Ethiopian mothers' experiences with micronutrient powders: Perspectives from continuing and noncontinuing users

Abstract As part of a formative evaluation of a micronutrient powder (MNP) trial in Ethiopia that was organized according to a programme impact pathway model, we conducted in‐depth focused ethnographic interviews with caregivers of children between 6 and 23 months who had accepted to try “Desta,” a locally branded MNP. After stratification into two subgroups by child age, respondents were randomly selected from lists of caregivers who had received MNP from government health workers between 1 and 3 months prior to the interview date. Thirty women who were either currently giving Desta to their child (“continuing users,” n = 14) or had stopped feeding Desta (“noncontinuing users,” n = 16) were purposefully recruited from both urban and rural areas in the two different regions where the trial was conducted. Interviews were recorded, transcribed and translated, and coded for both emerging and prespecified themes. On the basis of identifiable components in the caregiver adherence process, this paper focuses exclusively on factors that facilitated and inhibited “appropriate use” and “continued use.” For “appropriate use,” defined as the caregiver preparing and child consuming MNP as directed, we identified four common themes in caregiver narratives. With respect to “continued use,” the caregiver providing and child consuming the minimum number of MNP sachets over a recommended time period, our interviews spontaneously elicited five themes. We also examined caregivers' perceptions related to problems in obtaining refills. Attention to caregivers' perspectives reflected in their narratives offers opportunities to improve MNP utilization in Ethiopia, with potential application in other social and cultural settings.


2004; Central Statistical Agency
. The Global Alliance for Improved Nutrition (GAIN) and Concern Worldwide supported the Federal Ministry of Health (FMOH) in Ethiopia for a trial to deliver micronutrient powder (MNP) through the FMOH's Health Extension Program, which deploys government-salaried female health extension workers (HEW) to provide primary health care services in rural communities (Federal Ministry of Health: Health Extension and Education Center, 2006;Federal Ministry of Health: Health Extension and Education Center, 2007). MNP, a single-dose packet of dry powder containing lipid-encapsulated iron and other micronutrients that can be sprinkled onto any semisolid food (Zlotkin et al., 2005), is recommended by the WHO to address anaemia and iron deficiency in young children (WHO, 2011(WHO, , 2016. To inform the scale-up of MNP delivery through the FMOH's Health Extension Program, a theory-driven formative process evaluation was conducted using a mixed-method approach (Rawat et al., 2013).
Both survey methods and focused ethnographic study (FES) methods were employed to describe and understand how the intervention worked, with the goal of determining what aspects of programme delivery and adherence went well; identifying where problems occurred, the magnitude, and causes of problems; and providing knowledge and insights to support potential for impact and inform expansion of the intervention. For MNP interventions, adherence to MNP recommendations involves three main elements: initiation (caregiver determination to feed MNP and starting to do so), appropriate use (caregiver preparing and child consuming MNP as directed), and continued use (caregiver providing and child consuming the minimum number of MNP sachets over a recommended time period). Figure 1 identifies caregiver behaviours for the three elements, based on the specific instructions provided in the Ethiopia programme. This paper presents key findings from the FES that was conducted with caregivers who participated in the programme. The purposes of this paper are (a) to contribute to the growing body of empirical data on household responses to MNP interventions, with an emphasis on caregivers ' perspectives, and (b) to contribute to generating theories in implementation science of nutrition interventions, particularly with respect to interventions that are directed to vulnerable infants and children. In Section 3, we present the main themes that emerged from our ethnographic analysis of two components of adherence to MNP recommendations, specifically (a) appropriate use and (b) continued use. We focus on these two aspects of adherence because the cross-sectional survey results of programme outcomes showed that virtually all caregivers who received the MNP initiated feeding it to FIGURE 1 Programme impact pathway depicting household delivery system of micronutrient powder. Caregiver behaviours based on key messages communicated as part of Ethiopia programme. Adapted from  Key messages • Many mothers said their child differentiated between foods with MNP and foods without MNP.
• When health workers advise mothers about side effects, it moderates the mothers' reactions when those side effects occur.
• Findings call attention to the need for support systems for caregivers to manage negative child reactions to food prepared with MNP.
• Our study highlights the need for further research on mother-child interactions around food, as well as the development of taste preferences at different ages and stages of development, particularly in resourceconstrained environments. their children (98%; Tumilowicz et al., 2019). However, a much smaller proportion of caregivers who initiated use of MNP reported using it within 14 days prior to the survey (49%), making that one of the largest challenges for potential impact of the intervention.

