Factors influencing utilisation of maternal health services by adolescent mothers in Low-and middle-income countries: a systematic review

Eighteen Percent of the world population are adolescents, defined as individuals aged 10–19 years [1, 2]. Generally, the global discourse lays emphasis on adolescents aged 15–19 years as they fall within the broader reproductive age group (15–49 years) [3]. About 16 million girls within this 15–19 age group give birth every year, of which 95% of the births occur in low-and middle-income countries (LMICs) [4]. Girls aged 15–19 years contribute to 12% of global annual births however also make up 10% of global annual maternal deaths [4, 5]. Globally, complications during pregnancy and childbirth are the second leading cause of death amongst girls aged 15–19 years old [6]. Recent estimates from 144 countries suggests that adolescents between 15 and 19 years are about one and a half times more likely to die during childbirth when compared with women aged between 20 and 24 years [6], who are relatively better physiologically prepared for pregnancy and childbirth [7]. Ninety-nine percent of these adolescent maternal deaths occur in LMICs (82% occurring in just 20 countries) [6]. About three million girls within this age group undergo unsafe abortions every year, further contributing to these adolescent maternal deaths [8]. For those who survive pregnancy, evidence shows that they have higher risks for postpartum bleeding [9], anaemia, pre-eclampsia and other problems of pregnancy [10, 11]. They also have a higher risk of developing obstetric fistula [12].

Adolescent mothers are not only challenged by the physical threats to their health, as described above, but are also often socially disadvantaged. Many have to raise their babies as single parents, are unable to complete their education and consequently have a limited capacity to secure a job and sustain a livelihood to support themselves and their children [8]. Adolescent mothers have to deal with all these issues while still going through ‘adolescence’ with all its challenges as well as adapting to the maternal role concurrently [13‐15].

Furthermore, the health of babies born to adolescent mothers is also at risk as these babies are more prone to preterm delivery, low-birth-weight and of dying as infants compared to those born to 20–24 year-old mothers [9, 16, 17]. Particularly in LMICs, babies born to adolescent mothers face a 50% higher risk of being still born or dying in their first few weeks of life when compared to babies born to mothers between ages 20 and 29 years [8].

These vulnerabilities have been highlighted more recently in the development of the post-2015 agenda, as advocacy for more focus on health of adolescent girls, who have been described as being “left behind” in the era of the Millennium Development Goals (MDGs) has increased [18]. It is well established that utilisation of maternal health services (MHS) across the continuum of care, that is, antenatal, intra-partum (by skilled birth attendants) and postpartum care are critical in reducing pregnancy-related morbidities, decreasing maternal mortality of adolescent mothers and improving outcomes, survival, quality of life and health of their babies [19]. We argue that to better fulfil the promise of the sustainable development goals (SDGs) for adolescent girls during the post-2015 era, strategies that focus on preventing early marriage and early childbearing [20] must be complemented by more research that aims to better understand MHS utilisation patterns of adolescent who become pregnant. Such research would be critical in ensuring that the service needs of this vulnerable group are met.

Therefore, we conducted this systematic review of the literature to explore factors that have been found to influence adolescent utilisation of these life-saving MHS across LMICs, where the burden is greatest. Key questions that we aimed to answer were: How has MHS utilisation by adolescent mothers been assessed? And what are the factors affecting utilisation of MHS by adolescent mothers?

We used the PRISMA approach [21] to report findings of this systematic review on factors influencing utilisation of MHS by adolescent mothers in LMICs [22].

In this results section, we present a summary of search results, quality assessment results, distribution of studies that assess adolescent maternal health services in LMICs, characteristics of the included studies, and results of the thematic summary.

This systematic review mapped out the assessment of factors influencing adolescent MHS utilisation in LMICs, highlighting the distribution, quality and characteristics of studies that focus on this limited area of research. The review identified commonly used predictor variables in the assessment of adolescent MHS utilisation and predictor variables that have been shown to be significant, including the strength of their significance. The review also showed some evidence that there is poor utilisation by adolescent mothers compared to older mothers.

This review needs to be interpreted carefully, bearing in mind some of its limitations. Firstly, the search was limited to articles published in English language, as such, papers from developing Latin America and Francophone Africa countries may have been missed out. Secondly, the same group of researchers authored six out of the 14 included articles over a period of 3 years [34, 35, 40‐43]. This similarity in author profile could affect the conclusions that we reach, because of the potential for the authors to make similar decisions and processes in the conduct of their research.

