Multiple-micronutrient supplementation: Evidence from large-scale prenatal programmes on coverage, compliance and impact.

Micronutrient deficiencies during pregnancy pose important challenges for public-health, given the potential adverse outcomes not only during pregnancy but across the life-course. Provision of iron-folic acid (IFA) supplements is the strategy most commonly practiced and recommended globally. How to successfully implement IFA and multiple micronutrient supplementation interventions among pregnant women and to achieve sustainable/permanent solutions to prenatal micronutrient deficiencies remain unresolved issues in many countries. This paper aims to analyse available experiences of prenatal IFA and multiple micronutrient interventions to distil learning for their effective planning and large-scale implementation. Relevant articles and programme-documentation were comprehensively identified from electronic databases, websites of major-agencies and through hand-searching of relevant documents. Retrieved documents were screened and potentially relevant reports were critically examined by the authors with the aim of identifying a set of case studies reflecting regional variation, a mix of implementation successes and failures, and a mix of programmes and large-scale experimental studies. Information on implementation, coverage, compliance, and impact was extracted from reports of large-scale interventions in Central America, Southeast Asia, South Asia, and Sub-Saharan Africa. The WHO/CDC Logic-Model for Micronutrient Interventions in Public Health was used as an organizing framework for analysing and presenting the evidence. Our findings suggest that to successfully implement supplementation interventions and achieve sustainable-permanent solutions efforts must focus on factors and processes related to quality, cost-effectiveness, coverage, utilization, demand, outcomes, impacts, and sustainability of programmes including strategic analysis, management, collaborations to pilot a project, and careful monitoring, midcourse corrections, supervision and logistical-support to gradually scaling it up.

to combat maternal anaemia and iron deficiency, and to reduce the risk of low birth-weight (WHO, 2012). Efforts are also increasing to optimize weekly IFA supplementation targeting nonpregnant women and adolescents (Aguayo, Paintal, & Singh, 2013;Vir, Singh, Nigam, & Jain, 2008), to be used as a preventive rather than a therapeutic measure for improving iron status before pregnancy and preventing anaemia in pregnancy. Preconception care has the potential to positively impact a million pregnancies worldwide each year (Berti et al., 2017;Dean et al., 2013). In many countries, however, little progress has been achieved in ensuring adequate IFA coverage to improve pregnancy outcomes (Sanghvi, Harvey, & Wainwright, 2010). The increasing concern that IFA alone may not be sufficient to replenish the concurrent micronutrient deficiencies that often occur in pregnant women has encouraged the launch of multiple micronutrient (MMN) supplementation (Bhutta, Imdad, Ramakrishnan, & Martorell, 2012).
Scientific evidence demonstrates that compared with IFA, MNN supplements during pregnancy exert similar beneficial effects on maternal anaemia , mean birth-weight, incidence of both low birth-weight and small for gestational age (Fall et al., 2009;West et al., 2014), and neural tube defects and congenital heart defects (Czeizel, 2011). Because distribution systems are already in place to deliver IFA tablets to pregnant women, MMN supplementation may be a relatively cost-effective way of improving pregnancy outcomes in undernourished populations (Bhutta et al., 2012). Programmatic experiences of IFA supplementation may provide useful learning for the introduction of MMN supplementation. It is widely acknowledged that provision of micronutrient supplements is most effective when incorporated into a comprehensive context-based approach that integrates multiple interventions, including community-based supplement distribution, nutrition education, female/community empowerment, basic primary health care, sanitation, marketing activities, and national-level advocacy (Berti, Faber, & Smuts, 2014). Yet, little is known about how best to deliver these key interventions at scale, to ensure that impacts are achieved (Menon et al., 2014). Indeed, it is crucial to create actionable knowledge of relevant issues, including conducting operational research to identify methods and practices that effectively impact quality, coverage, equity, utilization, demand, outcomes, and sustainability of nutrition programmes (Menon et al., 2014). Given the incumbent call for addressing technical and evidence gaps to support policy makers and stakeholders with guidance on micronutrient supplementation interventions in pregnancy, in 2015 the World Health Organization (WHO), with the United Nations Children's Fund and the Micronutrient Initiative, launched a technical consultation "Multiple micronutrient supplements in pregnancy: implementation considerations for successful integration into existing programmes". The intention was to identify programmatic experiences of worth that were both successful and not, thereby extracting best practices for implementation and improved estimated adherence and sustainability. Purposely, in this paper evidence from large-scale prenatal IFA and MMN supplementation interventions is reviewed and synthesised to highlight factors, determinants and, mostly, actions (i.e., methods and approaches) that contribute to supplementation programme success or failure, with the objective of distilling important learning for effective planning and large-scale implementation in public health practice.

