The associations between women who are immigrants, refugees, or asylum seekers, access to universal healthcare, and the timely uptake of antenatal care: A systematic review

The World Health Organization (WHO) recommends that antenatal care (ANC) commence before 12 weeks' gestation to reduce the risk of obstetric and perinatal complications. Immigrants, refugees, and asylum seekers are at higher risk for late or non‐initiation of ANC, and exclusion from universal healthcare (UHC) may be a contributing factor.


INTRODUCTION
Antenatal care (ANC) is the care provided to a woman throughout the course of her pregnancy by experienced and skilled health professionals. It includes health education and promotion, fetal growth and wellbeing assessment, routine screening, and management of any medical or obstetric complications. 1 The World Health Organization (WHO) recommends that the first antenatal appointment be booked no later than 12 weeks' gestation, as ANC commencing in the first trimester is associated with better perinatal outcomes for women and babies. 1 In addition, delayed ANC has been correlated with maternal complications, such as anaemia, hypertensive disease, gestational diabetes (GDM), preterm birth (PTB), emergency caesarean section, instrumental vaginal birth (IVB), and neonatal complications. 2,3 Therefore, timely initiation of ANC is recommended for the prevention and reduction of maternal and fetal morbidity.
Women who are immigrants, refugees, and asylum seekers have been identified to attend later, on average, to ANC compared to their receiving-country-born counterparts. 4 Migrants can be defined as 'persons who change his or her location of residence, irrespective of the reason for migration or legal status'. 5 Refugees or asylum seekers, on the contrary, are 'displaced persons who require international protection for fear of persecution, conflict, violence, or other circumstances that seriously disturb public order'. 5 Therefore, refugees, and asylum seekers may attend later to ANC because of their disadvantageous social and political determinants of health, whereas migrants may attend later due to other reasons, such as difference in the cultural and normative values of healthcare, lack of knowledge regarding the receiving country's healthcare system, communication barriers, and employment barriers. 4,6 These challenges are slowly being addressed through the implementation of culturally appropriate health promotion by governments, more readily available interpreter services, and increasing employment of multicultural healthcare workers. 6,7 Substantial hurdles persist to accessing healthcare for women who seek asylum in another country due to their marginalisation by the receiving country. That is, women who do not hold citizenship or permanent resident status are usually ineligible for non-emergency healthcare under their residing country's health scheme. 8 Multiple countries employ a universal healthcare (UHC) system, and notably, the USA is the only country out of 33 highincome countries that does not provide UHC. 9 The WHO defines UHC as 'ensuring that all people have access to needed health services (including prevention, promotion, treatment, rehabilitation and palliation) of sufficient quality to be effective while also ensuring that the use of these services does not expose the user to financial hardship'. 10 Arguably, the most significant gap in UHC coverage is that it frequently fails to include immigrants, refugees, and asylum seekers who are often vulnerable by virtue of their social complexity. 8 Therefore, poor health outcomes can be further compounded in this population. 11,12 Therefore, it could be contended that many countries' versions of UHC are not, in fact, true UHC, as it prohibits a proportion of the population from accessing affordable medical care. This notion has been discussed at length in human rights literature, with some arguing that denying access to healthcare for immigrants, both legal and illegal, at a national level is of substantial ethical and humanitarian concern. 8,13 Within the maternal health scope, inadequate access to UHC is correlated to lower ANC uptake, as ANC is simply a specialised type of healthcare. 4 Current literature suggests that a lack of access to UHC is a major deterrent to presenting to primary health providers and hospitals. 11,14 In Australia, pregnant women who are immigrants and experience difficulties accessing Medicare may need to personally fund their own ANC. The cost for this service can be anything up to AUD $16,000, which covers routine ANC appointments and the birth itself. Treatment for obstetric complications, neonatal complications, or postpartum follow-up, however, incurs additional fees. 15,16 As a result, many women delay, limit, or forego maternity care altogether, as the financial burden is too great.
Moreover, although several papers have commented on limited access to UHC as a key issue, 8,13,17 a thorough review of the literature has not yet been performed to fully assess the degree of inequity of access and its relationship to delayed initiation of ANC.
We hypothesised that women who are immigrants, refugees, and asylum seekers would present later to ANC and have a greater risk of pregnancy complications.