| MNP integration into FMOH Health Extension Program
The study, which took place in Amhara and Tigray Regions, was designed to assess both programme delivery and household delivery of MNP, building on a model that was initially developed by Alive & Thrive (A&T) and implemented through the FMOH's Health Extension Program (Kim et al., 2015;Kim et al., 2016). The locally branded MNP, "Desta" (the Amharic and Tigrinya word for "joy, happiness"), together with behaviour change communication (BCC) activities, was delivered to approximately 71,000 children 6-23 months in six woredas in Amhara Region (Kobo, Gubalafto, Delanta, Dessie Zuria, Kalu, and Bati) and five woredas in Tigray Region (Saesi Tsaedaemba, Mereb Lehe, Werie Lehe, Gulomahda, and Ahferom). Building on the A&T foundation, GAIN created guidelines, a training manual, job aids, and materials for caregivers, which included information on the composition and purpose of Desta, instruction on food preparation with Desta, and contraindications and potential side effects of Desta. Implementation materials also included procedures for performance monitoring, supply chain management, and delivery of BCC activities protocols. GAIN trained "master trainers" from CWW and FMOH in October 2015, who subsequently trained CWW staff, FMOH health workers, and HEW from April to July 2016.
HEW or health centre staff provided a box containing 30 Desta sachets every 2 months to caregivers at the health post, health centre, or during home visits. At the initial distribution, caregivers were provided with a "Desta child card" that could be used to keep track of distribution dates. The card included appointment dates for caregivers to return to receive a new box of Desta, which was scheduled 2 months after the date of the current distribution. HEW and health centre staff were also trained to communicate the key messages shown in Table 1 2007). Average household size is 4.3, and 92% are from the Semitic-speaking Amhara ethnic group. Agriculture is also the main occupation. Interviews in Tigray were conducted in six rural woredas and one urban area, and in Amhara, they were conducted in two rural woredas and one urban area.