Despite overwhelming evidence suggesting that adolescents mothers are uniquely different from the general women’s population and that they are a particularly vulnerable and deprived population predisposed to worse maternal health outcomes compared to older age group women [6, 8‐12, 44, 45], findings of our review show that there is limited number of studies published in the area of adolescent utilisation of critical MHS. Ten years since the first adolescent MHS utilisation study was published in 2006, it appears that there remains minimal interest in the topic. This becomes even more apparent when a comparison is made with the plethora of research that has been conducted on utilisation of MHS amongst older age group women [46‐67]. The reason for the low focus on a vulnerable group like adolescent mothers is not particularly clear, but may not be unconnected to the inherent challenges in collecting data from this cohort. Firstly, data from demographic health surveys in several LMICs suggest that fertility rates amongst adolescents are lower than in women in their twenties and early thirties, as such the ‘chance’ of finding adolescent mothers for age-specific surveys are lower compared to mothers in their twenties [68‐72]. Secondly, the issue of adolescents getting pregnant remain a culturally complex one in many LMICs [73] and as such capacity to survey sufficiently large numbers for sensible analysis may be further complicated, due to barriers such as lack of consent, shame of the adolescent mother for having a baby and her lack of power to have a conversation on such matters with a ‘stranger’ [73].

When studies have been published, our review points to the need to address some quality issues in under to improve reports on MHS utilisation assessments amongst adolescents. In line with best practices [28], authors need to ensure that they describe the management of bias, missing data and discuss limitations of their study. In addition, as these assessments mainly constitute observational studies, there is a need to highlight percentage utilisation data of adolescent MHS utilisation before presenting factors influencing utilisation. Only two studies did this in our review [35, 40].

All studies included in our study [30‐43] used quantitative research methods in assessing MHS utilisation amongst adolescents. No study used qualitative research methods. Qualitative methods have been used extensively in healthcare [74, 75] and they offer a unique opportunity for researchers to be able to answer the “why” and not just the “what” [76]. Particularly as it relates to adolescents, there are several “why’s” that would need to be answered before effective strategies to improve their MHS utilisation can be implemented. In addition, qualitative methods may provide a more confidential platform for adolescents to discuss this sensitive topic. We believe that there is significant merit in supplementing survey-based approaches, using quantitative methods with qualitative methods for getting a better understanding of the challenges and other factors influencing adolescent mother’s care seeking patterns in different contexts. Use of such mixed methods approaches would provide the holistic perspective required for a broader understanding of adolescent MHS utilisation [77].

Only one study [32] collected primary evidence to assess MHS utilisation of adolescents. The remaining 13 studies [30, 31, 33‐43] used different country-specific secondary data sources like the DHS. The DHS series are generally well renowned for their robustness and quality [78, 79]. However, there has to be some concern about the time lag between the date of publication of the DHS datasets and the date that researchers analyse them. This is particularly important especially if such analyses are to be relevant for ‘up-to-date’ policy-making. Four country assessments were based on data that was over 10 years old already at the time of analysis and 38 country assessments were based on data between 5 and 10 years old. The reality is that datasets for sub-set (adolescent population) analyses, like that of the DHS, are not immediately available following completion of the primary survey that generated the data. This may be the reason for the delay in subsequent secondary analyses. Following such delays, the relevance of findings from these secondary analyses may be called to question, specifically for adolescents, who continue to change from generation to generation, even in the space of 10 years. The needs, aspirations and characteristics of Generation X are different from Y and so are the needs of Generation Y entirely different from Generation Z [80, 81]. Similarly, the factors that influence MHS utilisation may be different amongst adolescents across generations. It appears that when such considerably wide time interval between dataset availability and analysis is the case, then the adolescents from whom the data had been collected are not the same for whom planning and policy choices are required.

From our findings, there also appears to be lots of focus on adolescent mothers who are already married [30, 32, 34, 35, 37, 40‐43], ignoring the unmarried ones, who may be in even more precarious situations to be able to access MHS if they got pregnant [82]. This exclusion of unmarried adolescent mothers may in itself lead to some form of selection bias [83], thereby skewing results and affecting the interpretation of findings. The reason for the focus on married adolescent mothers is not too clear, but it may not be unconnected to possibly low numbers of unmarried adolescent mothers recruited in the primary surveys that were conducted to provide the datasets that the authors used for analysis. Secondly, some of the original surveys from which secondary analyses were subsequently conducted only collected data from within family settings that had married women [35, 37, 41, 42].

Excluding the multi-country studies, only two countries, Bangladesh [36, 38, 39] and India [37, 41] have had the same data source used for analysis on adolescent MHS utilisation multiple times. However, in both countries, the assessments used the same dataset for analysis, yet conclusions were not the same, regarding what factors were found to be significant. This, therefore, calls into question the quality of the analyses being done and highlights the need for more careful analysis and verification of findings. Also, we observed that even within the same countries, selection of predictor variables for consideration was not consistent. Our opinion is that selection of predictor variables for adolescent MHS utilisation must be based on the availability of reliable data, consideration for peculiarities of the specific setting and insight from literature focused on research conducted in similar settings.

