Nutritional care for patients with Ebola Virus Disease in Ebola treatment units – past and current experiences from practitioners

Background In November 2014, the World health Organization (WHO), in collaboration with UNICEF, and the World Food Programme, produced interim guidelines (iGL) on providing nutritional support to patients in Ebola treatment units (ETUs). They have been translated into French and issued by the Ministry of Health, UNICEF and WHO in adapted versions to be used in the current outbreak in the Democratic Republic of the Congo (DRC). This paper evaluates the use and usefulness of the 2014 iGL in the West Africa and current DRC Ebola virus disease (EVD) outbreaks and identifies experiences and lessons learned from practitioners on the operational aspects of nutritional care and support in ETUs. Methods Key-informants (n=26), from 12 organizations (Non Governmental Organizations, United Nations, Red Cross Red Crescent Movement) were interviewed who were actively engaged in the nutritional and/or clinical care of EVD patients. Results There was a consensus among key informants that the 2014 iGL initially served a guiding purpose. However, the vast amount of learning from the 2014-2016 and current EVD outbreaks indicates that the interim guidelines need to be revised. Practitioners struggled to find operational solutions for nutritional care, and the challenges were plentiful, especially regarding 1) the different perceptions of the importance of nutritional care among ETU staff; 2) the difficulties around food preparation and distribution for EVD patients; 3) how to take into account the patients’ dietary preferences; 4) the nutritional care needed in relation to specific EVD symptoms; 5) who assumed roles in nutritional care in ETUs; 6) if and how feeding support was organized; 7) whether malnutrition needed to be addressed and how; and 8) whether the intake of specific nutrients could contribute to improved treatment outcomes. Information from the key-informants interviews resulted in numerous lessons learned and recommendations for nutritional support during current and future outbreaks. Conclusions This investigation underscored the importance of documenting experiences of

Through key informant (KI) interviews, Kodish et al. reviewed the use of three interim guidelines co-issued by WHO in Guinea including the iGL. However, the findings were not specifically linked to this 2014 iGL [3]. A previous literature review found a limited number of publications on specific nutritional care in ETUs that often lacked detailed descriptions of the actual nutritional care provided (Ververs/Gabra to be submitted for publication).
The finding highlighted the importance of reporting on practitioners' experience to revise existing guidelines for increased applicability and acceptability.
This investigation is a follow on to the previous review with the objectives to a) identify During the interviews, conducted by Skype [4], detailed notes were taken and analyzed using MAXQDA version 2018.2 [5]. Within MAXQDA, a coding framework was designed based on four major themes emerging from the interviews, Box 1. Through coding and interpretation of the interview transcripts, information was consolidated into experiencedriven understanding of previous and current EVD outbreaks. Informed consent was sought through a verbal consent process during the intreveiws. This project was reviewed in accordance with the Centers for Disease Control and Prevention human research protections procedures and was determined not to be human subjects research.
Participant privacy was assured by de-identifying the project data and quotes. Interviews were conducted until information became repetitive.

Box 1. Coding framework on four main themes and sub-themes
Due to technical limitations, Box 1 is only available as a download in the supplemental files section.

