Interventions to Improve the Reproductive Health of Undocumented Female Migrants and Refugees in Protracted Situations: A Systematic Review

Evidence from a systematic review shows that subsidizing health care, strengthening health services, and implementing educational interventions have positive effects on undocumented migrant women and female refugees’ sexual and reproductive health outcomes.


INTRODUCTION
I nternational migration is a global and growing phe- nomenon.Its implications for health systems include challenges to ensuring necessary health and social services in transit and destination countries.It is estimated that in 2020, there were a total of 281 million international migrants worldwide (3.6% of the total population). 1From 2000 to 2020, the number of refugees has almost doubled, from 14 million to 26.4 million. 1In 2020, 48.3% of international migrants were women and girls. 1 Undocumented migrants and refugees in protracted situations find themselves transiting locations without meeting the requirements that a sovereign state has to allow them to reside, work, and access services in its territory.Women of these populations are often isolated, without documents or support networks, and subject to gender-based violence, which is often perpetrated by their partners and family members, migration authorities, and smugglers.Living in a country that is not their own, female migrants and refugees encounter legal frameworks, discrimination, and language barriers that make it difficult for them to access basic needs such as education, health services, or humanitarian resources. 2,3n most countries, women who lack a regular migratory status have limited access to health services; if they are on the move, the continuum of care provision can be further restricted because of their short stay in a single geographic setting.This population is often characterized by low socioeconomic opportunity, generalized insecurity, and precarious legal status.Barriers to health care access are accompanied by unequal reproductive health (RH) outcomes among undocumented migrant women and girls and refugees in protracted situations.5][6] Undocumented migrant women and female refugees are also at increased risk of unintended or forced pregnancies, rape, sexually transmitted infections, and unsafe abortions. 7Some of the factors identified that negatively affect access to RH for this population include fear of abuse and discrimination by health professionals, limited mobility, lack of transportation and distance to the health services, and lack of adequately trained and sensitized health providers. 8Therefore, the intersections between migration, women, and health requires special attention to comprehensive RH service delivery, with a focus on short-and long-term contraceptive methods, sexual violence support services, abortion information and counseling and postabortion care, and antenatal and childbirth care.
0][11][12] How many people benefit and how they are cared for is regularly documented, but the results obtained are hardly systematized and subject to verification.Moreover, RH interventions are even less frequently published in peer-reviewed journals.
To address this knowledge gap, we conducted a global systematic review to identify the outcomes of intervention studies focused on improving the RH of undocumented female migrants and refugees in protracted situations.We synthesized research that highlights RH outcomes and intervention components that have been tested among undocumented migrant women and refugees in protracted situations, regardless of whether the intervention was described as successful or unsuccessful, to provide evidence and recommendations for governmental and nongovernmental organizations (NGOs) to support their goals of improving the well-being of these populations.
This study contributes to understanding how to address the RH service needs of undocumented female migrants and/or refugees in protracted situations by exploring interventions and unveiling complexities at the intersections between migration and health.These intersections are of particular interest to global and national stakeholders committed to upholding international human rights frameworks and achieving the Sustainable Development Goals, which aim to end poverty, protect the planet, and ensure prosperity and well-being for all.

METHODS
We conducted a systematic literature review of intervention studies on RH following the Preferred Reporting Items for Systematic Reviews and Meta-Analysis (PRISMA) Statement guidelines.Because we aimed to identify the characteristics and impact of interventions, we included both qualitative and quantitative research in our analysis.Table 1 details the study's inclusion and exclusion criteria.
Following the 2019 Glossary on Migration of the International Organization for Migration, we defined an undocumented female migrant as a female nonnational who enters or stays in a country without the appropriate documentation.
We defined a female refugee using the 1951 Refugee Convention definition 13 : a person who, owing to a well-founded fear of persecution for reasons of race, religion, nationality, membership of a particular social group or political opinions, is outside the country of her nationality and is unable or, owing to such fear, is unwilling to avail herself of the protection of that country.
We defined a protracted situation as a situation in which refugees have been unable to return to their habitual residence for a long period of time and where the process for finding durable solutions, such as repatriation, integration in host countries, settlement in third locations, or other mobility opportunities has stalled. 13e decided to include and focus on these populations, given their limited access to health care because of a lack of documentation and living and housing conditions (e.g., refugee camps, informal settlements, or migrant shelters).There is a lack of consensus on when the refugee situation is considered prolonged.Some organizations and stakeholders emphasize the duration of displacement (e.g., 12 months or 3 years) as a reference point.Others consider that the situation is prolonged if people are not able to return to their country of origin, so their location is used as the main criterion.Finally, others identify a continued need for humanitarian action as the key element of the definition.For the present study, we defined refugees in protracted situations encompassing all situations in which the process of finding solutions has stalled. 13e defined RH as 14 : a state of complete physical, mental and social-wellbeing and not merely the absence of disease or infirmity, in all matters relating to the reproductive system and to its functions and processes.
To focus the review's scope on answerable research questions related to RH interventions, we refrained from broadening our search to genderbased and sexual violence services.Because of the high prevalence and long-term consequences of gender-based and sexual violence among undocumented female migrants and refugees in protracted situations, we believe that separate systematic reviews on these issues are needed.We excluded studies that