| Design and sampling
FES is a mixed-method research approach that is designed to answer specific sets of questions from the perspective of the target population that are important for various decision-making activities, including improving programme delivery and utilization of interventions (Pelto & Armar-Klemesu, 2015;Pelto, Armar-Klemesu, Siekmann, & Schofield, 2013;Pelto, Dufour, & Goodman, 2013). FES draws from several qualitative and quantitative ethnographic methods. In this study, we used in-depth interviews with open-ended, guided questions, which were administered with extensive probing to expand and interpret the initial responses. Our interview protocol contained four modules that covered a range of issues including demographic and socio-economic status (SES) characteristics; experiences receiving Desta and participating in BCC activities; motivation for initiating and continuing to feed Desta; motivation for discontinuing Desta; preparation of food with Desta; perceived changes in food mixed with Desta; perceived reactions of child when fed food with Desta, initially and over time; caregiver responses when a child refuses to eat food mixed with Desta; beliefs about how to keep children healthy; beliefs about the benefits of Desta; perceived changes as a results of the child consuming Desta; and how programme delivery of Desta could be improved . Interviews were conducted in mothers' homes from November to December 2016 and averaged 2-3 hr.
Respondents were randomly selected from lists of mothers of children 6-23 months of age who had been registered by HEW as receiving Desta within 1-3 months prior to the interview date. The sample design required filling respondent categories based on subgroups of the 6-to 23-month age range. We aimed to interview six mothers of children 6-11 months of age and nine mothers of children 12-23 months of age in each region. We also sought to interview caregivers who reported currently using Desta (referred in this paper as "continuing users") and women who had initiated using Desta but had stopped using it before the point at which they were contacted for an interview (referred to as "noncontinuing users"). Table 2 presents a summary of selected sociodemographic characteristics of the 30 respondents. All interviewed caregivers except for one who was the grandmother were the biological mother of the index child. Therefore, in the rest of this paper (and in the title), we refer to the respondents as mothers. The sample was relatively homogenous in important sociodemographic characteristics, such as education, occupation, and household composition. Most mothers said they were engaged in farming or in a combination of household responsibilities and farming, and some reported engaging in petty trade. Depending on the season, mothers spend many hours each week away from the house. The majority of respondents lived in nuclear families, whereas the rest lived in extended families of varying composition, including fathers-in-law, mothers-in-law, brothers-in-law, older generation mothers, fathers, and sisters, or grandchildren.
The study protocol was approved by the Amhara National Regional State Health Bureau Ethics Review Committee and the Government of the National Regional State of Tigray, Bureau of Health Ethics Review Committee. GAIN requested and received permission from the FMOH to undertake the evaluation. The FMOH communicated its permission for the evaluation to the regional health bureaus, zonal health offices, and woreda health offices. After the nature of the study was fully explained to the participants in their local languages, written informed consent was obtained from all respondents (for nonliterate respondents, a thumbprint was obtained in the presence of witnesses). There were no refusals. Respondents were informed that (a) they had right to withdraw from the study at any time, (b) individual study results would be treated confidentially, and (c) their responses would not affect any food or nonfood distributions.

| Data analysis
Thematic analysis of text (using the transcribed and translated recordings of the interviews) followed basic principles of qualitative text analysis (Miles, Huberman, & Saldaña, 2014). We used QSR International's NVivo Version 11 qualitative data analysis software to organize data and facilitate coding (QSR International, 2002). Data analysis of the caregiver interviews was conducted as a team through sharing NVivo project files, meetings, and memos summarizing actions taken regarding coding and emerging themes. First, double coding was completed on a sample of transcripts. From those initial transcripts, the team developed a common coding framework and assessed intercoder reliability. Coding then continued for the remaining transcripts with ongoing input from members of the study team. During weekly study team meetings over the course of 3 months, we adjusted the coding framework, identified emerging themes, and created matrices of quotations that were used to compare themes across transcripts.

| RESULTS
In this paper, we are concerned with two key components of caregiver behaviour with respect to MNP adherence: "appropriate use," the caregiver preparing and child consuming MNP as directed, and "continued use," the caregiver providing and child consuming the 30 MNP sachets within a 2-month period. We begin with "appropriate use," in which we identified and coded four themes in the mothers' • Desta is not medication.

Benefits
• Improves child's health and development.
• Improves the nutrition quality of child's food.

Administration regimen
• Give Desta every other day. The child needs to have regular consumption to benefit from Desta.
• One box contains 30 sachets. Give all of them to the child within 2 months. Then you will receive a new box until your child reaches the age of 2 years.

Preparation
• Wash your and your child's hands with soap and clean water before cooking and feeding child.
• • Desta should be mixed with a small amount of food that the child can finish during one feeding session.
• Pour the whole sachet of Desta into the child's food and mix well. Content of one whole sachet should be mixed with one meal for one child. It is not recommended to split one sachet into portions given over several meals.
• Feed child immediately or within 30 min of mixing Desta with food.
• Do not share the food mixed with Desta with other household members.
• Do not feed more than one sachet of Desta per day.
• Store Desta in a clean, cool, and dry place.
Side effects • There are no major side effects associated with Desta.
• In the first few days of taking Desta, the child's stool may be darker than usual. This is completely normal and caused by the iron contained in Desta. Continued use of Desta is advised.
• During the first days of taking Desta, the child may have softer stools, mild diarrhoea, or a mild form of constipation, which usually passes in a period of 4-5 days. This does not happen to all children. This is also normal, and it should not be cause for concern.
• If diarrhoea is severe, bloody, or with mucous, take the child to the health facility for care as you would have without concurrent use of Desta.

