ACCEPTABILITY OF OUTPATIENT READY-TO-USE FOOD-BASED PROTOCOLS IN HIV-INFECTED SENEGALESE CHILDREN AND ADOLESCENTS WITHIN THE MAGGSEN COHORT STUDY
Scaling up of antiretroviral treatment (ART) has dramatically reduced the mortality of HIVinfected children worldwide and thus allowed HIV infection to become, to some extent, a chronic disease.1 However, HIV infection has the potential to induce growth failure in children or to increase the severity of preexisting undernutrition. Consequently, severe acute malnutrition (SAM) and moderate acute malnutrition (MAM) remain common in HIV-infected children and adolescents, even when on ART,2,3 and are strongly associated with death.4 Since 2007, the United Nations has recommended ready-to-use therapeutic food (RUTF) for the outpatient nutritional rehabilitation of children with uncomplicated SAM,5 and most African countries adopted national guidelines for RUTF-based nutritional rehabilitation of children <5 years, including those with HIV. In the management of MAM, ready-to-use supplementary food (RUSF) is recommended as an option as well as other fortified foods and enhanced family diet.6 Provision of RUTF and RUSF, although originally developed for pediatric use, has become a widely used strategy among HIV-infected adults through foodby-prescription programs in low-income settings.7,8 Data on the acceptability of RUTF in malnourished outpatients are sparse and mixed. Qualitative studies conducted in Southern Africa among HIV-infected adults9,10 reported poor overall acceptability of RUTF. The RUTF was mostly found unacceptable by pregnant and lactating Bangladeshi women,11 whereas its organoleptic qualities were rated as good in another quantitative study from Vietnam carried out in HIVinfected children <5 years and adults.12 Data on acceptability outcomes of RUTF or RUSF therapies in undernourished HIV-infected older children and adolescents are unavailable. The ANRS 12279 MAGGSEN Pediatric Cohort Study aims to investigate the growth and metabolic disorders among HIV-infected Senegalese children. Recently, we reported a high prevalence of acute malnutrition in children aged 5 to 16 years, after a median duration of 4 years on ART.13 We implemented a pilot intervention within the cohort for outpatient nutritional rehabilitation using imported RUTF (Plumpy Nut; Nutriset, Malaunay, France) and RUSF (Plumpy Sup; Nutriset). Both ready-touse foods (RUFs) are energy-dense lipid pastes made from peanut butter, oil, sugar, and a high vitamin and mineral supplement (RUTF also contains milk powder, while RUSF contains soy proteins) packed in 92 g individual sachets providing 500 kcal.14 In the absence of specific and practical guidance for the nutritional rehabilitation of HIVinfected children 5 years, the intervention protocol was designed in compliance with Senegalese recommendations15 and adapted from the World Health Organization (WHO) framework for the nutritional care of HIV-infected children aged 6 months to 14 years.16 The objectives of the present analysis are to assess the acceptability of this intervention in HIV-infected children and adolescents through their description of organoleptic appreciation of RUF and compliance with the protocol modalities.
Methods Study Setting The study was carried out in Dakar from April to December 2014 within the ANRS 12279 MAGGSEN Pediatric Cohort Study. Full details of the rationale for the MAGGSEN Study and methods are provided elsewhere.13 Briefly, the MAGGSEN Study included HIV-1–infected children aged 2 to 1.5.15 Throughout the study, the social worker provided incentives, advice, and explanations to children and caregivers whenever needed. Population and Data Collection The study included HIV-infected children aged 6 months to <18 years, enrolled in the MAGGSEN Study or attending the study clinics, presenting at their follow-up visit with MAM or SAM without medical complications, and who succeeded at the appetite test. This was a convenience sample as children were enrolled during a 6- month period (April to September 2014) up to the financial and human resource limits for the cohort. At enrollment, a clinical assessment was performed, and patients’ socioeconomic and therapeutic data were recorded from the patients’ files. The Household Food Insecurity Access Scale—a list of 9 specific questions about the availability and accessibility of foods and related worry for the household during the previous month—was used to assess food insecurity.19 This access scale generates a 4-class variable of food insecurity— none, mild, moderate, and severe. In this study, we used a binary indicator—food-secure (none) versus food-insecure household (mild to severe). At each follow-up visit, the unused RUF sachets from the previous provision were collected and registered for consumption monitoring. Children were weighed in their underwear to the nearest 100 g with an electronic scale. At enrollment and at target weight visit, height was measured twice to the nearest millimeter with a roll-up measuring rod, and the mean was used in the analyses. Anthropometrics were measured using Seca devices (Hamburg, Germany). At week 2, a structured questionnaire covering the organoleptic perception of RUF, the mode of consumption, the associated stigma, and potential sharing within the family was administered to children 5 years and/or to the caregiver.
