Why don’t illiterate women in rural, Northern Tanzania, access maternal healthcare?

Globally, the maternal mortality ratio declined by 38% between 2000 and 2017. However, in 2017, 810 women died every day from pregnancy- or childbirth-related complications; 94% of these deaths occurred in low and lower middle-income countries (LMICs) [1], most commonly in rural areas [2]. Sub-Saharan Africa (SSA) has the highest regional maternal mortality ratio at 546/100,000, which is three times higher than the next highest region, South Asia (182/100,000). In 2015, Tanzania’s maternal mortality ratio (556/100,000) [3] was the 15^th worst among the 46 SSA countries [4]. Moreover, the adult lifetime risk of maternal death of 1 in 36 in SSA contrasts markedly with 1 in 3300 in high-income countries [5].

Common causes of maternal death in SSA include pre-existing disorders such as HIV, anemia, or malaria, exacerbated by pregnancy (28.6%), haemorrhage (24.5%), hypertension (16.0%), sepsis (10.3%), abortion (9.6%), other direct causes (9.0%), and embolism (2.1%) [6]. What is more, the pregnant woman is not the only one at risk; globally, in 2015, 2.6 million babies were stillborn [7] and in 2017, 2.6 million children died in their first month of life, of which roughly 2 million (77%) died within the first six days of life [8], most commonly in least developed countries [9]. In 2018, Tanzania’s neonatal mortality rate was 21.1, in comparison to high-income countries where the rate averages 3.1 [9]. Disparities between least-developed and high-income countries in the maternal mortality ratio, adult lifetime risk of maternal death, and neonatal mortality rates suggest that many more pregnancy-related deaths in least-developed countries could be prevented.

In order to improve maternal and neonatal health, the WHO recommends that pregnant women and neonates receive globally recognized standard of healthcare. This includes a series of antenatal care (ANC) visits [7], giving birth at a health facility with a skilled birth attendant, and postnatal care provided to both a mother her infant [10]. ANC enables early screening and recognition of problems and co-morbidities, prevention of complications, micronutrient supplementation, and, crucially, an opportunity for healthcare providers to communicate with and support women during a critical period [7, 11, 12]. In Tanzania, a woman who attends ANC is twice as likely to deliver in a health facility than a woman with no ANC visits [13]. Giving birth in a health facility with a skilled birth attendant, in turn, reduces stillbirths and deaths from intrapartum–related complications by about 20% [14]. Despite these impactful interventions, only about half of pregnant women globally receive even four ANC visits [11]. Similarly, in Misungwi District, Tanzania only 47% of mothers surveyed in 2016 had attended four or more ANC visits [15] with their last pregnancy and 61% delivered at a health facility [13].

Variables influencing whether pregnant women receive recommended maternal health interventions have been explored in several studies. In both India [16, 17] and SSA [18‐20], researchers identified illiteracy (defined as not knowing how to read or write) as an important risk factor for failure to attend ANC and higher maternal mortality. This is consistent with general linkages between low literacy and worse health outcomes [21] and lower utilization of health services by illiterate women [22]. Some studies have identified lower care-seeking due to the stigma associated with illiteracy: some illiterate individuals were hesitant to disclose their illiteracy to healthcare workers, compromising provider support for such individuals [23‐25]. Studies elsewhere have documented a positive correlation between level of maternal formal education and perinatal healthcare seeking behaviour [13, 26‐33].

Relatedly, a 2016 survey in Tanzania’s Misungwi District, Tanzania [15] documented that 35% of sampled mothers women were illiterate, versus 26.91% for Tanzanian women overall [34]. In this district, 41% of illiterate women report attending ANC at least four times compared with 49% of literate women, a significant difference (2 × 2 χ², p = 0.02). And, in comparison to the level of formal education received, number of children, and maternal age, illiteracy was the strongest predictor of having fewer than four ANC visits. Likewise, illiteracy in Misungwi District was significantly associated with not being attended by a skilled birth attendant when giving birth (2 × 2 χ², p = 0.0008) and was the third strongest predictor of giving birth without a skilled birth attendant after the number of children the woman had and level of formal education she had received [15].