Search Strategy
We used 6 search engines specialized in health and social sciences: Cochrane Library, MedLine (via PubMed), BIREME, PsycNet, SCIELO, and Web of Science.Screening the reference list of relevant systematic reviews as well as an intentional search in Google complemented the search strategy.Boolean operators were used for those search engines that allowed it, as well as the corresponding filters by type of study, date of publication, and language (English, Spanish, and Portuguese).We used 2 search strategies for every search engine (Figure 1).Each search strategy included a combination of terms related to population, intervention, and outcome.Keywords with high sensitivity and low specificity, such as women, contraception, or contraceptive, were removed after several search iterations to identify relevant terms in each search strategy.We added the term asylum as part of our search strategy to capture some of the articles about undocumented migrants and refugees in protracted situations.However, we did not include any article that solely focused on asylum seekers.Each word combination was conducted with the corresponding translation in each language.The searches were conducted independently by NL, LB, and LC.
In Google, we added the names of NGOs, academic institutions, and United Nations (UN) agencies known to work with the populations of interest to the previously selected search terms.The NGOs and UN agencies included in the search were the UN Refugee Agency, Médicins Sans Frontieres, Médicins du Monde, EMERGENCY, Marie Stopes International, Women's Refugee Commission, Inter-Agency Working Group on Reproductive Health in Crises, National Institutes of Health, and the Population Council.These organizations, as identified by the Population Council, have a history of providing health care services to migrant and refugee populations in humanitarian settings, documenting their interventions, and conducting monitoring and evaluation reports of their efforts.

Data Analysis
The research team used Microsoft Excel to download all returned citations from the searches and summarize data from the studies included in the analysis.All authors reviewed data screening and analysis.
Given the heterogeneity of interventions, designs, and outcomes of the studies that fit the inclusion criteria, we used a narrative synthesis methodology to analyze results.Studies included in the analysis were assessed for quality using checklists compatible with the study design and methodology: the Cochrane Collaboration Manual for randomized control trials, 16 the STROBE statement for observational studies, 17 and The Joanna Briggs Institute's Checklist for Qualitative Research. 18Eligible articles were given a quality score that was converted into the percentage of the total achievable score.As per previous reports, thresholds were set up to rate studies' quality as low (0%-33%), medium (34%-66%), or high (67% or above). 10Articles were divided between NL, LB, LC, SL, and LV, independently rated for quality, then reviewed in pairs.After peer discussion, if a consensus could not be reached, then a third reviewer rated the article.

