Combined protocol for severe and moderate acute malnutrition in emergencies: Stakeholders perspectives in four countries.

Abstract Each year, acute malnutrition affects an estimated 52 million children under 5 years of age. Current global treatment protocols divide treatment of severe acute malnutrition (SAM) and moderate acute malnutrition (MAM) despite malnutrition being a spectrum disease. A proposed Combined Protocol provides for (a) treatment of MAM and SAM at the same location; (b) diagnosis using middle‐upper‐arm circumference (MUAC) and oedema only; (c) treatment using a single product, ready‐to‐use‐therapeutic food (RUTF), and (d) a simplified dosage schedule for RUTF. This study examines stakeholders' knowledge of and opinions on the Combined Protocol in Niger, Nigeria, Somalia, and South Sudan. Data collection included a document review followed by in‐depth interviews with 50 respondents from government, implementing partners, and multilateral agencies, plus 11 global and regional stakeholders. Data were analysed iteratively using thematic content analysis. We find that acute malnutrition protocols in these countries have not been substantially modified to include components of the Combined Protocol, although aspects were accepted for use in emergencies. Respondents generally agreed that MAM and SAM treatment should be provided in the same location, however they said MUAC and oedema‐only diagnosis, although more field‐ready than other diagnostic measures, did not necessarily catch all malnourished children and may not be appropriate for “tall and slim” morphologies. Similarly, using only RUTF presented inherent logistical advantages, but respondents worried about pipeline issues. Respondents did not express strong opinions about simplified dosage schedules. Stakeholders interviewed indicated more evidence is needed on the operational implications and effectiveness of the Combined Protocol in different contexts.


| INTRODUCTION
Each year, an estimated 52 million children under 5 years of age suffer from acute malnutrition (WHO, & World Bank Group, 2019). At least 500,000 children die each year from severe acute malnutrition (SAM; Black et al., 2013); these children also have weakened immunity and are susceptible to long-term developmental delays and chronic disease later in life (Black et al., 2008;Lelijveld et al., 2016). Children suffering from moderate acute malnutrition (MAM) are also at higher risk of death compared with those who are not acutely malnourished (Chang et al., 2013), and moderate malnutrition can create lifelong problems due to cognitive development, and reduce lifelong earnings (Black et al., 2008). Food insecurity and hunger caused by ongoing and new conflicts and increasingly by climate change and migration mean that child malnutrition will remain an issue of the utmost concern.
Global treatment protocols, adapted by most countries for use at national level, call for MAM and SAM to be managed separately, despite malnutrition being a spectrum disease. SAM is treated with ready-to-use therapeutic food (RUTF) in Outpatient Therapeutic Programs (OTPs), supplied and led by UNICEF. In the absence of codified global guidelines, MAM is often treated with ready-to-use supplementary food (RUSF) or corn soy blend++ in Supplementary Feeding Programs (SFPs), supplied and led by the World Food Programme (WFP).
The International Rescue Committee (IRC) alongside other global actors has been researching the use of a Combined Protocol to manage both conditions, with diagnostic criteria adapted for emergency settings (only middle-upper-arm circumference, MUAC, and oedema instead of both MUAC and weight-for-height z-score, WHZ, together with oedema) as well as a simplified dosage schedule. These approaches are meant to provide harmonized technical guidance, cost-efficient supply management, and optimal care for the greatest number of children and families under challenging circumstances. This protocol was tested by the IRC in a randomized controlled trial known as the Combined Protocol for Acute Malnutrition (ComPAS) trial, with results pending (Figure 1; Bailey et al., 2016;Bailey et al., 2018).
Policy discussions about combining, simplifying, and/or improving current malnutrition treatment protocols have advanced among technical and financial partners at global and regional level, particularly in West Africa and to a lesser extent in East Africa. Versions of these approaches have found traction in conflict-affected and/or foodinsecure settings where logistical constraints make delivery of the typical MAM/SAM protocols difficult if not impossible (Hanson, 2017;Woodhead, Rio, & Zagre, 2019). However, it remains unclear the extent to which these protocols are deemed acceptable and appropriate in countries where their use is currently being proposed. It is important to understand factors that influence national stakeholders' consideration and adoption of novel malnutrition protocols in emergency settings.