| Search-strategy
After prioritising the key-topics to be reflected by the case studies, relevant articles and programme documentation were retrieved from a comprehensive search of PubMed/MEDLINE, Google-indexed scientific literature, websites of major agencies, and through hand searching of the bibliographies of retrieved publications. Search terms included "effectiveness," "nutrition," "supplementation," "micronutrient," "programmes," "pregnant women," "implementation," with no date or language limits. The strategy for selecting the final set of case studies for this review is schematised in Figure 1. After screening titles and/or abstracts for relevance, we retrieved the full-text publications of large-scale operational experimental studies and descriptive studies concerning nutrition interventions for reproductive-age women and children, including IFA or MMN supplementation and/or other health services, and the related implementation factors and sustainability strategies, to identify existing relevant countries' programmes and projects (Phase 1). Only articles written in English, French or Spanish were retained. Next (Phase 2), we selected a subset of countries' experiences for this review by critically analysing available data and information based on the following criteria, derived from the WHO/ CDC Logic-Model for Micronutrient Interventions in Public Health (De-Regil, Peña-Rosas, Flores-Ayala, & del Socorro Jefferds, 2014), and the principles proposed by Mason, Sanders, Musgrove, Soekirman, and Galloway (2006) for effective community health and nutrition programmes: programmes and large-scale experimental studies (i.e., programmes combining one or more interventions implemented at least at the provincial or state level); targeting; available data on impact, that is, at least one of the main outcomes and/or outputs was measured over time (e.g., nutritional status, obstetric outcomes, coverage, adherence, awareness, etc.); and amount of

Key messages
• Micronutrient malnutrition persists among childbearingaged and pregnant women worldwide • Actionable knowledge of relevant issues at various scales is a key factor to align efforts, optimize costeffectiveness and accelerate progress towards improved nutrition outcomes • Strengthening or establishing national technical and intersectoral committees/partnership, and securing assistance from bilateral and international agencies are vital in programme scaling-up and sustainability • Systematic and professional training of health-workers along with community motivation and mobilization are crucial for achieving coverage and compliance • Creating awareness and knowledge about optimal nutrition before and during pregnancy augments the potential to beneficially impact pregnancy outcomes information on example, implementation and challenges. Finally, to construct an objective picture based on a broad array of implementation features, we chose a final set of case studies intending to reflect regional variation, a mix of implementation successes and failures, and a mix of programmes and large-scale experimental (i.e., pilot) studies, insofar as the availability of relevant information and data would allow.

| Evidence-synthesis
To organise findings and interpretations, we used a simplified version of the WHO/CDC Logic-Model for Micronutrient Interventions in Public Health, adapted from the original (Figure 2). By explicitly outlining the different programme components (inputs, activities, outputs, outcomes) and how they related to each other to lead to the expected outcomes, the model served as an organizing framework to identify, analyse, and discuss the factors and methods contributing to success or failure in coverage, compliance, and impact within the selected case studies. After synthesising the available evidence pertaining to each component of the WHO/CDC Logic-Model, we additionally considered other relevant factors important for prenatal micronutrient supplementation interventions that fall outside of the WHO/CDC Logic-Model, including the importance of national nutrition surveys to initially inform micronutrient intervention design (Mason et al., 2006). Finally, where possible, we outlined brief conclusions and main lessons learned from the case studies about intervention implementation actions, including challenges as well as the evidence gaps. In particular, we sought to gather information that described the implementation methods and processes, including best practices in terms of content, supervision or monitoring.