Data Management and Analysis Children who reached their target weight are referred to as ‘‘recovery.’’ Those who experienced medical complications after their enrollment, failed to gain weight during 6 consecutive weeks, or refused to consume the RUF were discontinued from the study for clinical management and are referred to as ‘‘failure.’’ Those who discontinued their participation in the study are referred to as ‘‘default.’’ Height and weight were converted into WHZ, BMIZ, and height-for-age z score using WHO reference values.20,21Children’s characteristics were compared by type of malnutrition (SAM orMAM) using the Fisher exact andWilcoxon rank sum tests as appropriate and by outcome in the study (recovery, failure, or default). Consumption of RUF was assessed at each follow-up visit by subtracting the number of unused sachets from the number prescribed. Acceptability data were compared, using the Fisher exact test, by malnutrition status, by age as a categorical variable (<12 years, 12 years), and by outcome in the study. Continuous variables are expressed as median with interquartile range. Differences were considered statistically significant at P < .05. All statistical analyses were performed using SAS version 9.3 (Cary, North Carolina). Results Size and Characteristics of the Study Population During the inclusion period, 43 children, perinatally HIV infected, presented at the study clinics with MAM (n ¼ 24) or SAM without complications (n ¼ 19) and were enrolled in the intervention; 39 were on ART (3 drugs; Table 1). All passed the appetite test at the first attempt, except 1 who succeeded at the third. Most (38 of 43) were enrolled in the MAGGSEN Study. Median age was 12.2 years (9.3-14.8), and only 2 of the children were <5 years. In most cases, the mother was the caregiver; however, 7 of 24 MAM children and 9 of 19 SAM children were permanently placed with a grandmother (n ¼ 8), aunt (n ¼ 6), cousin (n ¼ 1), or neighbor (n ¼ 1). Of the 43 enrolled children, 25 (58%) recovered their weight; 9 failed to gain weight and/or to consume RUF and were discontinued from the study to be initiated on the first- (n ¼ 1) or second-line ART (n ¼ 8), of the latter, 2 also initiated an HIV status disclosure process; and 9 defaulted to follow-up due to distance from the clinic (Table 2). Children with MAM (n ¼ 18 of 24) were more likely to recover their weight than their SAM counterparts (n ¼ 7 of 19; P ¼ .01). Their nutritional rehabilitation was also faster, 54 days (31-90) in MAM children versus 114 days (69-151) in SAM children (P ¼ .02), but their daily weight gains were similar, 1.4 g/kg (1.0-1.9) versus 1.3 g/kg (1.1-2.8), in MAM and SAM children, respectively. Acceptability Outcomes Overall acceptability was assessed by questionnaire in 41 child–caregiver pairs at the 2-week visit. Organoleptic appreciation. Overall, 80%, 76%, 68%, and 68% of children initially rated the RUF color, taste, smell, and mouth feeling as good, respectively (Table 3). However, 12 of the 41 children reported that they felt disgusted by RUF, and there had been occasions when they simply refused to eat it. Intake modalities. In most cases, the caregiver was responsible for supervising RUF intake, and the child needed encouragement to eat (Table 4). Half of the children ate the RUF in place of breakfast. The main mode for eating the RUF was directly from the sachet; however, some children/ caregivers used other feeding strategies like diluting the sachet into water to make a gruel or spreading it over a piece of bread. Perceptions and behaviors surrounding RUF. Most children (n ¼ 34) had understood that they were prescribed RUF due to their malnutrition status, 3 because they were sick, and 4 did not know the reason; 38 of the 41 perceived RUF as a drug. At least 15 of the 41 children stated they used to hide from others when eating RUF. Reasons given to explain this embarrassment were the fear of jealous reactions (n ¼ 5), teasing (n ¼ 2), or arousing pity (n ¼ 1), and 7 did not wish to respond to the question