This study draws on four case studies to analyse national policy discussions and decision-making around acute malnutrition policies in Niger, Nigeria, Somalia, and South Sudan. In each of these countries, rates of child malnutrition have reached emergency levels due to protracted political crises (Somalia and South Sudan) or more localized conflicts (Niger and Nigeria), often exacerbated by seasonal droughts and the effects of climate change. The objective of this article is to examine the perspectives of country-level stakeholders with respect to proposed modifications to current MAM and SAM treatment protocols under the Combined Protocol in emergency settings.

| METHODS
Case study methodology is used to reconstruct phenomena to reveal underlying processes and is thus well-suited to policy analyses (Yin, FIGURE  were conducted with Abuja-based respondents). Semistructured interviews were conducted in countries (n = 46 interviews, n = 50 interviewees) and with stakeholders at global/regional level (n = 8 interviews, n = 11 interviewees; Table 1). Respondents included personnel of the MOH, UNICEF, WFP, and various NGOs in most countries. All respondents were asked about policy discussions at the national level including key actors and recent events; specific barriers or hesitations about modifying protocols; operational, financial, and practical considerations; and the importance of scientific evidence and global guidance. Most interviews took place in person, although some took place remotely via Skype or phone. Interviews lasted an average of 54 min (range: 33-82 min) and were conducted in English for stakeholders in Nigeria, Somalia, and South Sudan and in French in Niger and with some regional/global respondents. A small number of respondents requested not to be recorded; notes were taken for these interviews. Recorded interviews were transcribed verbatim then verified and completed by interviewers.
Data analysis was iterative and concurrent with data collection, beginning with regular debriefing discussions between lead data collectors to discuss emerging themes and adjust interview questions, combining analytical strands from the document review and interviews. Data from documents were abstracted into a standardized coding form to systematically track key themes, allowing for comparison across cases of theoretical and contextual categories (Maxwell, 2005). In-country data collection concluded with debriefing sessions to share initial impressions with and seek any clarification from IRC's country nutrition program focal points. Subsequently, interview transcripts were analysed using thematic coding, with categories developed a priori based on the theoretical literature and research questions; emergent codes were also developed after testing the coding structure on a first set of interviews. Interviews were coded using NVivo software (Version 11); the final codebook included codes on the content of current malnutrition policy and protocols; policy processes in countries; policy interactions at country, regional, and global level; national context; and the origins, rationale, and arguments for or against combined/simplified protocols. A synthesis report was reviewed by in-country IRC staff in all countries for accuracy of facts and interpretation, resulting in minor adjustments.  (Table 2). Differences were particularly notable for MAM treatment, likely due to a lack of global guidance. In Somalia and South Sudan, children received supplementary feeding in Targeted Supplementary Feeding Programs (TSFPs); however, treatment for MAM was virtually absent in Nigeria, both from the protocol and in the field (in Borno state and nationally). Treatment for MAM was included in Niger's protocol; however, RUSF or other supplementary foods were available only in insecure regions and/or those with very high global acute malnutrition rates. Guidelines varied in terms of preference for MUAC over WHZ as screening and discharge criteria, although all had incorporated MUAC in some way. Modes of dispensing treatment also varied, although in most cases MAM and SAM care were not provided in the same place on the same day. In all countries, national guidelines were meant to be applied nationwide including in emergency settings. Next, we examine stakeholders' views on the four proposed components of the combined protocol, as laid out in Figure 1.   In Somalia and South Sudan, there was also strong acceptance of providing MAM and SAM treatment in the same place, and respondents frequently characterized the main problem with malnutrition care as being the fact that OTPs and TSFPs were often not placed in the same location:

| Agreement that MAM & SAM should be treated at the same location
