Monitoring and surveillance for multiple micronutrient supplements in pregnancy.

Abstract The World Health Organization (WHO) recommends iron–folic acid (IFA) supplementation during pregnancy to improve maternal and infant health outcomes. Multiple micronutrient (MMN) supplementation in pregnancy has been implemented in select countries and emerging evidence suggests that MMN supplementation in pregnancy may provide additional benefits compared to IFA alone. In 2015, WHO, the United Nations Children's Fund (UNICEF), and the Micronutrient Initiative held a “Technical Consultation on MMN supplements in pregnancy: implementation considerations for successful incorporation into existing programmemes,” which included a call for indicators needed for monitoring, evaluation, and surveillance of MMN supplementation programmes. Currently, global surveillance and monitoring data show that overall IFA supplementation programmes suffer from low coverage and intake adherence, despite inclusion in national policies. Common barriers that limit the effectiveness of IFA—which also apply to MMN programmes—include weak supply chains, low access to antenatal care services, low‐quality behaviour change interventions to support and motivate women, and weak or non‐existent monitoring systems used for programme improvement. The causes of these barriers in a given country need careful review to resolve them. As countries heighten their focus on supplementation during pregnancy, or if they decide to initiate or transition into MMN supplementation, a priority is to identify key monitoring indicators to address these issues and support effective programmes. National and global monitoring and surveillance data on IFA supplementation during pregnancy are primarily derived from cross‐sectional surveys and, on a more routine basis, through health and logistics management information systems. Indicators for IFA supplementation exist; however, the new indicators for MMN supplementation need to be incorporated. We reviewed practice‐based evidence, guided by the WHO/Centers for Disease Control and Prevention logic model for vitamin and mineral interventions in public health programmes, and used existing manuals, published literature, country reports, and the opinion of experts, to identify monitoring, evaluation, and surveillance indicators for MMN supplementation programmes. We also considered cross‐cutting indicators that could be used across programme settings, as well as those specific to common delivery models, such as antenatal care services. We then described mechanisms for collecting these data, including integration within existing government monitoring systems, as well as other existing or proposed systems. Monitoring data needs at all stages of the programme lifecycle were considered, as well as the feasibility and cost of data collection. We also propose revisions to global‐, national‐, and subnational‐surveillance indicators based on these reviews.


| INTRODUCTION
Iron deficiency remains the most common micronutrient deficiency worldwide, particularly affecting children and women of reproductive age, including pregnant women (www.who.int/nutrition/topics/ida/ en/; Kassebaum et al., 2014;Stevens et al., 2013). The World Health Organization (WHO) currently recommends a daily dose of 30-60 mg elemental iron (60 mg preferred in setting where anaemia prevalence >40%) and 0.4 mg folic acid taken as early as possible in pregnancy to improve maternal and infant health outcomes (World Health Organization, 2012;World Health Organization, 2016).
Continued supplementation beyond pregnancy is also recommended if iron deficiency prevalence in the country is high, or if the pregnant woman is anaemic (Stoltzfus & Dreyfuss, 1989). Iron and folic acid (IFA) supplementation during pregnancy reduces anaemia and iron deficiency at term, as well as the risk of low birth weight (Peña-Rosas, De-Regil, Dowswell, & Viteri, 2012). Other micronutrient deficiencies, beyond iron and folate deficiency-such as vitamin A, vitamin D, vitamin B 12 and zinc-are also widely prevalent among women of reproductive age, particularly pregnant and lactating women in developing countries (Bhutta et al., 2013;Dalmiya, Darnton-Hill, Schultink, & Shrimpton, 2009;Jiang, Christian, Khatry, Wu, & West, 2005;World Health Organization, 1992). Poor maternal nutrition is frequently caused by poor access to nutrient-adequate foods, cultural practices that discourage women from gaining weight, long hours of physical labour, and recurrent infections (Christian, 2010). In 1999, the United Nations Children 0 s Fund (UNICEF), United Nations University (UNU), and WHO agreed on the composition of a proposed multiple micronutrient (MMN) tablet called United Nations International Multiple Micronutrient Preparation, which provides the recommended daily allowance of vitamin A, vitamin B1, vitamin B2, niacin, vitamin B6, vitamin B12, folic acid, vitamin C, vitamin D, vitamin E, copper, selenium, and iodine with 30 mg of iron and 15 mg of zinc for pregnant women (UNICEF/ UNU/ WHO, 1999). Efficacy and effectiveness trials have confirmed that MMN supplementation during pregnancy can improve outcomes beyond anaemia, including deficiencies of other vitamins and minerals, low birth weight, small for gestational age (Dalmiya et al., 2009;Fall, Fisher, Osmond, & Margetts, 2009), and preterm birth (West et al., 2014) & Bhutta, 2015). There were no significant differences for other maternal and pregnancy outcomes such as preterm births, maternal anaemia in the third trimester, miscarriage, maternal mortality, perinatal mortality, neonatal mortality, or risk of delivery via a caesarean section (Haider & Bhutta, 2015). Though the evidences were with insufficient data on neurodevelopmental and long-term follow-up of maternal and offspring outcomes, these findings still support the potential replacement of IFA supplements in pregnancy with MMN supplements in populations at risk (Bhutta et al., 2013).
According to the WHO/Centers for Disease Control and Prevention (CDC) logic model for vitamin and mineral intervention in public health programmes (De-Regil, Peña-Rosas, Flores-Ayala, & Jefferds, 2014), most micronutrient programmes are expected to follow particular intervention processes. These processes require inputs to support relevant policies, adequate production and supply, delivery of the programme, quality control, and behaviour change communication and intervention strategies in order to achieve expected outputs and outcomes ( Figure 1). All programmes also require effective programme management and monitoring. However, global surveillance data show that IFA supplementation programmes often suffer from low coverage and intake adherence of supplements, despite many countries include supplementation during pregnancy in national policies. Common barriers limiting the effectiveness of these programmes, which would also apply to MNP supplementation programmes, include weak supply chains, low access to antenatal care (ANC) services, gastrointestinal side effects, and poor quality interactions between healthcare providers and women, low-quality behaviour change interventions to support and motivate women, and weak or non-existent monitoring systems used for programme improvement (Bhutta et al., 2009;Deitchler, Mathys, Mason, Winichagoon, & Tuazon, 2004;Galloway & Mcguire, 1994;Nguyen et al., 2008;Sununtnasuk, D    Public health surveillance is the ongoing, systematic collection, analysis, interpretation, and dissemination of data for a health-related event that will be used in public health action to reduce morbidity and mortality and to improve health (CDC 2001;CDC 1999). Nutrition surveillance is the ongoing, systematic collection, analysis, interpretation, and dissemination of data regarding nutritional status and nutrition programmes to make policy and programmatic decisions that will lead to improvements in the nutrition situation of a population. While national nutrition surveillance focuses on programmes in an individual country, global surveillance examines the types and number of nutrition programmes and key indicators, such as coverage in countries around the world.
Logic models are useful tools to structure and organize information, and they can be used to convey information and  high-priority indicators based on the review of practice-based evidence and the sources mentioned above.

