers to promote obstetric care and early and frequent ANC use during home visits was an effective strategy to involve and inform men [
55]. Supporting income generating activities for women such as revolving funds might be a suitable mean to reduce delay due to lack of economic means needed for ANC or other maternal health services particularly among women who lack support from their husband or partner.
Third, exploratory analysis comparing adolescent and adult pregnant women in terms of social and economic support during pregnancy confirmed that adolescent pregnant women were less likely to be married than adult pregnant women. These findings are in line with a study from Uganda reporting that adolescent first time mothers were more disadvantaged in terms of their likelihood to be rejected by partners [
29]. On the other hand, in line with a study from Kenya [
31], adolescents in this study were more likely to receive advice to attend the ANC clinic than adult women. However, this advice was mostly given by their mother than by their husband, partner or the child's father. These findings suggest that close family members rather than the husbands, partners and child fathers play an important role in supporting adolescent pregnant women. The fact that lacking support from the husband or partner showed no significant effect in the sensitivity analysis that included young women who visited the ANC clinic for the first time (data not shown) further supports this argument.
While the support of relatives seems to be sufficient for adolescents to initiate ANC around the same time as adult pregnant women, the consequences of disadvantages in terms of social and economic support on the overall ANC attendance and - even more importantly - for delivering with skilled attendance and postnatal care for themselves and their child needs to be further investigated.
Conclusions
The majority of pregnant women delayed ANC attendance starting at an average of five months gestation. Adolescents had no greater delay in ANC initiation than adult pregnant women despite being more likely to be single. However, first ANC attendance at four months is recommended, so it is likely that some women missed important services offered during ANC such as preventive health measures, risk screening and health education.
This study found that many women rather attended due to norms and rituals than awareness about the health benefit of prenatal care; and that they delayed ANC initiation due to late perception of pregnancy, perceived bad quality of care and lack of social and economic support. These findings call for combined interventions at the community and health system level. Promotion of early and frequent ANC utilization through community-based interventions - involving also male community volunteers - could potentially be scaled up at low cost and adapted to local needs. Supporting income generating activities for women such as revolving funds might complement the approach in order to reduce delay due to lack of economic means needed for ANC or other maternal health services. At the same time, the quality of antenatal care services needs to be improved to attract women to use medical care throughout pregnancy, birth and the postpartum period; outreach services should be offered on a regular basis in order to bring services closer to women living in very distant settlements; and informal rules created by health workers in order to force women to attend the ANC clinic should be replaced with informing women about the benefits of maternal health services, but also the use of pregnancy tests.
Acknowledgements
We thank all the women for their willingness to participate in this study and the district authorities for their cooperation and interest. The study was funded by the Swiss Commission for Research Partnerships in Developing Countries (KFPE), the Novartis Foundation for Sustainable Development and the Rudolf-Geigy Foundation of the Swiss Tropical and Public Health Institute. None of the funders had a role in study design, data collection and analysis, decision to publish, or preparation of the manuscript. We acknowledge the institutional support of the Ifakara Health Institute. Special thanks go to all members of the ACCESS Programme, especially Judith John, Irene Ngoja and Hilda Mwakabusi, and to Phemy Muhaku, Prudenciana Kassim, Lucy Ngwatali and Salum Mbunda for their help with data entry, transcription and translation. We also thank the two reviewers for their helpful and constructive comments and inputs.
Competing interests
The authors declare that they have no competing interests.
Authors' contributions
KG was responsible for the design and implementation of the study, carried out the data collection, the data management and analysis, and wrote the manuscript. SA assisted with data management, statistical analysis and contributed to the interpretation of the results and the discussion of the manuscript. TRG supported statistical analysis and commented on the manuscript. JS and BO supported the design and coordination of the study and contributed to the discussion of the manuscript. All authors have read and approved the final manuscript.