Consumption of RUF in children who recovered. The rate of RUF consumption was assessed at each follow-up visit up to week 10, which is approximately the median duration of the study among children who recovered. Median proportions of RUF consumed compared to RUF prescribed were 50% (50-95), 71% (32-87), 50% (33-89), 50% (45-89), and 50% (50-93) at week 2, 4, 6, 8, and 10, respectively. There was no difference between malnutrition groups in the rates of RUF consumption. Discussion With regard to initial acceptability, most children participating in this outpatient nutritional rehabilitation protocol reported a positive organoleptic appreciation of RUF. This result is consistent with those of a previous study that also assessed organoleptic qualities using a structured questionnaire in a similar context.12 However, such a quantitative approach might insufficiently explore the children’s actual perception as they may answer what they think is the ‘‘desirable response,’’ particularly in the context of a food aid program. In our study, extensive investigations revealed that most children needed continuous encouragement from the caregiver and that feelings of nausea, disgust, and refusals were common. Moreover, organoleptic acceptability is likely to decrease over time, due to saturation with RUF’s taste and dietary monotony. Qualitative studies conducted in HIV-infected adults on ART enrolled in similar RUF protocols (Plumpy Nut or Plumpy Sup) reported that most outpatients were pleased to receive RUF and perceived it to be beneficial as a therapeutic product for improving general health.9,22,23 In spite of this, most of them indicated that they disliked the RUF taste, which was perceived to induce nausea and vomiting, and that both the taste and consistency
made it difficult to swallow. A leading reason for nonadherence in these studies was boredom with having to consume the product so frequently over a long period of time. In low-income settings, Plumpy Nut and Plumpy Sup are thus far the most widely available and accessible products for outpatient therapy. Studies comparing the acceptability of a novel milk-free soy–maize–sorghum RUTF versus peanut-based RUTF (Plumpy Nut) in healthy children24 and not-wasted HIV/tuberculosis adults25 could not demonstrate the superiority of this alternative product. As palatability of therapeutic foods constitutes a primary criterion for the acceptability of rehabilitation protocols, further research is urgently needed to offer products with improved organoleptic qualities, particularly for adolescents and adults. Although children and caregivers were advised to eat the sachets at snack time to preserve family meals, many substituted their habitual breakfast with RUF and/or consumed a sachet at an inopportune time, just before or just after meals. Almost half of the children could not consume a full RUF dose in a single feeding. To cope with nausea and lack of appetite, children were advised to have small bites of RUF throughout the day until the amount prescribed was consumed. Feeding strategies, which consist of alternating RUF consumption with meals and dividing intake throughout the day, may not be feasible in practice, due to school and transportation constraints, especially when a child is prescribed up to 4 sachets per day. Moreover, self-stigma was reported by some children who so feared adverse reactions that they hid from others when eating RUF. In this study, where barely one-third of children were informed of their HIV status, HIV-associated stigma with RUF was not mentioned. In Southern Africa, where HIV-associated stigma with RUF is widespread, this has ultimately led to poor adherence in adult outpatients.9,10,22 In Senegal, a low HIV prevalence country, RUF seems at this time to remain associated with child health and nutrition. Sharing of RUF with people beyond the immediate family was common (and probably underreported) in our study, thus depleting the nutritional supplement available to the child. Food insecurity was highly prevalent in these HIV-affected households, and it is likely that other family members would need nutritional support. Sharing were commonly reported in individual-targeted HIV food programs where ‘‘food by prescription’’ is mostly used to focus on HIV-related wasting.9,22,26 The effectiveness and feasibility of food-based interventions for the entire family, distributed by HIV clinics to complement individual-targeted food aid, need to be evaluated. In