With education of the adolescent mother being reported as statistically significant for MHS utilisation in all surveyed countries (except Malawi), there is a case for focusing on broader girl child education strategies. Similarly, education of the husband was reported to be significant in seven studies, and with the influence of the husband reported to be the most influential in making adolescent mothers use MHS [32], it is critical to include men to increase uptake of MHS by adolescents.

Findings from our review suggest that adolescent mothers are more likely to utilize MHS for their first pregnancy/delivery, but less likely to utilize MHS when they have more children [30, 34‐36, 40‐43]. There is, therefore, a need to make adolescents more aware of the additional risks that they face in pregnancy because of their ‘adolescence’. Our review suggests that there is an opportunity to leverage ANC attendance as a platform for advocacy to encourage and stimulate subsequent SBA utilisation by adolescents, especially as all five studies in our review that considered ANC utilisation as a predictor variable reported it as significant for SBA and PNC utilisation [30, 34, 39, 40, 42], which interphases with arguably the most critical period of the entire pregnancy for adolescents - delivery. The World Health Organization recommends that health care providers should be “seizing the opportunities” that patient engagements like ANC provide [84]. Evidence from the literature shows that ANC offers an opportunity to sensitize adolescent mothers about utilisation of MHS and promote healthy lifestyles that could potentially improve long-term health outcomes for them and their yet unborn child [19, 85]. For example, family planning counselling could be integrated into ANC, continued as part of PNC and this could potentially have a positive impact on the adolescent’s use of contraception after delivery. It is also a platform to implement a birth preparedness plan, ensuring that adolescent mothers can be better prepared for the birth itself including identifying the closest facility to manage them in the case of complications. However, this integration of services needs to be achieved, without overloading the already stretched workforce in many LMICs as well as providing an inclusive service for both married and unmarried adolescents [86].

Seven out of nine studies that looked at media exposure as a predictor variable, mass media exposure was found to be statistically significant. Going forward, with the proliferation of access to social media of young people globally, including in LMICs [87], ‘access to social media’ needs to be considered as a variable to be explored. We also opine that there is an opportunity to conduct research via electronic data collection, even via social media in order to target more adolescents, who otherwise will be uncomfortable talking to adults openly about their pregnancy etc. On the outcome side (MHS utilisation), while it is straightforward to report outcome measures such as attended ANC or not or attended PNC or not, there is need to capture indicators that describe the quality of care that adolescents also receive across the whole continuum of care. We note that four of the 14 included studies [30, 35, 37, 41], all conducted in India, actually reviewed whether adolescent mothers received Tetanus toxoid injection, folic acid and iron tablets. This is particularly important for adolescents because of their higher risk for poor maternal health outcomes. For them, it is not just about utilizing the services, but more about how well the services have been utilised.

No article was retrieved that assessed impact of intervention(s) in increasing MHS utilisation amongst adolescents. However, there have been many studies that reviewed the effectiveness of strategies in the wider women of reproductive age group, as evidenced in this recent systematic review [88]. More recently, another systematic review published in 2014, assessed the impact of user fees on MHS utilisation for all women [89]. To ensure that appropriate interventions are being properly targeted at increasing adolescent MHS utilisation, there is a need to build on the needed evidence to base decisions upon.

Even when broader age groups are being researched, it is critical to highlight adolescent mothers and conduct some form of subset analysis of adolescent mothers, because of their afore-described peculiarities. In our review, four studies did this [31‐33, 37]. Such disaggregation of data is critical for planning and for better understanding and design of health systems. More recently, there have been global calls for presenting disaggregated data to ensure that inequities may be better highlighted [90], as may be the case with adolescent health MHS utilisation when compared to older women. In addition, such data may be able to support ‘business case’ development for the need to focus on adolescent MHS utilisation.

Clearly, there are notable similarities between countries with regards to factors that affect adolescent MHS utilisation, especially maternal education and wealth index. Emphasis thus needs to be placed on educating girls and ensuring that financial barriers do not limit their access to critical care. However, there may be some context-specific factors in different countries, which need to be considered when designing interventions aimed at improving adolescent MHS utilisation. This study highlights the need for more robust evidence on how to achieve this. We need innovative approaches that incorporate both real-time quantitative and qualitative research methods in studying access, utilisation and quality of MHS for adolescent within specific settings. These studies should include ‘all adolescents’ and not the ‘easy to capture’ adolescents [18]. This will bridge the equity gap and promote universal health coverage.

Increasing access to and utilisation of quality MHS for adolescents especially in the 20 countries responsible for 82% of global adolescent maternal deaths [6], will contribute significantly to a reduction in maternal mortality. Efforts geared towards improving maternal health care for adolescents are consistent with the SDGs, which also focus on girl child education, preventing early pregnancy and removing financial barriers to care [91]. One thing we cannot afford to do again in the post-2015 era is to “leave them behind”.