AIMS AND OBJECTIVES
This systematic review aims to provide a narrative synthesis of the evidence regarding the relationship between access to UHC by immigrants, refugees, and asylum seekers and routine ANC uptake.
Furthermore, our study seeks to investigate the downstream effects that barriers to UHC have on maternal and neonatal health, through its evaluation of secondary outcomes, including maternal and fetal health outcomes. Finally, our study endeavours to add to the research and literature which aims to promote the basic health rights of those who currently do not have access to UHC and to evaluate whether perinatal outcomes are affected in women who are immigrants, refugees, and asylum seekers.
With the global migrant population estimated at 272 million and increasing, 18 particularly due to increasing civil conflicts, political turmoil and economic instability, the right to healthcare for this population has never been of greater importance.

Registration
Our review was reported in accordance with the MOOSE guidelines for meta-analyses and systematic reviews of observational studies. 19 This systematic review is registered with PROSPERO, registration ID: CRD42021225242.

Study design
A systematic review was conducted to synthesise the current available literature on this topic.

Search strategy
The PICO approach guided the development of a systematic search strategy: • P: pregnant women were immigrants, refugees, or asylum seekers residing in a country with UHC.
• I: lack of access or significant barriers to accessing healthcare in the receiving country.
• C: pregnant women who were citizens of the receiving country who were eligible for UHC.

Study selection
The systematic review included observational cohort studies.
Articles were first screened by title and abstract against our inclusion and exclusion criteria. For inclusion in the review, studies must have reported the time taken to initiate ANC in pregnant women who were immigrants, refugees, or asylum seekers who did not have access to healthcare or had substantial barriers to healthcare and were living in a country with an established UHC system. The full text of these potentially eligible papers was then appraised for final inclusion in our study by the principal investigator (PI) (S.L.S.) and either of the two co-investigators (K.M.L. or K.L.S.). Where there was a discrepancy in the decision to include the study in the final review, agreement was reached by discussion and consensus with all three investigators.

Data collection/extraction and data items
Data were extracted using a data extraction form template (Appendix S2), with the outcomes for main analysis adapted to match the outcomes of our review. Data extraction was performed on 12 articles by the PI (S.L.S.) and either of the two co-investigators (K.M.L. or K.L.S.). Extracted data included study design, study population and setting, intervention, comparison, outcomes, risk of bias (RoB), timing of ANC, demographical characteristics, and maternal and neonatal outcomes.

Risk of bias
Risk of bias was assessed using the ROBINS-I tool for nonrandomised studies. 31 This was completed based on the following categories: confounding bias, inclusion bias, bias due to classification of interventions, missing data bias, outcome measure bias, and selective reporting bias. The methods used to determine bias in each criterion, as well as overall bias, can be found in the ROBINS-I tool. The studies were then graded as low, moderate, or high for RoB per criterion and for overall bias.

Summary measures
The odds ratio (OR) or relative risk (RR) measures with corresponding 95% confidence intervals and P-values were recorded where available, as reported by the study authors.

Synthesis of results
Meta-analysis was not performed on the data due to the heterogeneous nature of the studies. Data were reported narratively.
Key information from each article, including country of origin, participants, intervention, comparator and outcomes, was tabulated.
Primary outcomes were also tabulated and then reported on intext. Similarly, secondary outcomes were discussed in text.

Study inclusion
Twelve articles were included in the final review (see Fig. 1).
Individual study characteristics are summarised in Table 1. Table 2 provides a summary of the RoB in each study. Low overall risk inferred a low risk in each criterion. Moderate overall risk was determined by moderate risk in at least one criterion, without high risk in any criteria. High overall risk was determined by high risk in at least one criterion.

Primary outcomes
Each included study reported that women who were immigrants, refugees, or asylum seekers were more likely to present late for ANC compared to receiving-country-born women (see Table 3), and of eight studies which reported a P-value, all reported the outcome to be statistically significant between groups (P < 0.05). [32][33][34][35][36][37][38][39]

Demographical outcomes
Eleven of 12 studies reported demographical data. Nine studies reported on maternal age, with no statistically significant difference identified between groups in any studies. Only two studies reported on maternal BMI. Råssjö and colleagues 36 reported BMI categorically and found receiving-country-born women more than twice as likely to be underweight compared to women who were immigrants (OR: 2.39, 95% CI [1.30-4.37], P < 0.05). However, Vanneste and colleagues 38 reported no statistically significant difference in BMI between groups (P > 0.05).
Eight studies reported on maternal ethnicity. In all of these studies, the control group comprised of women with ethnicity of the country of asylum. The ethnicities of the immigrant groups were diverse, with ethnicities including those from parts of Europe, 33,35,39,40 Asia, 40,41 the Middle East, 37,42 Africa, 36,40 the Pacific Islands, 32 and the Americas. 34,41 Three studies reported on maternal level of education. Each of these studies found that receiving-country women were more likely to have had a higher level of education compared to women who were immigrants (P < 0.05). Five studies reported on parity,

Records excluded (n = 233)
Full-text articles assessed for eligibility (n = 31) Full-text articles excluded, with reasons (n = 7) Study from a country without UHC.