RESULTS
We obtained a total of 21,453 citations, of which 17,419 corresponded to titles and abstracts obtained through specialized search engines, 114 through the intentional search in Google, and 3,920 through screening the references of relevant systematic reviews (Figure 2).Of the total citations, 8,118 were excluded due to double counting and 12,966 were excluded in the abstract screening for not meeting the inclusion criteria.In addition, using the search word "transit" yielded a large number of unfavorable results related to transport or digestive processes in medical sciences.We assessed 306 articles for eligibility for full-text reading, of which 273 were excluded for not meeting the population, intervention, and outcome criteria.Thirty-three articles were included for a full-text critical reading, 23 of which did not meet the inclusion criteria because they lacked a measure of comparison or impact estimate (n=22) or because the data were found on an NGO website, and it was not known whether the data went through an editorial review process (n=1). 19dditional reasons for exclusion were gaps in data analysis and other methodological weaknesses, such as small sample size (<30 participants) or insufficient data regarding the intervention.][22][23][24][25][26][27][28][29] All studies included were observational; there were no randomized or field trials. There were 5 quantitative, [20][21][22]26,28 3 qualitative, 24,25,27 and 2 mixed-methods studies.23,29 As shown in Table 2, 7 studies were rated high quality according to the measures used.20,21,[23][24][25]28,29 However, in most cases, the methodological design was not adequate to report impact results, as they were ex post facto studies and/or did not have a comparison group.The 10 studies were conducted in the United States, 20 Pakistan, 21,22 Lebanon, 23 Thailand, 24,25,29 Guinea, 26,28 and Uganda.27 Nine studies were conducted by teams working in refugee camps or by university-led research groups. The intervetions were diverse: 2 evaluated the improvement in access to RH services through the provision of public (Medicaid) 20 or private health insurance 21 ; 2 studies evaluated the impact of improving health infrastructure and staff capacities in border areas 22,23 ; 2 evaluated referral services and/or access to safe abortion 24,25 ; and 4 evaluated the impact of educational interventions, implemented either by trained personnel or peers, on increasing RH knowledge (Table 2).[26][27][28][29] Two financial support intervention studies done in the United States 20 and Pakistan 21 showed that enabling undocumented migrant and refugee women to obtain RH services for free or at low cost promotes RH service utilization. The stud in Pakistan showed that, compared to only making health services available, subsidizing health care through public or private insurance plans doubled the frequency of a woman having heard about contraceptive methods (44.9% vs. 88.9%), and increased approval of contraceptive use among women, their friends, and partners. 21 Likewse, the use of contraceptive methods increased by 29 percentage points in the group of women with access to subsidized health care, compared to the group of women who did not have access to health care subsidies.21 Similarly, the U.S. study showed that, compared to migrant women in Florida state where Medicaid was restricted for this population, the proportion of pregnant undocumented migrant women that attended prenatal care during the first trimester was 18 percentage points higher in the states of California and New York, where subsidized services through Medicaid were provided. 20 Inaddition, among migrant women with access to Medicaid, 22% more women received adequate prenatal services compared to those without access to Medicaid.20 Two studies conducted in Pakistan 22 and Lebanon 23 showed that health systems interventions that established or strengthened health service delivery structures targeting refugees or located in refugee areas generally had a positive impact on reproductive and maternal health.In Pakistan, the establishment of emergency obstetric care units combined with community-level efforts to educate refugee and undocumented migrant women and their families and facilitate links with the unit was related to a 41% increase in the number of women that attended at least 3 prenatal visits, a reduction in maternal mortality (291 vs. 102 per 100,000 women), and a 0.2% prevalence in maternal deaths after cesarean delivery.22 The prevalence of maternal deaths related to cesarean deliveries was consistent and within the <1% goal of the United Nations at that time. 22 In Lebanon, stengthening primary health care services in areas with a high population of refugees showed high RH service utilization among refugees; in some health areas, service use was higher among refugee women than local women in the same area (Table 2).23 This intervention is described as "supporting" facilities, but no further detail was provided.23 These 2 interventions occurred during a period of at least 3 years and included aspects of cultural and gender sensitivity.22,23 The 2 studies conducted in Thailand focused on facilitating access to safe abortion for migrant women.24,25 One evaluated a referral program to access free safe abortion services.24 The other evaluated an intervention with community health promoters, who provided counseling on safe abortion and access to a free supply of oral misoprostol.25 Both interventions showed positive results in terms of referral to services (64% of all women    who requested the referral were given one) 24 and effectiveness of the intervention (96% of women who accessed misoprostol had a complete abortion).25 Women expressed satisfaction with the initiative and a perception of increased empowerment and capacity to advocate for their RH rights after the intervention.One study informed that the intervention with oral misoprostol was safe and the side effects reported were tolerable. 25 The 2017 study describethat one-fifth of participants reported that they were denied abortion services mainly due to the subjective interpretation of health personnel on legal grounds. 24th interventions had an important subsidy component; the costs of medical supplies, transportation, and interpreters were absorbed by the implementing organizations and not by the women themselves.Four studies conducted in Guinea, 26,28 Uganda, 27 and Thailand 29 evaluated the influence of educational interventions on RH outcomes of migrant and refugee women.All but 1 of the studies included a qualitative component in the evaluation.28 Two educational interventions trained community promoters and implemented community outreach strategies.28,29 One included an educational campaign that raised awareness among health care workers and women (including refugees) on the prevention of neural tube defects with the intake of folic acid during the early stages of pregnancy.29 The other study trained refugee women to provide RH education, referrals, and contraceptives for women in their communities.28 The third followed a peer-education methodology to increase the knowledge of migrant women sex workers regarding human rights, life skills, advocacy, sex work, and sexual and RH.27 The last study evaluated the impact of RH literacy sessions on refugee women living in camps.26 These studies suggested variable effects of the interventions on knowledge and behaviors, with higher gains in the former.While 1 study showed a significant increase in knowledge among health care workers on the importance of folic acid intake, it did not observe a corresponding uptake in folic acid consumption by women.29 Another study identified a mean difference of 21.7% in the socialization with partners and other family members of knowledge acquired by women regarding safe sex. 26In contrast, there were no statistically significant differences in antenatal care seeking between women who participated in interventions and those who did not 28 nor in increased folic acid consumption practices.29 The 4 study findings suggested that educational interventions can increase knowledge, especially when led by peers, but knowledge shifts do not always translate into changes in practices or increased careseeking behaviors.Still, promoting safe spaces to learn about or discuss RH seems to provide a sense of self-confidence, 26 to inspire women to become advocates of their RH, 27 and is particularly relevant for stigmatized populations. 27 Amng migrant sex workers and transgender women, peer education helped address concerns of privacy and confidentiality, reporting increased access to information, management skills, and ability to discuss and disseminate knowledge with peers and clients alike.27