Findings
In total, 26 KIs from 12 organizations were interviewed, Table 1. Sixteen KIs had direct contact with EVD patients in the ETUs. All KIs were international staff. The patient load in ETUs varied between 40 -200 patients per day, and the majority of patients were adults.
There were no refusals among KIs contacted. Interim GL WHO-awareness and usefulness as perceived by the interviewed practitioners to be determined later Twenty-four KIs were familiar with the 2014 iGL and two only had knowledge of the guidelines issued in 2018 by WHO/Ministry of Health DRC/UNICEF [6]. Most organizations (8 out of 12) developed their own protocol/guidance on nutritional care during the West Africa EVD outbreak. Some aligned their protocols to the 2014 iGL upon publication.
One-third of the KIs found the 2014 iGL not to be sufficiently practical. While the technical aspects were acceptable, the feasibility was questioned. Protocols developed by individual organizations were more detailed on what types of foods to provide and how. However, at least half of the KIs stated that the 2014 iGL were still useful, specifically the decision tree. KIs stated that despite the lack of evidence base for the 2014 iGL, it provided some guidance, especially in the beginning of the West Africa outbreak. Most KIs agreed it was essential to develop the guidance continuously and to contextualize it within the countries affected.
Some KIs questioned the recommended use of therapeutic milk in the 2014 iGL, because the local community was not regularly using milk as part of their diet. This was particlayrly problematic for adults. Additionally, it was questioned whether milk would aggravate diarrhea.
KIs also recognized limitations in their performance during the West Africa outbreak because of the focus on the use of specialized products (e.g., therapeutic milks, ready-touse therapeutic foods (RUTF), fortified porridges such as Corn-SoyBlend (CSB)), despite the desire of patients to consume local diets: 'iGL recommended certain specific food products and local foods. 'Patients did not accept the porridge made from specialized products. A lot of food was wasted. Critically sick patients preferred their local food. But local foods are not fortified.' 'We asked initially support from experts who were specialized in acute malnutrition, but we should have asked experts that knew a lot about very sick patients, and not those. We needed rather advice from experts from hospital settings, e.g., perhaps protocols needed as if they were surgical patients'.
Nutritional care for EVD patients in ETUs-practical experiences from interviewed practitioners Importance of nutritional care KIs and their colleagues perceived the importance of nutritional care in EVD patients in different ways. Although most KIs noted the importance of the role of nutrition in patient care, some KIs mentioned that clinical staff was predominantly concerned with medical treatment: 'Nutrition wasn't such an issue in Ebola-There were a lot of other clinical issues of concern than the nutritional status of patients. The biggest risk of clinical care was heat stress for patients and staff.' 'There was no focus on food and nutrition. Nobody was checking if people had actually eaten their food.' 'As there was not much malnutrition, staff also did not prioritize nutrition.' 'Teams are under pressure, and they forget that patients need to eat.' 'We had medical doctors from country X; they said 'we are here to save lives', nutrition was not seen as important.' 'Medical doctors stated that nutrition was not important, even if they were malnourished and treated, they might still not survive. They just need to survive EVD.' Other KIs were members of teams that acknowledged the role of nutrition in the care of the patients. It was seen as a responsibility of the general nursing care.

Food preparation and distribution Food preparation and transport
Half of the organizations providing food to EVD patients, prepared in onsite kitchens run by the organizations responsible for ETUs; the other half contracted local catering companies. One organization initially ran a kitchen and then transitioned to catering, while a second organization reversed this pattern. Both options had advantages and disadvantages, see Table 2.

Table 2. Perceptions expressed by Key Informants on advantages and disadvantages of using catering and onsite kitchens in preparation of foods for EVD patients in Ebola
Treatment Units.
Most KIs had experienced challenges in the provision of foods in West Africa and DRC: 'They were constructing the ETU. This included a kitchenette for the staff meals. When I asked about the meals for the patients, everybody remained silent. They had forgotten the issue on meals preparation for patients'.
'Food supplies were probably a greater problems in MoH hospitals where families were normally bringing in support and resources./../One hospital was not able to provide adequate support for nutrition. They mostly had doctors; many nurses had left; they had problems concerning food for patients.' 'We found a major problem in ETUs in terms of food that was provided to patients. In the beginning, food was provided once a day. The food supply was low'.
'Catering was the only option, … but delivering food is not the same as nutritional care.' 'There were food availability problems in the beginning: Non-Governmental Organizations (NGOs) knew they were not feeding well. Some died maybe because they were not eating well. We knew we had to provide nutritional care.' KIs noted that foods were transported to the patients the ETU ward with confirmed EVD patients (Red Zone) through slides, hatches, or staff entrances. Food was primarily provided on hard plastic plates or polystyrene compartment trays. Additionally, plastic cups, sippy cups, bowls, small plastic bags, or mineral bottles were used. Spoons were typically the only utensil provided. However, KIs noted the usefulness of straws, which were less frequently used. Hard plastic materials were disinfected after use or burned, and all other materials and leftover food were disposed of as hazardous materials. Foods were individually packaged; some ETUs labeled commodities with patient numbers, patient names, or just the composition of the meal. Food was placed on tables or chairs at the bedside or in communal areas for those who could eat at a table. Various challenges were reported: 'We had initially no idea how to get food in the Red Zone except through changing rooms.
But we did not want to mix food with health staff; we had to set up a food hatch system. However, one night, a child tried to escape from the Red Zone through the hatch.' 'Food was sitting in big bags. Nurses would give food in polystyrene trays. Soup was in a plastic bag. Adult patients were not able to untie the knot. They were not strong enough to eat with their hands. Food would just stay there. Nurses would be busy. Nobody was checking if people had actually eaten their food.' 'For fluids, big 1-1.5 liter plastic bottles (for Oral Rehydartion Salts (ORS) were provided, but people were too weak to pick up. Water was too hot. 40 degrees Celsius.'