TABLE 1 (Continued)
with over half of participants in each of these studies being primiparous. However, no P-value was given for this outcome in any of these studies.

Maternal outcomes
Eight of 12 included studies reported on maternal outcomes.
Five of these reported on the incidence of GDM. Two studies found that women who were immigrants were more than twice as likely to develop GDM compared to receiving-country-born Two studies reported on prolonged second stage of labour.

Neonatal outcomes
Neonatal outcomes were seldom reported in the studies included in this review; only five of 12 studies reported on these, with no single study reporting on more than two of the four outcomes screened for in the review.
Wilson-Mitchell and Rummens 41 found no statistically significant difference among breastfeeding rates at discharge between women who were immigrants and the control group (91.02 vs 86.45%, respectively, χ 2 = 2.077, P > 0.05).
Two studies reported on neonatal intensive care unit (NICU) and special care nursery (SCN) admission. de Jonge and One study reported on respiratory distress (RD), 35 with neonates born to women who were immigrants more likely to experience RD than neonates born to receiving-country-born women (11.8 vs 6.6%, respectively, P < 0.05).
Apgar scores less than seven at 5 min were reported by three studies. Both Liu and colleagues 42

DISCUSSION
Access to UHC by immigrants, refugees, and asylum seekers is of international concern. Pregnancy is a time of particularly increased vulnerability in this population, and access to timely ANC is essential to both mother and fetus.
Our study has shown that compared to receiving-country-born women, immigrants, refugees, and asylum seekers consistently present later for ANC, which does not meet the WHO recommendations. Therefore, immigrants, refugees, and asylum seekers should be considered at higher risk of poor pregnancy outcomes. The link between delayed ANC and pregnancy outcomes, as observed in our secondary outcomes, remains inconclusive, due to the het- in Germany, undocumented migrants who present for any type of non-emergent care will almost certainly face deportation.
Moreover, doctors can also be punished by law for providing care to those without documentation. 46 In Canada, if payments are not received in full upfront, patients can similarly face deportation. 44 The UK simply has a blanket rule that states that anyone without insurance must pay in full for non-emergency care. 47  Agreement. 51 However, immigrants are often not aware of their entitlements. This highlights how unfamiliarity within a foreign healthcare system ultimately leads to delayed ANC uptake.
Our review showed that regardless of the laws which governed immigrants, women who were immigrants, refugees, or asylum seekers appeared equally as likely to present later for ANC as their receiving-country-born counterparts. This suggests that actual versus perceived exclusion from UHC leaves patients as equally disadvantaged in the context of ANC. It should be noted that our sample size was small, with only 12 studies included in the review.
Further research should be conducted to explore the degree to which perceived and actual exclusion from UHC worsens patient outcomes in the maternal health sphere.

Limitations
Our review has some internal bias, with many studies ranked as 'moderate' as per the ROBINS-I tool, most commonly due to moderate confounding bias. [32][33][34]36,38,39,41 Social confounders such as employment, income and education are known independent risk factors for delayed ANC. Despite this, other studies demonstrated that ANC care was still delayed in women without access to UHC, even after confounders were adjusted for. 35,37,40,42

Future directions
Our study underlines the necessity to promote the basic health rights of women who are undocumented, or have refugee or asylum seeker status, and face substantial inequities in accessing healthcare. Our objective is that the review's findings will provide evidence to inform policy adjustment and guide the development of programs to provide access to all to achieve WHO's Sustainable Development Goal of equitable access to sexual and reproductive health for all by 2030. 10 Future research should be conducted on how elements of UHC can be made more accessible to this population.
The impact of delayed ANC due to improper UHC on subsequent maternal and neonatal outcomes remains uncertain, as metaanalysis could not be conducted. Additional research is required to make conclusions on the effect of such ANC. Larger studies should be conducted prospectively with research outcomes agreed a priori, leading to a wider collection of common outcome measures.

CONCLUSIONS
In summary, women who are immigrants, refugees, and asylum seekers who are excluded from UHC, whether actual or perceived, present later to ANC than their receiving-country-born counterparts with full UHC access. This phenomenon was observed in nine countries with UHC systems. However, the effect of delayed ANC on pregnancy outcomes in this population remains uncertain due to heterogeneity of studies. Further prospective research is required to direct policy and services to address issues of inequity.