Hence, illiteracy is a well-established barrier to receiving the recommended antenatal, giving birth and postnatal care, yet there is little documented about the lived experiences of illiterate pregnant women and mothers with newborns. Indeed, only one paper [25] providing qualitative data on this topic was found and none were found for rural regions or for LMICs. This is concerning since effective public health policy should be informed by local context [35‐37]. What is it, for example, that renders illiteracy a risk factor for not receiving the recommended standard care during pregnancy, when giving birth, and after giving birth? In addition to missing data on lived experiences of illiterate women, specific obstacles to meeting recommended standards of perinatal care are sorely lacking for rural regions of LMICs. This study, consequently, was designed to begin to address this knowledge gap by using qualitative methods to capture the lived experiences related to care-seeking by illiterate, pregnant women and mothers in a rural region of an LMIC. Specifically, this study asks the following question: why don’t illiterate, pregnant woman in rural Tanzania access the recommended standard of maternal healthcare services?

During design, data collection, and analysis, this study adhered to the consolidated criteria for reporting qualitative research (COREQ) when possible as outlined in Additional file 1: Appendix [38].

The 81 illiterate women of reproductive age had a mean age of 32 ± 11 years, a mean of 1.7 ± 0.82 children under the age of 5 (Table 1). None of these respondents had any education. All were subsistence farmers engaged in farming small plots of land and 84% were married. All of these women spoke Sukuma, and none spoke or understood Swahili.

Two content-driven themes related to antenatal care, birth and postnatal care access by women emerged from the analysis: 1) a communication gap arising from a) these women’s inability to read public health messaging documents provided by health facilities, and b) healthcare providers speaking Swahili, but not the local dialect, Sukuma, and 2) a dependency of these women on family and neighbors to negotiate their illiteracy and inability to speak Swahili. Each of these themes was apparent in transcripts from all participant categories. Moreover, there was no indication in the interviews that these women lacked health literacy; they knew that they should be attending the clinic for care but found it difficult to do so due to the communication gap.

Globally, pregnant, illiterate women and illiterate mothers with newborns are less likely to receive the recommended standard care, with detrimental health sequelae. In Misungwi District, in comparison to literate women, illiterate women, who represent 35% of the women in this district, were significantly less likely to have at least four ANC visits or give birth under the care of a skilled birth attendant [15]. Why these women were less likely to access maternal care was unclear, requiring a deeper consideration of the link between these women’s perceptions, behaviors, and socio-cultural environment. Toward that end, this study endeavored to document the lived experience of illiterate women in rural Tanzania as it relates to the recommended standard of maternal healthcare.

Across the FGDs, KIIs, and IDIs with the illiterate women, CHWs, and healthcare providers in this study a consistent theme emerged: the inability of illiterate women to read, speak, or understand the language of healthcare providers was a formidable obstacle to illiterate women seeking maternal healthcare. Illiteracy is associated with two obstacles to receiving the recommended standard care and one coping strategy. First, illiterate women cannot understand written information provided to them by healthcare providers and CHWs. Clearly, this renders these women’s ability to recall appointment dates and read public health messaging a challenge. There is no apparent initiative on the part of healthcare providers or CHWs to determine which women are illiterate and illiterate women may be shy about admitting their illiteracy [23]. In the words of one CHW, determining who is illiterate is rendered more difficult by the fact that “illiterate mothers just listen and don’t ask many questions, keeping their lack of understanding hidden from me so that they can just finish the appointment and leave” (KII). If the CHW and healthcare provider are not aware of who is illiterate, their ability to address this issue is compromised. If they were aware of the woman’s ability to read and write, the CHW and healthcare providers did seek to assist the woman. For example, an illiterate woman noted that if you tell the CHW and healthcare provider that you are illiterate, they “will know how to help you or they will tell you to find anyone who can help to read the information” (FGD). Second, exacerbating their inability to read is the fact that the women in this study received no formal education, so did not learn to speak Swahili. These women therefore cannot understand those healthcare providers who do not speak their language, Sukuma. This inability to speak the language of those providing them with healthcare compromises their healthcare experiences. As noted above, a key benefit of ANC is the opportunity for healthcare providers to communicate with and support women during a critical period [7, 11, 12]. For example, exchanges between healthcare providers and pregnant women facilitate the relay pertinent information that may not be apparent to the healthcare providers, such as previous pregnancy complications, number of miscarriages or neonatal deaths, or a history of gestational diabetes. This benefit is lost for the women in this study. Even if they do receive the minimum standard of care when pregnant and with a newborn, the language barrier suggests that these illiterate women may gain little beyond the vaccines their newborns receive. Hence, the problem is larger than whether or not the mother can read or write; rather, illiteracy is a marker of ability to speak the language of healthcare providers. Furthermore, this problem may not be limited to illiterate women, as one CHW pointed out that “even if [a woman] can read Swahili, some will not speak Swahili” (KII).