Contraindications
• Children suffering from severe acute malnutrition should not receive Desta. Desta is unsuitable for treatment of severe acute malnutrition and children should receive the usual ready-to-use therapeutic foods (RUTF) instead. Desta can be used safely after a child is rehabilitated.
• Children suffering from malaria should temporarily stop consuming Desta. Once malaria treatment has been completed, children can continue to consume Desta.
narratives. These results are followed by the presentation of five themes pertaining to "continued use" that emerged spontaneously in the caregivers' narratives. The last part of the results section presents the findings from narratives that were obtained from a general question near the end of the interview, which was phrased as "Could you tell me how you collect a new supply of Desta?"

| Appropriate use
Appropriate use refers to mothers preparing and children consuming MNP as directed in programme guidelines that are intended to be communicated by HEW. Mothers overwhelmingly reported following recommended preparation instructions, and these instructions were followed by both continuing users and noncontinuing users ( In theory, MNP is a bland powder that can be added to any food, provided it is not a liquid and that the food is not so hot as to melt the lipid layer that protects the iron from interacting with the food. However, most of the respondents (n = 24) reported organoleptic changes, including alterations in colour (n = 16), taste (n = 12), smell (n = 10), and/or texture (n = 7). Eleven mothers (seven continuing users and four noncontinuing users) noted some of these organoleptic differences only occurred if certain factors were at play. For example, some respondents said there would not be any change if Desta was "well mixed into the food." Others described changes if Desta was mixed into food that was still being cooked or hot or if the food "stayed too long." Mothers' awareness of these changes, and the accompanying Living with father of index child (n, yes) 15 13 Is biological mother of index child (n, yes) 14 15 Schooling of caregiver (n) Illiterate 5 4 Primary education (Grades 1-8) 7 9 Secondary education (Grades 9-12) 3 2 Occupation of caregiver in addition to household duties (n) Farming 9 5 Petty trading 2 3 None 4 6 Household size 4.3 ± 1.0 (3,6) 4.6 ± 2.1 (2,9) Child characteristics Age of index child (n, month) 6-8 2 2 9-11 4 3 12-23 9 8 Sex of index child (n) Male 6 10 Female 9 5 Breastfed (n, yes) 15 15 Currently being fed food with Desta (n, yes) 8 6 Mean and standard deviation; range in parentheses. bOne caregiver was the grandmother of the child.
preparation explanation, was attributed by them to their attendance at a cooking demonstration, by experiencing the change first-hand, or from verbal education they received from the HEW: If it is added on hot food it will bring change. Twelve mothers said that their child initially did not seem to like the food mixed with Desta (e.g., made a "bitter face" or a face "as if she ate something sour") or refused to eat (e.g., turn their head away from the food) but eventually "got used to it" or accepted it. A few mothers offered (or were prompted to provide) an explanation as to why they thought their child initially did not seem to like it. One said because her child's first experience with Desta was also his first experience with complementary foods, so perhaps, "[his unpleasant expression] could be because of his first-time exposure to food in general." Another mother thought her child initially did not like it simply "because it's a new thing." One suggested that her child was not taking any food well when she first started feeding Desta, but the Desta improved her child's appetite so that the child was now taking all foods, including those mixed with Desta.

| Theme 4. Mothers reported various strategies and techniques to encourage children to eat food with Desta
Almost all mothers (n = 26, thirteen continuing users and 13 noncontinuing users) described feeding techniques or problem-solving strategies they used to encourage their child to eat food with Desta.
Some of these strategies were employed proactively, and some were reactive. Preparing Desta with a variety of foods, or trying Desta with different foods if a mother perceived her child did not like it with certain foods, was the most commonly reported strategy (n = 11). This was followed by descriptions of playing with or "fooling" the child (n = 8) and then by force-feeding (n = 5) or feeding Desta at a time the mother knew her child would be hungry (n = 4). Other less common strategies included waiting a bit of time and then trying again, either hand-feeding or letting the child self-feed, and making the food "tastier." Two continuing users also described mixing in only half of a Desta sachet into food until their children got "used to it," perhaps for the first week or so. Although this practice is not in line with the instruction that the mother should mix the entire contents of the sachet into the food, the mothers' technique may have facilitated the acceptance of foods with Desta by these two children. The most commonly reported strategy for ensuring children ate all the food was to mix the Desta in with the amount of food the mother thought her child could finish eating at one time (n = 12).