DISCUSSION
Our review of literature from the past 2 decades identified 10 observational studies that, according to our inclusion criteria, provided sufficient documentation and measures of intervention impacts on RH of migrant women and female refugees.][28][29] Promising approaches for future research and programming interventions in protracted situations have a focus on geographic proximity, are cost-sensitive, involve opportunities for peer sharing and learning, and are culturally appropriate.Additionally, interventions must recognize and address undocumented migrant concerns around legality and access to resources.
One reason for the limited number of studies evaluating intervention impacts on RH outcomes in refugee women is that most interventions were implemented in adverse political environments and contexts that may pose unique challenges for study purposes.
For example, in our study, of the 6 countries with reported studies, none favor broad and free access to comprehensive RH services for refugees and undocumented migrants.6][37][38][39][40] In some instances, the countries that temporarily host migrants or refugees are neighboring states to those in conflict and are susceptible to forced migration implications (e.g., Ghana, Lebanon, and Pakistan).These unstable circumstances directly affect the implementation of interventions and the collection of data required for research and evaluation. 26,28Nonetheless, some studies indicate the possibility of documenting the impact of intervention through quantitative and qualitative study designs.Future implementation research should consider both what is pragmatic and flexible in protracted migration settings.
Findings from this systematic review suggest that providing free or low-cost services to migrant women and female refugees in protracted situations is a critical strategy in successfully ensuring access to the continuum of sexual and reproductive, maternal, and newborn care services.The findings also concur with other studies documenting increased use of contraceptive methods and prenatal care following interventions. 6Studies with nonmigrant populations similarly report that increased enrollment in economic transfer programs or a reduced cost of services results in increased utilization of reproductive and prenatal services. 41,42Subsidies reduce financial barriers to accessing health services and necessary medical supplies, 43,44 as well as potentially motivate migrants who may otherwise hesitate to spend out-of-pocket given the precarious socioeconomic circumstances that they are in.Also, when the subsidy is public, undocumented migrants are provided with conditions similar to those of citizens, facilitating migrant women's ability to exercise their right to RH care. 45eview findings coincide with other studies around the need to strengthen health systems structures to ensure access to available health services among refugee and undocumented migrant women.Staffing health care units with culturally competent and trained personnel influences the prevalence of contraception use and reduces maternal and neonatal complications and adverse outcomes. 11,12,46,47Studies included in our review do not adequately describe aspects important for the replication of interventions.For example, the establishment of a unit of care entails hiring qualified, trained, and certified personnel; thus, studies should document financial considerations and mechanisms for sustainability.A study in Tanzania with a nonmigrant population identified a median per-patient cost of US$290 across 6 evaluated facilities.The 2 items associated with the highest expenditure were the hiring of personnel and the purchase of equipment (32% and 28% of the total budget, respectively). 48,49A mention of the minimum budget required to implement and maintain these interventions in border areas and/ or with refugee and migrant populations would be helpful to those aiming at replicating specified intervention dimensions.Even though the budget to be allocated in each country would be different, a good estimation exercise would consist of observing the proportion of the budget allocated to RH services compared to the total expenditure on health in the countries that have already shown favorable impacts on RH outcomes among undocumented female migrants and female refugees in protracted situations.
Our review of access to comprehensive RH services, such as abortion information and counseling and postabortion care, aligns with findings of studies with nonmigrant populations regarding the safety and effectiveness of involving trained lay health care workers as facilitators of access, 50 particularly for migrant women who face similar legal RH restrictions as reported by nonmigrant women but have additional concerns such as deportation and needs such as understanding and navigating the host country's health system and overcoming language barriers and cultural differences regarding service provision. 24In addition, the relevance of establishing a referral system to enable health care facility access to comprehensive RH services is implicit.Tousaw et al. is an example of the feasibility of establishing referral systems for accessing services that are not always readily accessible to migrant women. 24Future studies on interventions facilitating refugees' and undocumented migrants' access to comprehensive RH services through lay workers, however, should discuss practical, ethical, and legal specificities while training personnel, as any attempt to replicate these initiatives will benefit from such considerations.
Although we acknowledge that educationfocused interventions are important to empower women to act on the knowledge they have gained, our review points to educational interventions as potentially having a greater impact if implemented after or concurrently with health system capacity improvements (e.g., infrastructure and specialized staff).According to the results of this systematic review, interventions aimed at increasing RH knowledge raise awareness but do not necessarily influence behavioral change. 26,27,29Interventions that link and conduct follow-up with service users whenever health care capacities are already installed show better results in RH outcomes of migrant and refugee women 51  alike. 52Moreover, the inclusion of nearby communities and women's partners in educational interventions can be beneficial for women, as these allies become facilitators (and not barriers) to adopting health care-seeking behaviors. 53Lastly, educational interventions providing information show a self-reported empowerment effect on participants, especially when the intervention includes the creation of safe spaces and peer-support groups. 26,27Thus, educational interventions, combined with community-level strategies and availability of health services, can increase the probability of RH self-care or self-efficacy in the short, medium, and long term. 54ne common element that emerged across study narratives was that cultural competency and security should inform intervention design and implementation.We affirm this need and recommend that these elements are included as essential aspects of future intervention design and evaluations.Examples from the interventions analyzed include naming the emergency obstetric care unit a "minor operating theater" to respect refugee women's desire for privacy 22 or reconsidering the use of the label "sex worker," being that it is often a temporary and forced source of income that women do not use to define themselves. 27s reviewed studies included primarily monitoring and auditing data, there is a need for investigators to support as much as possible a health system culture of de-identified electronic record keeping, provided that the security and confidentiality of the data can be guaranteed-particularly in conflict areas, as records can get lost during violent attacks. 22,26,28,55Our review also shows that local social conflict may arise when interventions are not inclusive of all of the population living in a locality, such as some refugees receiving subsidized care while others do not 21 or refugees receiving more care than uninsured local populations. 23he latter is particularly more visible among refugees (compared to undocumented migrants), given the financial support that countries might receive from the UN Refugee Agency. 55ur review's identification of prevailing gaps suggests that further evaluation is needed to inform the states and the community of practice aiming to improve RH outcomes of migrant and refugee women living in protracted situations.Particularly important is evidence on how to address difficulties that arise because of the mobility of this population and the constant changes in staff providing health care in these contexts.
In addition, although our review shows the pivotal role that NGOs have in providing access to RH for migrants and refugees in protracted situations, the responsibility of states to provide comprehensive RH services to all women remains.Evaluating interventions and publishing results can help in enticing states to get involved in the initiatives and fulfill their responsibilities.For instance, 2 of the 10 included studies commented on the creation of collaborations between NGOs and public health care institutions or government officials, after the interventions' results became available. 21,22How data from evaluating interventions implemented by NGOs influence public policy and government actions in these contexts is an area that merits further exploration.
Other areas to explore include considering telehealth and digital platforms for information on RH topics, service options, and self-care to improve RH outcomes and other innovative interventions, as well as implementation science and participatory approaches.Studies included in this review do not mention the referral systems they use to provide comprehensive care, nor do they mention a trauma-informed approach to the RH services offered.Referral to mental health services is particularly important in the migratory context since migration is often related to traumatic violence experiences before or during the migratory journey.Similarly, future implementation and evaluation of RH interventions should consider a trauma-informed approach for RH (i.e., providing a trustworthy environment and ensuring care does not revictimize service users).Such approaches can be part of the training on cultural sensitivity for health care providers and other personnel working with undocumented and refugee populations.Traumainformed approaches and resulting outcomes should be assessed using implementation research.Documenting, assessing, and evaluating efforts to improve undocumented migrant women and female refugees' RH outcomes-possibly drawing on perspectives of these populations-are needed to strengthen care provision.A knowledge management platform for quick referencing could further allow NGOs and governmental institutions working toward improving RH outcomes in migrant and refugee populations to explore what other interventions have been implemented, where they have been implemented, and the results of these interventions.