| RESULTS
Applying our inclusion and selection criteria to the potentially relevant records that we retrieved through our comprehensive search strategy, we ultimately selected a final set of eight case studies. Most retrieved documents were reviews and reports on nutrition-sensitive programmes/interventions or nutrition-specific interventions other than micronutrient supplementation, descriptive studies, action plans/roadmaps/guidelines exclusively describing implementation strategies and methods without any quantification of inputs and/or outcomes related to IFA or MMN supplementation, qualitative studies, and situational analyses with no data on programme/intervention effectiveness. In contrast, the most relevant sources of useful data and information were studies, reports, and summarising reviews of country cases' nutrition programmes, including micronutrient supplementation for women of reproductive age and children.
3.1 | Overall shortcomings in the current state of evidence: What is missing Our search revealed a paucity of literature documenting the impact of large-scale integrated nutrition and health programmes on obstetric and foetal outcomes. Indeed, while many trials have considered the effects of IFA supplementation on pregnancy and birth outcomes (Lassi, Salam, Haider, & Bhutta, 2013;Peña-Rosas, De-Regil, Dowswell, & Viteri, 2012), our search retrieved no reports on these outcomes in the context of large-scale IFA programmes. Additionally, we did not find any relevant paper that addressed exclusively challenging or failed supplementation programmes, although some cases are mentioned in literature. For example, South Africa was cited as a country where  (Liu et al., 2013;Wang, Pei, Song, Chen, & Zheng, 2013;Zeng et al., 2008), we could not find publications that focused in detail on the implementation aspects of these interventions. Moreover, because technical guidelines developed specifically for micronutrient deficiency are available as well as on-going nutrition plans or roadmaps for Ethiopia, South Africa, and Kenya, reports of related impacts and analyses of implementation issues are unlikely to be available for some years yet. Furthermore, our findings likely pointed out a lack of MMN or IFA supplementation programmes in high-income countries given that only studies of single-micronutrient interventions such as folic acid were found (Branum, Bailey, & Singer, 2013;Nilsen et al., 2006).

| Selected case studies
All were considered successful with respect to implementation and impact. Their general characteristics are summarised in Table 1. We included one case from Central America, one from South Asia, five from Southeast Asia, and one from Sub-Saharan Africa. Seven were case studies of IFA supplementation interventions, and one included both IFA and MMN. Three case studies were national programmes targeting pregnant women, and five were pilot studies targeting women of reproductive age (both pregnant and nonpregnant women).
All evaluated the intervention impact on incidence of anaemia and/or maternal or women's haemoglobin level, except for the Indonesia pilot study. Only two pilot studies investigated the impact on other health outcomes, that is, the association between neonatal birth-weight and an integrated deworming and IFA-supplementation intervention in Northwest Vietnam, and between the reduction in early infant mortality and the use of skilled birth attendants in Indonesia. Below, we present evidence from each case study on effective implementation strategies adopted, in the attempt to identify key intervention factors, that is which way worked and which failed.