Meal composition
KIs agreed that the provision of meals and snacks evolved over time with much trial and error. There was little standardization across organizations and menus and approaches varied. Some organizations were limited on the types of foods available, which led to the use of specialized products (RUTF, High Energy Biscuits, therapeutic milks, ready-to-use supplementary foods (RUSF), and/or CSB). The ease of use of the products by staff was also a factor in their selection. All organizations attempted to provide a balanced diet through the aid of specialized products. These specialized products were generally the recommended commodities in the 2014 iGL. KIs mentioned that the rationale behind the use of RUTF and therapeutic milks was the nutrient density of the products, easier consumption for patients with difficulty swallowing, and consumption without utensils.
Overall, three meals and 2 to 3 snacks were provided in most ETUs. Meals were provided in different consistencies: standard, soft/semi-solid, and liquid. Typical meals were as follows: Breakfast: porridges (often made from CSB), RUTF, therapeutic milks, fruits/juices, bread, local doughnuts, eggs Lunch: often of a standard 'family' meal made from staple foods and meat/fish/chicken, sauce, beans, oil, local soup Dinner: similar to lunch or breakfast Snacks were provided upon request or followed fixed schedules and included bread, fruits, biscuits, RUTF, RUSF, crisps, sweets, etc. Liquid diets for very sick and young children were mainly therapeutic milks, soups, or RUTF diluted with water. Semi-solid/soft foods mainly consisted of mashed regular meals (listed above), RUTF, RUSF, porridges, or soups.

Patients' dietary preferences and intake
The majority of adult patients in ETUS covered by the KIs did not prefer specialized products. Reasons given for the dislikeof the products included the monotony of diet and lack of palatability. Similar experiences were captured in both West Africa and DRC. This resulted in patients not consuming the supplied foods, but also significant wastage in the Red Zone, which created a health hazard. KIs mentioned numerous challenges: 'At arrival, the food systems and how they were organized were based on the use of RUTF, RUSF; Many people did not like this and wanted to have "family" meals.' 'These specialized products, patients did not eat it. This was piling up in the Red Zone, and health workers told me it was becoming a health hazard inside, and leftover food had to be burnt.' 'Adults in critical stage got RUTF, but the feedback was not positive, so we continued with local foods.' 'Adults, receiving RUTF, were not happy. They wanted local food. We saw that patients weren't eating and food was rotting at the bedside. We organized more rigorous criteria for who got therapeutic milks or RUTF, the dosage, and even local food; but I still pushed for patients to eat RUTF packets.' 'In DRC we have many fruits; the nutritional care should not be with these specialized products. It works maybe well in other African countries. Some organizations insist on giving 2 RUTF sachets/day and leaving meals as secondary. This is problematic for both patient and health workers.' Many KIs were torn between what they thought was best in the given context and the reality of the refusal of foods by adults. Some organizations shifted to a more patientcentered approach, taking into account the patients' preferences for local foods. However, other organizations continued to provide only standard menus with limited flexibility to meet patient preferences. As one KI stated:' Not eating by patients was a big problem; that was less a problem later when we gave them what they wanted. We had to individualize care because some were not eating, and many were dying from hypoglycemia.' Liquid diets were problematic for some patients. This was likely a combination of the commodity (therapeutic milks) and the patients' weakness, lack of appetite, and the monotonous diet. Providing sufficient nutrition to critically ill patients was a significant issue. Nausea, vomiting, anorexia, weakness, and hiccups led to difficulty in consuming foods, which resulted in weight loss in some cases. KIs stated that once patients began to recover, and their appetite returned, patients could consume large amounts of food.
The West Africa outbreak identified two local foods which were preferred and consumed by adults, including extremely ill patients: pepper soup and coconut water. In West Africa, various KIs mentioned that pepper soup was the preference of very sick patients, with fever, in a catabolic state, anorectic, and/or unwilling to eat. One KI stated: 'As they were sick, nothing worked so we changed the whole menu to pepper soup because the patients did not want anything else'. Nutritional care related to EVD symptoms Some KIs mentioned that there was an unwritten rule that the first 7-10 days after admission to ETUs were crucial. Regarding nutritional support, one KI stated, 'If they

survive the first week and especially the first few days, nutritional support is very important for survival. Exactly then we need to listen and give them what they want.'
In the West Africa outbreak, anorexia, vomiting, and diarrhea were particularly severe [7].  'Adults who received IV fluids had little to no oral intake. Once they were off the IV they were given foods. I am convinced we can get them eating earlier if people had looked into nutrition sooner.' Some KIs worked diligently to improve oral fluid intake by EVD patients. As one KI stated: 'Patients would take ORS, but would come to ask for soda and coconuts (which we gave).