The CHWs speak both Sukuma and Swahili and could translate for the women. Indeed, CHWs indicated that they speak Sukuma with women from their community as much as possible, with one noting that “if you speak to them in their language, they can express themselves very well” (KII). However, translating the healthcare providers’ Swahili for the illiterate women may not be feasible as CHWs are generally not present during ante- or post-natal care visits or when a woman gives birth with a skilled birth attendant.

These findings are consistent with the robust body of work noted above that finds a strong association between women's formal education and child health elsewhere [13, 26‐33]. The current study considers the association between the level of formal education and use of healthcare services more deeply. Here, the driver for not receiving the recommended maternal healthcare is neither the lack of formal education nor the lack of understanding about the need for such care, generally, but, specifically, is the inability to understand the language of both healthcare providers and inability to read public health documents due to the lack of formal education.

The illiterate women in this study did not report feeling stigma about their inability to read or write, contrary to other research [23‐25]. Within social groups, stigma is believed to function to reinforce power inequities, maintain social norms and avoidance of disease [65]. As regards norms, stigma is argued to be tightly tied to a society's ideas of how things should be [66] and arises when a phenomenon is perceived as a threat to material goods or symbolic goods such as beliefs, values, or ideology [67]. Thirty-five percent of the women surveyed in Misungwi District are illiterate [15]. As illiteracy is not uncommon in these communities and may have little bearing on a woman’s productivity as a subsistence farmer, it may not be perceived as a threat to norms and material and symbolic goods so may not be stigmatized. That said, the fact that the CHWs and healthcare providers stated that illiterate women did not reveal their illiteracy may indicate that these women do feel stigmatized and, as such, were remiss to admit feeling stigmatized in the FGDs or KIIs. If they do feel stigmatized for their illiteracy, this would very likely be a barrier to healthcare seeking [66].

The agency of these illiterate women is manifest in their efforts to negotiate their illiteracy via assistance from their relatives or neighbors to help them to understand public health messaging and to remind them of appointment dates written in their healthcare card. While these actions may well serve their purpose for some, being illiterate was the most important risk factor for failure to attend at least four ANC visits [15]. As above, even if an illiterate woman’s coping strategy gets her to her healthcare appointment on the assigned day, if the healthcare provider speaks only Swahili and there is nobody to translate Swahili into Sukuma, her healthcare experience may be compromised.

Interestingly, Tanzania’s 2015–2016 Demographic Health Survey (DHS) [13], finds that 67.4% of women aged 15–49 in the Lake Zone reported that they have serious problems accessing health care when they need it. In this DHS, women were asked if the problem accessing health care was due to insufficient money for treatment, distance to health facility, not wanting to go alone, and/or failure to obtain permission to go, and replied in the affirmative 52%, 47%, 30%, and 16% of the time, respectively [13]. A communication barrier is not noted in the DHS, likely reflecting the fact that these women were not asked if illiteracy or the ability to speak Swahili was a barrier and suggests that those responsible for the DHS may not have anticipated communication barriers.

Public health messaging about the benefits of receiving the recommended care may have no impact on behavior and be misguided if local perspectives are not considered. The local perspective that emerges here is that the illiterate women in Misungwi district can neither understand the verbal communication of their healthcare providers nor can they read the information provided by them. Notably, these women do have health literacy and do appreciate the need to visit the community clinic. They know what they should do but gain little in attempts to access healthcare services.

Given that there are 128 languages in use in Tanzania, it is unrealistic to expect healthcare providers to learn the languages of each community they serve. Hence, in the short term, the needs of this at-risk population may be met by providing 1) a protocol for CHWs and healthcare providers to identify those who are illiterate or cannot speak Swahili; 2) a local individual with training in confidentiality or a cell-phone application to translate for those unable to speak Swahili; and 3) graphic public health messaging that does not require literacy. In the long term, this barrier may be overcome by ensuring that all Tanzanians have access to an education in which they would learn to read, write, and speak Swahili. A failure to address the obstacles to maternal care for these illiterate women will likely ensure that the maternal mortality ratio and neonatal mortality rate remain high in this region.