| Continued use
The final step in the adherence process is the continued use and feeding of MNP. Themes 5 through 9 relate to continued feeding versus stopping based on the narratives of both continuing and noncontinuing users. Continuing users reported feeding Desta between 1 and 9 months before the interview. As shown in Table 4, nine of the 16 noncontinuing users reported feeding Desta to their  3 Instructions that mothers reported following regarding preparation and feeding of food mixed with Desta (n = 30, including 14 continuing users and 16 noncontinuing users) Continuing user (n, yes) Noncontinuing user (n, yes) Wash your and your child's hands with soap and clean water before cooking and feeding child 9 6 Only add Desta to soft or mashed food. Do not mix Desta with liquids 13 15 Do not add Desta during cooking or to hot food. Mix Desta after food has cooled 11 14 Desta should be mixed with a small amount of food that the child can finish during one feeding session 12 14 Add entire content of sachet at one meal. Do not split sachet over several meals 13 16 Feed child immediately or within 30 min of mixing Desta with food 13 16 Do not share food mixed with Desta with other household members 11 14 Do not feed more than one sachet per day 13 14 Give Desta every other day 8 7 Give 30 sachets of Desta to the child within 2 months 3 3 Store Desta in a clean, cool, and dry place 1 2 No perceived benefit "They have told us to go and get Desta but we didn't go. Even though they told us to collect it, I said to myself, the rest of my kids before him grow up just fine without Desta." 6. Amhara 18 4 Child refused food with Desta "At first she ate food with Desta properly, but one day she refused the food and I thought she got tired of it. I also tried to give her after few days but she still refused food with Desta. I tried to give it to her for days and she still hates it. I stopped giving her when there are still four sachets left." 7. Amhara 18 2 Missed refill appointment, did not know how to obtain more Desta "I was not in town on the day Desta is distributed, I was away. When I returned I didn't know when the next appointment was, they didn't tell me when it is given or when I should come." 8. Tigray 7 <1 Child refused food with Desta "Initially she was taking Desta very well but then she disliked it. She preferred breastfeeding at that time. It could be because of the amount of food I used to mix with Desta. I probably decreased the size of the food I used to mix with the Desta … At the beginning, she was taking it without hesitating but over time she decreased and resisted to eat food mixed with Desta. I am not clear how she could detect it was not good whether it is by taste or smell. For me, I did not sense any smell with it." 9. Tigray 6 <1 Child has poor overall appetite for foods; child refused food with Desta "My child has a poor appetite … He is not good with foods other than breastmilk! … But when I started feeding I saw a yellow colour in the and I was astonished by the colour I saw in the food, and I questioned what it could be but I gave to my child and he ate a small amount. But later on the totally refused eating and consequently, I stopped getting Desta from the health post." 10. Tigray