| Process for selecting indicators
Guided by these definitions of monitoring and surveillance, we reviewed practice-based evidence including monitoring manuals and guidelines, implementation manuals, and country reports from public health supplementation programmes; and expert opinion to identify existing and potential programme monitoring and evaluation indicators and potential data sources. In addition, we reviewed published literature, including literature related to micronutrient supplementation on monitoring and evaluation; efficacy trials and effective ANC and micronutrient supplementation programmes; cost; and systematic reviews on micronutrient supplementation. We also contacted experts in the areas of micronutrient supplementation implementation, ANC, and programme monitoring and evaluation. Programme input, activity, and output indicators were primarily found in guidance documents, whereas outcome indicators (e.g., related to status) were generally For example, we do not list any indicators for the outcome box "Improved development, performance, and productivity" as they are complex and costly to collect and are not needed for most programme context. Second, we expanded the indicator titles into an indicator matrix so that each indicator title included a description of the type of indicator (e.g., input), calculation of the indicator or operational definitions, and potential data collection methods or data sources, suggested frequency and timing of data collection, and possible targets for some indicators. Finally, we identified a set of key indicators for different public healthcare settings.   Table 1 includes 30 potential input, activity, output, and outcome indicators that can be selected and adapted to any setting or specific programme depending on resources and the stakeholder and programme needs and priorities. Appendix A presents the indicator matrix for the 30 potential indicators identified from the literature and the expert panel, including descriptions of the type of indicator, calculation of the indicator or operational definition, potential data collection methods or data sources, suggested frequency and timing of data collection, and the potential expected target for each indicator. Some indicators are further divided into two or three subindicators to fully capture the programme 0 s progress and impact.  pregnancy, were you given or did you buy any iron tablets or iron syrup?) and consumption/intake adherence (Q.421: During the whole pregnancy, for how many days did you take the tablets or syrup?).