| National programmes
Nicaragua (Mora, 2007) The available literature evaluating the Nicaragua national Integrated Anaemia Control Strategy (IACS) provided the most detailed information on implementation and impact of all the case studies included.
The IACS was developed and gradually improved, as part of a National Micronutrient Plan, with IFA supplementation to pregnant women provided as a part of a comprehensive approach including other strategies for anaemia control such as flour fortification; behavioural change communication (BCC); target-training to health service and non-governmental organization (NGO) personnel and community health volunteers; and development of a system for programme monitor and evaluation.
The key programme characteristics are listed in Appendix A. The IACS implementation succeeded in improving IFA supplementation's FIGURE 2 Conceptual framework of the WHO/CDC Logic-Model for Micronutrient Interventions in Public Health highlighting indicators tied to expected intervention processes, as adapted from De-Regil et al. (2014). The model is organised according to four main categories or components: Inputs: Resources invested in the intervention, including personnel, partnerships, politics and governance with different agendas, direct and indirect support from organizations, communities, and private sector, infrastructures, money, materials, and nutrition know-out. Activities: Actions, events and processes of programme implementation such as developing protocols, passing legislation and regulation, designing production and supply delivery systems, engaging stakeholders, providing training, setting quality control systems, planning dissemination, education, counselling and advocacy communication strategies. Outputs: Direct effects or results of programme activities, such as procurement of annual supplies and availability of the supply in the country; staff trained and motivated to deliver and counsel participants on the intervention; availability of the intervention in communities or markets; and access and coverage to the intervention. Outcomes: Benefits or changes among target populations during or after the intervention in terms of the impact on both the micronutrient deficiency-related issues (i.e., intake, nutritional status, morbidity and mortality, health functions), and the long-term viability components (i.e., behaviours, knowledge, motivation, decision making, skills, individual/systemic/strategic/operational capacity etc.) coverage and adherence and resulted in a significant reduction (~68%) of anaemia rates among women of reproductive age. The analysis of inputs and activities highlighted the presence of most factors acknowledged as successful for achieving favourable outputs and outcomes, including baseline country analyses, political commitment, and partnerships. The specific facilitative actions, that is, the strategic ways that contributed to the successful IACS implementation were oriented to: establishing and disseminating clear IFA supplementation policies and target-technical guidelines among all the health care personnel; involving community health volunteers in supplement delivery, follow-up and counselling, coupled with increasing knowledge of health care providers and community health volunteers on anaemia and supplementation; and creation of a coordination task-force including several stakeholders; conducting operational research. In contrast, some activities were not properly engaged due to practical challenges, thereby preventing any conclusion on their final role in the expected results.   (Mora, 2007); Nepal (Pokharel et al., 2011); and Vietnam a (Ninh et al., 2003) Pilot-studies targeting women of reproductive age Philippines (Nutrition Reviews, 2005); Indonesia (Shankar et al., 2009); Vietnam (Casey et al., 2010;Nutrition Reviews, 2005;Passerini et al., 2012); and Ghana (MacDonald et al., 2007) Type and mode of supplementation IFA Nicaragua (Mora, 2007); Nepal (Pokharel et al., 2011); Vietnam (Casey et al., 2010;Ninh et al., 2003;Nutrition Reviews, 2005;Passerini et al., 2012) (Mora, 2007); Nepal (Pokharel et al., 2011); and Vietnam (Casey et al., 2010;Ninh et al., 2003;Passerini et al., 2012) Data on health outcomes Anaemia and/or iron status Nicaragua (Mora, 2007); Nepal (Pokharel et al., 2011); Vietnam (Casey et al., 2010;Ninh et al., 2003;Nutrition Reviews, 2005;Passerini et  Vietnam (Ninh, Khan, Vinh, & Khoi, 2003) The national Iron-Deficiency Anaemia Control Programme in Vietnam The programme features are outlined in Appendix A. Several implementation strengths were evident in the programme planning.
The programme was delivered through the existing preventive health system, and it promoted community participation in the programme's management. IEC activities were implemented widely, focusing on iron-rich foods, to increase knowledge and awareness of the role of iron. Although a reduction of anaemia (~39%) was achieved in pregnant women, concurrently a significant decrease in programme coverage occurred in terms of both targeted women and benefited districts, probably due to lack of supplement supplies and low utilization of community health volunteers. In contrast, the survey in 2000 found out a likely role of IEC strategies in decreasing anaemia, as a significant difference in anaemia prevalence was seen between people who were and were not exposed to IEC messages. The overall shortage of information about activities and outputs, including adherence and IEC, limits understanding of the extent of the intervention impact on anaemia.