"If you see a mom with a SAM with complications child at your [stabilization center] and … you say to the mom, 'I will refer you in a two kilometers or three kilometers away to the OTP program,' she will face some obstacles, starting from the transportation."  Strong objections to a MUAC and oedema-only protocol were also raised in Niger along different lines: quality of care. In Niger, national policymakers perceived that they had been "lobbied" for MUAC and oedema-only by stakeholders from international donors starting in 2014-15. Government officials opted to retain WHZ as one possible criteria for admission, and a necessary criterion for discharge. Respondents, particularly those in government, appeared annoyed when asked about a MUAC and oedema-only protocol, asserting strongly that WHZ was essential because it is a superior measure for growth monitoring, and is already used throughout the health system for this purpose: "You agree with me, a correct anthropometry requires weight-for-height." (Government official, Niger) "It's true, when you have to measure weight, height, it requires … several people and it takes time, but it's not "For [MUAC] to be an independent admission criteria, it required a memorandum from the Ministry … and it was a long battle for that to be accepted." (International NGO, Niger)

| Reservations about the appropriateness of RUTF as the single product
Respondents expressed different opinions on whether RUTF and RUSF were interchangeable products from a clinical standpoint. The majority of respondents said they were very similar, but others said there were significant differences in terms of composition or nutri-  with cost breakdowns of US$ 83 for outpatient treatment costs and US$ 149 for inpatient treatment costs, respectively (Isanaka et al., 2017). Indeed, among 10 interventions recommended to improve maternal and child nutrition, SAM management was by far the most expensive, estimated at a total cost of $2.6 billion to achieve 90% coverage in the 34 highest-burden countries (2010 international dollars; Bhutta et al., 2013). Yet, the same study also showed that effective management of MAM and SAM was among the most cost-effective interventions per life saved, potentially saving up to 435,000 lives per year, when scaled up alongside the other nine interventions.
As a result of ongoing research, evidence is beginning to emerge that simplified approaches can provide safe, effective treatment in emergency settings. In Sierra Leone, a cluster-randomized controlled trial used an integrated protocol with a single product (RUTF) and MUAC-only admission (<12.5 cm) to test recovery and coverage rates (Maust et al., 2015).  . SAM children who were discharged in the second period using the MUAC-based criterion also had shorter lengths of stay and fewer adverse program outcomes.
Further evidence is awaited from the ComPAS trial (Bailey et al., 2018) and research by WFP, UNICEF, European Civil Protection and Humanitarian Aid Operations, and other partners in West Africa (Kaul, Husain, Tyrrell, Gaarder, & Jimenez, 2018).
There have been fewer trials of combined/simplified protocols in South Asia, although this region is home to the world's largest number of severely wasted children (up to 70% of the global burden) (Ahmed et al., 2014). Nonetheless a few country programs and projects have adopted some of the proposed modifications discussed here, albeit in nonemergency settings. In India, a CMAM program initiated by Médecins Sans Frontières in Bihar state that admitted children using MUAC < 115 mm for admission and MUAC ≥ 120 mm for discharge attained low mortality and high cure rates, in part by lowering the threshold for severity compared to WHZ < −3 SD (Burza et al., 2015). Pakistan's CMAM program has also achieved good recovery and survival rates using Lady Health Workers to screen and refer children based on MUAC < 115 mm (Aguayo et al., 2018). However, care should be taken in comparing these programs given that acute malnutrition in the settings studied in our analysis is driven by severe food insecurity (often prefamine warning levels), whereas the drivers of wasting may be different in South Asia.
Nonetheless, it is clear that more evidence is needed on specific aspects of combined/simplified protocols before national decisionmakers will substantially revise national protocols. For example, the hesitancy to adopt MUAC and oedema-only diagnosis observed in our study is understandable given the lively and ongoing scientific debate on this topic Grellety & Golden, 2016a (Hanson, 2017;Phelan et al., 2015). Although some analyses have suggested that up to three-quarters of severely malnourished children are missed by MUAC <115 mm criteria (Grellety & Golden, 2016b), most proposed MUAC-only protocols use a higher threshold than 115 mm. Interventions have sometimes used the sitting height to standing height ratio (or Cormic index) to correct of weight-forheight measurements for the body morphology of some ethnic groups (Roberfroid et al., 2015;Salama et al., 2001), but the issue of geographical variations body shape requires further study . Evolving understandings of the frequency of multiple anthropomorphic deficits and the overlap and dynamic interactions between wasting and stunting will also influence this debate (Myatt et al., 2018;Wells et al., 2019). Although it is beyond the scope of this article to adjudicate the debate around MUAC versus WHZ, greater clarity around the specific conditions under which MUAC is an appropriate diagnostic tool, emphasizing its practicality and ease in difficult field situations, will be needed to assuage stakeholders' observed misgivings.