<1
Child has poor overall appetite for foods; child refused food with Desta "She took it easily, for a while but, later refused to eat, even food without Desta.
I tried to change the food which Desta will be added but she totally refused." 11. Tigray 8 6 Child had diarrhoea and mother scared that giving Desta would continue the diarrhoea "I stopped giving Desta because my child was sick with diarrhoea repeatedly. The HEW treated my child and counselled me that this is quite normal, it is temporary and the diarrhoea will stop by itself so I should not worry. But I was scared of the diarrhoea so I stopped giving DESTA." 12. Tigray 8 3 "At that time he was not healthy, he was rejecting any type of food including that with Desta. I took him to the health center get treated and he became well, so (Continues) Waiting to restart Desta until child's appetite returns after illness; too busy to obtain refill f rom HEW because of his illness he stopped taking Desta but only for that specific time.
But I will wait some time to continue Desta until his appetite reverses … Desta is useful it does not cause illness, it's my own problem, I failed to bring another Desta because I was so busy doing farming for the last three months." 13. Tigray 7 6 Perceived Desta as aggravating diarrhoea from existing stomach infection "I do not think his illness is only because of Desta because even when I stopped Desta his appetite is still low and also he has some diarrhoea though it is not as severe as before. I believe that he has bacteria in his abdomen and this is being aggravated by the introduction of Desta." 14. Tigray 16 4 Child refused food with Desta; perceived child appetite decreased because of Desta "One day I tasted it to check why she is refusing it and realized that it has some bitter taste. I realized that it is because of the bitter taste that she is refusing … the HEW taught us that Desta increases the appetite but in my child the reverse happened. She has significantly decreased her appetite once I started feeding Desta.
They told us that her stool colour might be changed. I have seen and accepted that it is due to Desta. But when her appetite gets lower and she refuses Desta, I just stopped giving her." 15. Tigray 8 10 Away from home without appropriate food to mix Desta; too busy to obtain refill from HEW "Initially I was taking Desta because I had enough time for a refill and I was giving Desta comfortably and I had time, but now I am one of social justice committee and the circumstances that I stay in meeting the whole day, and I took my kid with me and I do not consider to add Desta in food because most of the time the food I took for my child when I have meetings is dry food 'kita' and can't get time to add and bring Desta." 16. Tigray 8 <1 Child refused food with Desta; organoleptic changes to food with Desta; perceived Desta caused diarrhoea and dark stool; perceived child's health diminished; too busy to obtain refill from HEW "There were many reasons [for stopping Desta]. My child totally refused foods prepared with Desta. Most of the time I dumped the food prepared with Desta because it changed to a yellow colour and the taste become offensive … The other reason was that the child had diarrhoea and at the same time the stool became darker and darker … Totally the health of the child has diminished and I decided to stop feeding because I assumed that the reason for the change in the health of my child was because of Desta. Meanwhile, the last reason that made me stop feeding Desta was because I was busy and I was not able to visit the health post during the appointment time and then I stopped feeding." children for 3 months or longer before discontinuing its use. The most frequently cited reasons for discontinuing Desta were child refused food with Desta (n = 9), no perceived benefit (n = 4), trouble obtaining refill (n = 4), and diarrhoea (n = 3). Mothers often mentioned several challenges rather than just one cause for discontinuing use of Desta.
3.2.1 | Theme 5. Perceptions of positive changes in their children are reported commonly by caregivers who continued to use Desta Closely tied to learning about Desta's benefits for children, which emerged strongly in mothers' decisions narratives about why they decided to try preparing and feeding foods with Desta to their child, is their articulation of observing benefits when they used it. Two thirds of mothers (n = 20, twelve continuing users and eight noncontinuing users) reported perceiving a variety of positive changes in their child after starting Desta (Table 5). The following quotations from continuing users in Amhara describe various aspects of this theme: At first, she did not take food well, but since I started feeding her Desta her appetite has improved … after I started using Desta she is eating foods well.
Yes, she became strong, she knows her surroundings, she is active. She used to have a stomach ache before, but now she is free from that. She used to get sick, but now that doesn't happen anymore. She is very fine.
When they first gave me Desta I didn't think it was useful, until I noticed a change in her appetite … I felt happy, because she gained weight and her appetite increased.
I was happy when I saw that it makes a child gain weight and build the body. I said to myself that the government must bring this thing, because it is useful.
Four mothers explicitly articulated the role that positive changes played in their continuing to give Desta. The following quote from a mother in Tigray illustrates this idea: I didn't see any complication and the response was positive, like increasing my child's appetite, I decided to continue and I never missed my appointment so far.