| Pilot studies
These studies were chosen as examples of operational research investigating the effectiveness of specific delivery-platforms such as community-based nutrition programmes, health system platforms integrating nutrition interventions, and/or market-based interventions at achieving either process or impact outcomes. Appendix A shows the main elements and implementation features of each case as well as the main outputs and outcomes. A summary of campaigns and relevant actions undertaken in each case is described below.
Western-Pacific countries: Philippines and Vietnam (Nutrition

Reviews, 2005)
These one-year pilot-effectiveness projects were conducted jointly by Department of Health and United Laboratories (UNILAB), the Philippines' largest private pharmaceutical company, with the support of the WHO Regional Office for the Western Pacific. "Social marketing" was applied as a framework to be applied for selling and dispensing preventive-IFA to women of reproductive age by community workers, and motivating them to both adopt and maintain the behaviour for a long period of time. These studies followed national nutrition surveys on the prevalence of anaemia among pregnant women. A master protocol was developed and adapted in each study based on contextual elements and rapid situation assessments, in order to study impact and processes of implementing a weekly-supplementation approach in countries with high anaemia rates but different socioeconomic systems and cultures.  but also because the methodologies applied to deliver interventions are often inadequate (Menon et al., 2014). From existing programmes, it is already known which are the key features and actions essential to initiate and sustain a micronutrient supplementation programme (Mason et al., 2006). Nevertheless, coverage and adherence rates often remain low. Hence, existing experiences can also be analysed to understand the ways in which inputs and factors, that is, capacities and strategies, were specifically and sustainably implemented. In this review, by using the WHO/CDC Logic-Model for Micronutrient Interventions in Public Health framework to evaluated programmatic components and subcomponents (what), we identified and summarised the actions and methods (how) which contributed to success in coverage, compliance, and impact within the selected case studies.

| Shortfalls
The

| Successful factors and strengths
The case studies we selected succeeded in reducing anaemia rates among both pregnant women and nonpregnant women. The analysis of inputs and activities highlighted the presence of factors acknowledged as successful for achieving favourable outputs and outcomes.
Indeed, the successful planning of the IFA interventions was foreseen by accurate baseline country analyses aimed at documenting the iron deficiency's magnitude, identifying all the potential risk factors, assessing the status quo (i.e., available inputs/resources and existing policies) and understanding which capacities needed to be developed.
These actions have been highlighted as crucial to align efforts and actions, and to accelerate progress towards improved nutrition-health outcomes (Deitchler et al., 2004;Olney, Rawat, & Ruel, 2012;Pelletier et al., 2012). All programmes and projects we selected, provided supplementation interventions through context-specific integrated delivery methods, by using existing delivery platforms. Developing context-specific action plans, along with operation research aimed at addressing key constraints and establishing the feasibility of delivery models, was also shown to be useful for the proper sequencing of actions for scaling up effectively and sustainably. These elements helped optimize cost-effectiveness and maximize available human resources, as also demonstrated in a systematic review about integration of targeted health interventions into health systems (Atun, de Jongh, Secci, Ohiri, & Adeyi, 2010). Furthermore, all experiences we reviewed had both high political advocacy and good governance, which were shown in West Africa to be vital in financing, research, policy-making, and programme implementation (Sodjinou et al., 2014).
Likewise, our cases showed national technical and intersectoral committees/partnership, and assistance from bilateral and international agencies. Government/Private sector/Community coalitions were particularly crucial to promote and expand the preventive IFA supplementation programme in the Western Pacific countries' pilot studies (Nutrition Reviews, 2005). Cases in Nicaragua (Mora, 2007), Nepal (Pokharel et al., 2011), Northwest Vietnam (Casey et al., 2010;Passerini et al., 2012) and Ghana (MacDonald et al., 2007) outlined the role of local and international NGOs as both active forces in filling public-health system at macro and microlevels, capacity building of human resource, and implementing health activities. They covered a broad range of tasks from providing regular, intensive technical support and training, funds, staff and expert consultants to conducting monitoring and adequacy evaluations or undertaking large-scale health interventions as observed in the MICAH programme (Berti, Mildon, Siekmans, Main, & Macdonald, 2010). Most importantly, the selected cases evidenced that during the implementation period, anaemia and potentially related complications such as maternal or infant mortality and low birth-weight, showed a significant decrease that occurred concurrently to the increase of women of reproductive age taking IFA and attending ante-natal care. The critical implementation delivering-activities to enhance coverage and adherence included the following: training and delivery within existing programmes (using what is already in place reduces costs); measurement and recording of supplement supply and adherence (using fieldworkers and health-workers), monitoring and supervisory systems; fixed time to deliver to home (home visits); and providing education and motivation to women (explaining benefits, side-effects, asking to note benefits; creating awareness/ knowledge).