| Key indicators for MMN supplementation programme monitoring
In 2014, in response to the DHS call for suggestions to modify their core questionnaire, the CDC, Strengthening Partnerships, Results, and Innovations in Nutrition Globally (SPRING), and MI suggested the following revisions to replace these two questions with a set of six questions. The questions would improve our understanding of where women obtain supplements, the formulation they take (e.g., iron, IFA, or MMN), whether the supplements were obtained free of charge or were paid for, and how many tablets were received during the pregnancy. It was suggested that Q.420 be revised into five questions (Q.420a. During this pregnancy, were you given or did you buy iron or IFA tablets, iron syrup, or multiple micronutrients? Q.420b. If more than one form was received or purchased, in which form did you receive or purchase the most? Q.420c. Did you get iron tablets, iron syrup, or multiple micronutrients during an antenatal care visit, during another visit to a health facility, at a pharmacy, from a community worker/volunteer, or from another source? Q.420d. Did you purchase your iron or multiple micronutrients supplements or receive it free of charge? Q.420e. During the entire pregnancy, how many iron/IFA tablets, syrup, or multiple micronutrient supplements did you receive or purchase [all forms; iron/IFA tablets, syrup, and multiple micronutrients]?) It was recommended that Q.421 be revised, too (During the entire pregnancy, how many iron/IFA tablets/syrup or multiple micronutrient supplements did you take?). We plan to continue to work closely with DHS to consider these revisions in future iterations of the DHS survey.
SPRING has developed a simple method to use DHS data to make a rapid, initial assessment of the distribution and consumption of IFA supplements among pregnant women (Sununtnasuk et al., 2015).
Briefly, this method assesses four sequential points at which an IFA supplementation programme commonly falters: ANC attendance/  importance in improving maternal-and child-health outcomes.

| Data collection systems/approaches
One major challenge to monitoring the key elements of ANC services is that systems are often established vertically and the individual components are the responsibility of different departments within the health sector (e.g., nutrition, family planning, malaria, and immunizations). This can result in fragmented programme monitoring systems that are inefficient, wasteful, and discouraging to frontline staff as they face an ever-increasing workload (Chaulagai et al., 2005). Developing a streamlined monitoring system that makes better use of time and resources will require improved communication and coordination within the health sector. At the district level, one potential approach to promote intersectoral integration is to convene district health management meetings to coordinate ANC monitoring activities. For instance, the Indian Ministry of Health and Family Welfare recently established "integrated monitoring teams" in highfocus districts to review action plans and discuss how to improve programme implementation and monitoring (Prasad, Chakraborty, Yadav, & Bhatia, 2013).
There are also biological reasons for establishing integrated monitoring systems in settings with widespread infections. First, impaired micronutrient absorption and metabolic alterations occur during states of acute and chronic inflammation, especially in the case of iron, where reduced iron erythropoiesis is mediated by hepcidin (Allen, 2005).
Second, safety concerns have been noted around providing iron in settings with recurrent infections, particularly malaria (WHO, 2007). The current evidence indicates that as long as women receive adequate healthcare, the benefits of supplementation during pregnancy likely outweigh the risks (Sangaré, van Eijk, Ter Kuile, Walson, & Stergachis, 2014).
The WHO recommends daily IFA supplementation during pregnancy with the caveat that "in malaria-endemic areas, provision of iron and folic acid supplements should be implemented in conjunction with measures to prevent, diagnose and treat malaria" (WHO, 2016). This recommendation may also be expected to apply to MMN supplementation; thus, integrating nutrition and malaria monitoring systems becomes advantageous. The World Malaria Report There are many gaps in current monitoring and surveillance systems, and more than one system will often need to be used to capture the necessary programme indicators. For example, the current HMIS cannot obtain information related to intake adherence because women are not taking the actual pills at the clinic and HMIS do not currently monitor whether women reported taking pills at a later ANC visit. In DHS surveys, women are asked about the number of pills consumed during their last pregnancy, which could be up to 5 years ago, creating a major issue with recall bias. The HMIS in different countries do not necessarily follow a standard approach, and the data are asked differently across countries. A report (Dwivedi et al., 2014) from 13 sub-Saharan Africa and south Asia countries found the majority of countries (12 of 13) did include IFA indicators in their HMIS, but much fewer include an indicator for consumption of 90 or more pills (six countries). There was also variations in whether it was captured in facility register, reported to higher levels in the facility monthly report, and/or client card. The use of HMIS and DHS data can be triangulated to get a better sense of where current programmatic bottle necks exist, but there also needs to be new and innovative approaches to collecting high-priority indicators, such as intake adherence.
The importance of input indicators emerged as a high priority because early on the big hurdle will be building political commitment to switch from IFA to MMN supplementation. Once there is buy-in and new guidelines and processes are in place, only small modifications will need to be made to the downstream programme indicators in countries where IFA supplementation programme previously existed. The indicators listed in this paper are useful for monitoring and surveillance, but they can also help document changes, advocate for change, propose/ revise policies, and help identify where to put funding or resources.
In summary, all the indicators presented in this manuscript are useful for consideration when designing and implementing an MMN supplementation in pregnancy programme monitoring system and for surveillance in various settings. At a minimum, to ensure a successful MMN supplementation programme, most programmes should include key indicators related to the monitoring of supply, coverage, and intake adherence. The suggested monitoring indicators described in this paper are relevant for many situations, but they need to be considered with regard to local data needs and requirements. They can be adapted or other monitoring indicators considered depending on the programme context and circumstances.

CONFLICT OF INTEREST
The authors declare that they have no conflicts of interest.