| Theme 6. Positive statements about Desta from relatives or neighbours encouraged mothers to maintain the practice of giving it
Just as the influence of others played a role in mothers' decision to try Desta, spontaneous reports of positive comments from others appeared as a theme related to continuing use. This theme emerged not only in the narratives of women who were continuing users (five mothers) but also in noncontinuing users (three mothers). Women Child's health deteriorated 0 1 Other positive changes: "Make child normal," "child became beautiful," "child can differentiate people," and "child is alert, knows surroundings." discussed how husbands, neighbours, and even older children encouraged and supported them to keep feeding Desta: We will ask each other whether my child is taking Desta or not? I also ask her whether her baby is taking Desta. The theme of failure of problem-solving strategies appeared primarily in the narratives of noncontinuing users. A few of the noncontinuing users expressed their frustration or unhappiness that their attempts did not work. For example, I tried all my best … since it is very essential for her … I wanted to feed her … but she refused … I tried … but … she refused.-Noncontinuing user from Amhara Eight noncontinuing users and one continuing user described discussing the problem of their child's refusal of foods mixed with Desta with their HEW. One continuing user received counselling from her HEW on how to better prepare foods with Desta. Six of the eight noncontinuing users also reported receiving advice about what to do in the face of child refusal. However, according to their narratives, the suggestions these six mothers received were primarily to "keep trying" or to "try with different foods." An idea expressed by three noncontinuing users indicates that there was also a feeling, even among those who described problem-solving techniques, that nothing could be done if a child refused the food. One noncontinuing user also tied this sentiment to her explanation about why she did not seek help from the HEW: Interviewer: "So if you felt concerned this much, why didn't you try to make some more efforts? Or take her to [the HEW]?"Caregiver: "No, I didn't take her. What can the HEW do, if she once refused to eat the food?" 3.2.5 | Theme 9. The perception of negative effects associated with consuming Desta emerged as a theme for discontinuing its use The theme of negative effects of Desta was identified in some of the narratives from mothers. Five mothers brought up the problem of loss of appetite and one mother said that her child experienced an overall decrease in health. The following quotes from two noncontinuing users in Tigray illustrate this theme: The HEW taught us that Desta increases the appetite of the children, but with my child the reverse happened; she has significantly decreased her appetite, once she started eating Desta added to food … when her appetite got lower and refused Desta, I just stopped giving it to her.
Totally the health of the child has diminished and I decided to stop feeding because I assumed that the reason for the change in the health of my child was because of Desta.

| Difficulty obtaining a new supply of Desta
The findings in this section were obtained from a general question: "Could you tell me how you collect a new supply of Desta?" By way of background for this question, the programme initially gave mothers a box of Desta containing 30 sachets, together with use and dosing instructions and the Desta child card. The child card included appointment dates for mothers to return to the health post to receive a new box of Desta, which was scheduled 2 months after the date of the current distribution. In the next paragraphs, we present the results from this broad question.
Only half of mothers in the study spontaneously reported that the instructions from their HEW included the information that they should return to the HEW to receive more Desta 2 months after receiving the first box. Six mothers reported that they had returned to their HEW to obtain more Desta 2 months after they received the first box. However, this number reflects the fact that some of the mothers in the sample had not yet reached the end of their initial 2 months at the time of the interview, and some of the noncontinuing users had already decided to stop giving Desta for reasons that had nothing to do with problems in obtaining the next box.
Summarizing the narratives that were elicited with the basic question, we found that several mothers expressed confusion about how to get a refill of Desta or said they had experienced problems. A few mothers did not know when or how to get a refill; several of them give twice on the same day.