Country-level stakeholders also need more information about the appropriateness of RUTF as the single product to treat both MAM and SAM. Both products are lipid-based nutrient pastes containing similar amounts of protein, although usually from (more expensive) animal sources for RUTF or from soy or whey for RUSF; amounts of vitamins and micronutrients also vary between the two products (Isanaka et al., 2010). RUSF was developed for treatment of MAM following the observed success of RUTF for treating SAM; however, RUSF is meant to supplement, not replace, other sources of nutrients.
Yet the physiology of weight gain is essentially the same between MAM and SAM (Briend et al., 2015), suggesting potential exchangeability between products, although few studies exist to test this prop- ). An earlier study in Niger found a lower risk of wasting with RUSF compared with RUSF; however, this difference was attributed to factors related to a previous intervention (Isanaka et al., 2010). More research is needed about the relative efficacy, including weight gain but also micronutrient absorption and functional outcomes, of RUTF, RUSF, and their variants, as well as the cost effectiveness of the different options (Briend et al., 2015;Schoonees, Lombard, Musekiwa, Nel, & Volmink, 2019).
This study has some limitations. Not all interviewees were able to be reached in-person in all study countries due to the short duration of visits for data collection. In Somalia in particular, the researcher's time in country was restricted due to security protocols, which also made it difficult to reach many potential respondents, although additional stakeholders were interviewed in Nairobi. We mitigated this limitation by conducting additional phone interviews and triangulating with documents, reaching saturation on most relevant points.
However, given small sample sizes, it remains possible that some perspectives were not fully represented. Next, data were collected by a team of two researchers one in Niger and northeast Nigeria (SD) and the other in Somalia and South Sudan (NK). To ensure congruity, the two researchers held regular and frequent discussions, including debriefing during or shortly after field visits and throughout data analysis. Next, given that NK is an IRC employee, and SD an IRC consultant at the time of data collection, it is possible that some respondents may have tailored their answers on the Combined Protocol, which is often associated with IRC. We used triangulation between respondents, as well as probing on potentially controversial aspects of the protocol, to mitigate this risk. Finally, a handful of interviews at both country and regional level took place with two participants, rather than individually. Given that participants in these settings may have refrained from sharing negative views, particularly regarding their own organization, we interpreted these interviews as representing an institutional standpoint, rather than two individual actors.
Field-level stakeholders are interested and aware of the potential benefits of simplified approaches to treating child wasting in emergency settings, including those included in the Combined Protocol.
Further research on specific aspects of these protocols is needed to answer scientific and operational questions, as is already recognized by regional-and global-level stakeholders, especially on issues where is a lack of consensus. Ongoing dialogue between these levels, and especially with national and MOH officials, will ensure planned research incorporates and answers field-level concerns. At country level, stakeholders should be informed about ongoing research and be alert to ways to improve the effectiveness and cost efficiency of their own acute malnutrition programs, including as these may be affected by any global guideline revisions.

This project is funded by Elrha's Research for Health in Humanitarian
Crises programme, which aims to improve health outcomes by strengthening the evidence base for public health interventions in humanitarian crises. The Research for Health in Humanitarian Crises programme is funded equally by the Wellcome Trust and the UK Department for International Development. Visit www.elrha.org/ r2hc for more information.
This research was made possible by the support of staff at IRC field offices in Niger, Nigeria, Somalia, South Sudan, and the IRC Somalia satellite office in Kenya. We also thank Sophie Woodhead for providing valuable context and resources on the latest work on simplified approaches in West Africa and at global level.