The self-administration, acceptance, and willingness to take in fluids were important!'
Some organizations ensured that there was cold water, especially in the context of the hot climate of West Africa: 'Plastic packets of water, we stuffed the fridges full of those. We threw them over the fence for people to drink. They were nice and cold and appreciated.' Others organized to have ice for patients with sore throats.
The use of sodas was discouraged in the 2014 iGL. "Owing to the high osmolarity of sugary carbonated beverages or juices, it is important that they are not given to patients with diarrhea. These products are low in electrolytes and nearly all essential nutrients" [2].
However, some KIs argued that they were useful, including for patients with diarrhea, frequently requested by patients, and helped to increase/maintain fluid intake. As one KI Sip feeds, designed for specific medical purposes, were included as an option for liquid diets in the 2014 iGL. Sip feeds often have a low renal solute load, low osmolarity, and come in lactose-free options. However, no KIs reported having used these products in ETUs but some stated that it might have been a missed opportunity. Sip feeds, though not ideal, may be more acceptable to adult patients than therapeutic milks or RUTFs. Various reasons of non-use were reported: unavailability of product in organization supply catalogs, unfamiliarity with product, substantial lead time for ordering and receipt, customs issues, the expense of product, and limited shelf life.

Roles in nutritional care
Generally, clinical staff determined what was provided to patients, including the type and consistency of food. Most KIs noted that there was no recording of information on dietary needs or preferences, only verbal sharing of information.
In the West Africa outbreak, many organizations had nutritional staff. However, nutritionists were not always allowed in ETUs, which was perceived as an impediment to improving the nutritional care of patients. Feedback from patients and health staff to nutritional managers was minimal. Only two KIs stated that they were able to monitor the content and quantity of food consumed by patients through a recording chart they developed. Their system monitored improvement or decline in appetite and provided insight into patient status with regard to nutrition: 'We knew who was deteriorating as they were not eating (anymore)'. Others tried but did not have the capacity for Feeding support Three KIs estimated that at any time during the height of the outbreak in West Africa, an ETU would have approximately 40% severely ill, 20% moderately ill and 40% ambulatory patients. Feeding support provided to patients varied by ETU, as well as by organization.
Patients in need of support were assisted by different types of staff: nutritional support staff, nutritionists, nursing aides, community members, community health workers, general infection prevention and controlstaff and sometimes nurses, doctors as they had increased access to the patients. Patients also supported each other. Some KIs mentioned it was often a collaborative effort at the meal times. Whoever was present assisted.
However, many KIs stated that in the beginning there was simply not enough staff to effectively provide support for feeding in light of the essential tasks to run the ETU. As the response matured, EVD survivors started to play a significant role in some ETUs as feeding support 'staff', often receiving a per diem. At least six out of the 12 organizations 'employed' EVD survivors to provide feeding support. While some EVD survivors found this role to be psychologically difficult, others appreciated the opportunity to be useful while delaying their return to their communities and the associated stigma. Two KIs described the construction of simple furniture, wooden triangle or 'pillows', which enabled patients to sit upright to help with the feeding process.

Malnutrition-screening and treatment
KIs stated that very few EVD patients were acutely malnourished upon admission. Some patients lost weight and possibly developed malnutrition during their stay. KIs stated that though most intended to perform anthropometric measurements at admission, this was infrequently implemented because of the 'no touch policy' or lack of time. At best, weight was measured. Some organizations only screened children at discharge for referral purposes or provision of food upon discharge. One organization only measured mid-upper arm circumference (MUAC) when children <5 years of age were suspected to be acutely malnourished. However, this was still limited by the need to sterilize tapes after each measurement. If needed, most organizations had the means to treat malnutrition as products used for the treatment of malnutrition were recommended for EVD patients in WHO's 2014 iGL. Overall, KIs stated that malnutrition was not a significant issue; however, no prevalence data were available, as no data were collected.
One KI questioned the use of therapeutic milks, normally used for treatment of SAM, for non-malnourished children: 'In DRC, every child is prescribed F75 or F100, we don't find this is appropriate. If a child is not malnourished it can eat normal food, and they want that! …There is no reason to take these specific products from other sites where they need it, away from community management of acute malnutrition programmes.
Therapeutic milks are still limited in availability./…../ And children do not necessarily like these milks, they want real food: cassava leaves, rice, chicken/meat and fufu.