| DISCUSSION
The results presented in this paper were obtained through in-depth interviews, using a focused ethnographic methodology. The format of the questions and the interviewing method were designed to maximize the opportunities for mothers to articulate their experiences with MNP, their interpretations of its purposes and effectiveness, their management of the problems they encountered, their understanding and interpretations of their interactions with the programme delivery system and its representatives (the HEW), and the challenges and opportunities they experienced in the adherence process.
Together with the results of a larger survey , which provided the opportunity to conduct a quantitative analysis of the adherence process, the mixed-method approach that was used to conduct the formative evaluation of the Desta MNP project in Ethiopia is intended to contribute to future nutrition interventions to improve infant and young child nutrition, as well as to provide guidance for the programme in Ethiopia.  (Nicklaus, 2009). This inference is supported by survey data analysis, which showed among children whose mothers initiated feeding Desta that children aged 12-17 months and 18-23 months were 32% (P < 0.001) and 38% (P < 0.001) less likely respectively to have been fed it in the 14 days prior to the interview as compared with children aged 6-11 months-indicating continuing use was more problematic among older children . Most studies on the development of taste preferences and caregivers' role in encouraging children to eat less desired foods come from high-income countries (Mura Paroche et al., 2017). Our study highlights the need for further research on the development of taste preferences and mother-child interactions around food with and without MNP, at different ages and stages of development in resource-constrained environments. Moreover, it is unwise to assume that all infants and children will readily accept foods with MNP for long periods of time.
With respect to the challenges that mothers faced to overcome child refusals of food with Desta, one is reminded of the earlier insistence of nutrition and health professionals that "breastfeeding is natural, so women should not have problems with it," and "mothers just need to get on with it" (Shannon, O'donnell, & Skinner, 2007). Fortunately, the development of lactation counselling and lactation counsellors has created an environment in which many women today have access to the help and support their need to sustain breastfeeding (Patel & Patel, 2016). By analogy, the results of this study, including not only the findings concerning reports of negative child reactions but more importantly the fact that those negative reactions lead to cessation of giving MNP, call for attention to developing support systems for women to address negative child reactions to food prepared with MNP and to instituting them as an integral part of programs to improve child nutrition through MNP.
The reported experiences with side effects are, in our view, another important finding from the study. It is very encouraging to see that when health workers advise mothers about side effects, to the level of being very explicit about potential observable symptoms, it moderates the mothers' reactions when those side effects occur. It was a wise decision on the part of the programme to include these in the BCC messages for mothers.
It is also encouraging to note that the programme provided nuanced information to mothers, in the form of suggesting there are signs and symptoms "your child may experience," laying the groundwork for differential reactions, as contrasted with a blanket statement about reactions. This tacit message about individual differences accords with mothers' experiences in other domains of health and development. It is, in that sense, a "meta-message" that can be simultaneously reassuring for women and open future lines of dialogue between health providers and mothers.
Regarding mothers' spontaneous reports of side effects and the finding that only some respondents commented on them, we note that this differential reporting may reflect actual differences in observable symptoms or differences in mothers' attention to and sensitivity to their children's responses and behaviours. Without observational studies, this question, clearly, cannot be answered.
Concerning the problems women reported about getting refills, the narratives provide evidence that confusion about what to do was a significant barrier to the final step in adherence-continued use. We note that the set of messages HEW were given to convey to mothers about this aspect of the programme was inadequate, and mothers' confusions about the process may stem from that programme-originated problem. This interpretation gains plausibility because the messages related to dosing, which were well developed, were understood and followed by half of the women we interviewed.
At the same time, we note the appearance of other "refill barriers," particularly constraints on women's time to acquire refills, and absence from their home base, as factors that also affected continuing use of MNP.
Finally, it is important to point out that all the data presented in the paper represent the perspectives of mothers who tried Desta.
The paper does not address the issue of factors that prevented mothers from trying Desta after being presented with the product or learning about it, or who did not learn about it, even though they lived in the project catchment areas. On the basis of the neutral or positive insights about the mother's experience with programme delivery and the overwhelmingly positive response to learning about the product itself and its benefits for children, we surmise that the main limiting factor for not trying Desta was not encountering the programme delivery, the BCC activities, or the product at all. This inference is supported by the survey results, which showed that among mothers who received Desta, 98% tried it . Thus, we conclude that ensuring mothers learn and receive Desta is critical because they are very likely to try it initially if they are exposed to the product and learn about its use and benefits.