Limitations
Our review has several limitations.First, this was a secondary study that directly depended on the contents and quality of data from the primary studies.As already mentioned, many studies did One common element that emerged across study narratives was that cultural competency and security should inform intervention design and implementation.Documenting, assessing, and evaluating efforts to improve undocumented migrant women and female refugees' RH outcomes are needed to strengthen care provision.
not include in-depth evaluations of the impact of interventions on the RH outcomes of migrant and refugee women.Quasi-experimental and pre-post evaluations with a control population group could shed more light on the impact of interventions on the target population.However, given the challenges of conducting rigorous evaluations among these populations, elevating the voices of successful experiences and conducting pragmatic implementation research should be prioritized in partnership with the community of practice.Obtaining information on the differential outcomes that interventions may achieve on subpopulations (e.g., adolescent versus adult women) could also provide indications on how to improve practice.Another limitation was that because the focus of the inclusion criteria was on RH, interventions aimed at addressing other SRH needs, such as sexual violence and sexually transmitted infections, were beyond the study's scope, even though these needs are highly prevalent among migrant communities. 56Further studies should focus on the impact of interventions to address gender-based violence and sexually transmitted infections among this target population.Additionally, the search strategies were done in 3 languages, and it is possible that we could have found more studies if the number of languages had been expanded.Finally, the heterogeneity of the data did not allow for reaching summary measures through a meta-analysis.Nevertheless, this heterogeneity revealed some of the diverse populations, contexts, and circumstances of mobile communities worldwide.Our findings represent a good portion of the variety of implementation contexts for NGOs and governmental institutions, allowing our conclusions and recommendations to be transferrable.