Intake of specific nutrients
Four KIs (3 medical doctors, 1 nurse/nutritional manager) expressed significant concerns on hypo-albuminemia in EVD patients. The following captures the concern of practitioners: 'EVD patients are all low in protein (hypo-albuminemia). In DRC, we have constant lab analyses; we see the biochemistry constantly and see this disaster of lack of albumin.' 'We found that albumin is very low in EVD patients; almost no patients have normal albumin levels. We have no scientific evidence re albumin, but we know for a fact it is low (for all patients not only the very sick) and it has a link with nutrition. And we know high energy/protein intake can help.' 'They had low proteins, inadequate stores. IV fluid can make hypo-albuminemia worse esp.
because we don't give fluids with it, such as blood.' 'Patients have hypo-proteinemia-it will aggravate the fluid loss. We lost a few patients because they did not get the food they wanted and were not eating in a week. Many doctors want the patients to be fasting because of vomiting, diarrhea, risk of sepsis, but if they don't eat the drugs won't work enough. Those doctors want the GI tract to be empty, then they give IV and then re-evaluate but often forget that the patients need to eat, so for one week they won't get enough. This focus on fasting paralyzes the nutritional care and follow up of the patient. The new slogan here is aggressive rehydration, referring to intense fluid management-so we give a lot of fluid but patients still loose water within tissues, and hypovolemia is present, capillary leak and edema exists; in fact the patient is dehydrated. Therefore the patient needs food and needs the proteins to become reduce the hypo-albuminemia.' KIs mentioned the need for specific micronutrients, such as magnesium, potassium, zinc, and vitamins A and K. Concerning the 2018-19 DRC outbreak, 2 KIs reported that laboratory analyses are increasingly assisting in determining the dietary needs of patients.

Lessons learned according to the interviewed practitioners
There was a general consensus among KIs that the 2014 iGL, though initially useful for some organizations providing some kind of guidance, became less useful because of the focus on specialized products and did not evolve over time.
KIs were asked what they perceived as lessons learned based on the experiences they had on nutritional care for EVD patients in ETUs. Box 2 describes the most important reported lessons learned as expressed by the KIs on the nutritional care for EVD patients in ETUs.

Box 2. Lessons learned as expressed by the practitioners on the nutritional care for EVD patients in ETUs
Due to technical limitations, Box 2 is only available as a download in the supplemental Recommendations for future and ongoing outbreaks concerning nutritional care according to interviewed practitioners KIs were asked what they would do differently in the context of ETUs and working on nutritional support for EVD patients in a subsequent outbreak. Answers were expressed as recommendations for current and future outbreaks.
KIs recommended that greater concerted action is required to work collaboratively to develop context-specific guidance on nutritional care for EVD patients in ETUs. This guidance should address past lessons learned and include more practical guidance on implementation. Box 3 lists the main recommendations related to nutritional care for EVD patients in ETUs, according to KIs. professionals, but this was not related to care within ETUs itself [3].
While waiting for more scientific evidence to be generated on what nutritional support should be provided, there is a need for continuous monitoring of practice and generation of lessons learned. What may work in DRC may not translate to other outbreaks. However, from this investigation, it is clear that some nutritional care practices are universal and relate more to the disease than the geographical locations (e.g., preference for local foods by adults, the importance of feeding support).
Additionally, it is essential to obtain information on the perspectives of EVD patients themselves on the nutritional care they receive or have received. If future research findings are able to inform metabolically optimal nutritional care protocols, but the patients themselves do not consume what is recommended, treatment outcomes might be negatively impacted.
A step for the near future would be to share the recommendations expressed by the practitioners with a broader audience and assess the support and feasibility of their implementation. While good nutrition cannot cure EVD, it is believed that maintaining the nutritional status of the people affected could improve their response to the treatment.

Limitations
Limitations of this investigation included a sample size that does not allow the representation of all practitioners working in nutritional care in ETUs. However, most international organizations that have responded or are currently involved in EVD have been represented. Nevertheless, the experiences from Ministries of Health have not been documented through this investigation and therefore findings are skewed towards lessons learned recorded from international staff. This investigation did not specifically assess the challenges around infant feeding in the EVD context in ETUs. This topic merits more indepth studies. While findings from West Africa response outbreak cannot necessarily be extrapolated to the context in DRC, this investigation did not find striking differences in practitioners experiences between those that worked in DRC or West Africa. Availability of data and materials

Conclusions
The datasets generated and/or analysed during the current study are not publicly available because it concerns personal experiences from practitioners but de-identified data can be made available from the corresponding author on reasonable request. organisatons they worked for at the time they worked on EVD).