| Strategic implementation actions: How were key intervention factors implemented effectively?
IEC strategies BCC strategies to promote behaviours that support adequate intake, ensure availability of supply, and help to empower communities to be more self-reliant are considered relevant factors to be implemented.
A systematic analysis of approaches to social and BCC for preventing and reducing anaemia suggested that supplementation could succeed by creating awareness and knowledge about the optimal behaviours to reach health and well-being within both consumers and all stakeholders involved . In our set of cases, the IEC implementation actions successfully employed improved the overall KAP concerning anaemia implications and IFA importance during pregnancy, as well as how to enhance supplements' iron-absorption through diet. They consisted the following: carrying out, on a regular basis, popular educational and creative events at clubs, religious places, and shops; disseminating field-tested communication materials among Women's Union collaborators filled the gaps in the existing national programme's IFA distribution system, which had contributed to the relevant decrease in programme coverage observed from 1995 to 2000 (Ninh et al., 2003). In the experiences we selected, the methods which appeared to facilitate supplement adherence consisted of delivering reminder pamphlets to women on how and when taking weekly tablets (e.g., always on the same week day before bedtime eating foods rich in iron and vitamin C) or dosing-schedule tips, and developing supplements' covers including educational information and listings of recommended locally available iron-rich foods. Furthermore, community health volunteers were effective for improving overall KAP concerning IFA, and promoting overall healthy behaviour among both expectant mothers and eventually their families. Community agents were also crucial in increasing ante-natal care attendance by promoting its early use, which in turn allowed women to benefit from available package of services such as deworming medicine and skilled birth assistants. Only a well-implemented and locally adaptable community health volunteer workforce can deliver highly cost-effective interventions that improve maternal and child health (Singh & Sachs, 2013).
The need to systematically and professionally manage, support, motivate and train lay community members, appropriately to context, task load and expectations, to be a part of the health workforce emerged as a core component of primary health care systems in low resource settings (Columbia University Technical Report, 2011;Solon, 2006).
As an example, the Indonesian SUMMIT programme (Shankar et al., 2009)

CONFLICTS OF INTEREST
The Authors declare no conflict of interest. C. Berti received financial support from WHO to prepare this paper.

CONTRIBUTIONS
Research strategy: All authors contributed to prioritize the key-topics to be reflected by the case-studies and the strategy to be followed in order to identify relevant cases. CB collected available articles and reports; CB and MFG screened the retrieved documentation and identified eligible case-studies. CB, MFG, ZAB, IC discussed and selected the most relevant case studies to be analysed.
Manuscript writing: CB analysed and interpreted data from the selected cases, wrote the first draft of the manuscript, attended the technical consultancy meeting, and finalized the manuscript accordingly; MFG contributed with data analysis and discussion, and with manuscript revision; ZAB and IC revised the manuscript. All authors approved the final version for publication.