CONCLUSION
During the migratory process, undocumented migrant women and female refugees in protracted situations find themselves in contexts that put them at higher risk of sexual violence and unwanted RH outcomes.There is limited highquality research on migrant-centered RH interventions.However, interventions that reduce disease, unintended pregnancy, and fatal RH outcomes include subsidizing or giving financial assistance for accessing RH services and establishing or strengthening health services, including referral services, for undocumented migrant women and female refugees where they live.The successes and challenges to providing RH services identified within this review can inform future programming worldwide, particularly in regions like Latin America and the Caribbean, where fewer interventions have been implemented and evalu-ated.Intervention budgets and RH service financing should include evaluation as an essential element of future programmatic research.This review, consistent with other studies, suggests combining interventions in a culturally acceptable comprehensive approach will best meet the RH needs of migrant and refugee women.

FIGURE 1 .
FIGURE 1. Search Strategies in English, Spanish, and Portuguese Used in the Identification of Records for the Systematic Review

FIGURE 2 .
FIGURE 2. Selection Process for Systematic Review on Interventions to Improve the Reproductive Health of Undocumented Female Migrants and Refugees in Protracted Situations

TABLE 1 .
Inclusion and Exclusion Criteria for the Selection Process on Interventions to Improve the Reproductive Health of Undocumented Female Migrants and Refugees in Protracted Situations

TABLE 2 .
Continued *Reason to attend PNC * Danger signs in pregnancy * c Study rated low quality according to measures used.d Study rated medium quality according to measures used.e It is not clear in the study results if women at baseline and follow-up were the same women.
and of nonmigrant women