this study, overall adherence with ‘‘by weight’’ RUF prescription was rather low and stable from the beginning, at around 50%, corresponding to an average of 2 sachets consumed daily. Similar to our results, a study in Vietnam reported that 65% and 81% of RUTF prescription was consumed by HIV-infected children <5 years and adults, respectively.12 Of note, in that study, children received 1 dose and adults received 2 doses daily during a 4-week period. The combination of previously mentioned factors, such as boredom with RUF, lack of time to consume, self-stigmatization, and sharing within family, may explain this poor adherence result. Assessment of adherence to RUF in HIV-infected adults, prescribed 2 to 4 sachets daily for 3 to 12 months, suggested that 2 sachets would be a physiologically and socially acceptable option.9,22,23 Ultimately, 25 of the 43 children enrolled in outpatient RUF-based protocols had a successful weight recovery. Median duration to recovery in MAM children was 54 days, whereas for SAM children, it was nearly twice as long, at 114 days. Daily weight gains, 1.4 and 1.3 g/kg in MAM and SAM children, respectively, were by far lower than those reported in HIV-infected SAM children <5 years under similar RUTF protocols27-31 but were found to be highly consistent with HIV-infected adults’ weight gain outcomes.8 In older children and adolescents, slow weight gain and its corollary, longer recovery duration, are likely to strongly affect acceptability of such outpatient RUF-based protocols. Effectively, improvement in weight is hardly perceptible by the patient, his/her caregiver, and the health worker and might present a risk of discouragement (such situations were observed within the study). In addition, coming to the HIV clinic Cames et al 33every 2 weeks for a long period of time involves huge effort and constraints for undernourished children, who are largely at school and have a long commute to reach the clinic. This was the reason given by 9 children, all on ART and with undetectable viral load, for discontinuing their participation in the study after 63 days on average. To partly alleviate this burden, monthly follow-up visits would be more appropriate for children in school. However, it would involve massive RUF provision (1 box ¼ 12 kg), which in turn could create difficulties with transportation because of physical weakness and stigmatization due to noticeable branded boxes, household storage, and increased sharing as reported in HIV-infected adults.9,10 This study has 2 important limitations. As a pilot intervention, it was designed for a short duration and had limited funding, so thus enrolled only a small number of patients as and when they came to their routine care. Second, the results were obtained in a research context, providing free medical support and remuneration for travel costs to attend the frequent visits. Although the findings may not be generalizable to all undernourished HIV-infected children and adolescents, they provide initial data on the overall acceptability and feasibility of outpatient RUF-based protocols among undernourished HIV-infected children 5 years. As such, they allowed for the funding and scaling up of these therapeutic protocols within an operational research study in HIV-infected children that started in early 2015 and is currently ongoing at the national level in Senegal. We reported a satisfactory rate of successful weight recovery in children on ART, despite high levels of food insecurity and a suboptimal consumption of RUF. However, this pilot study highlights several acceptability issues of major concern, as they may lead to decreasing adherence to RUF protocols in the long term and directly impact the effectiveness of outpatient therapies. All in all, nutritional rehabilitation is likely to remain a long-term challenging experience that is mainly tolerated by the young patients. Beyond the organoleptic barriers concerning RUF, a simplified intervention design as well as strengthened and continuous psychosocial support appear to be determinant and beneficial areas to invest in order to improve the social acceptability of outpatient nutritional rehabilitation. As such, our findings are a source of operational evidence on the management of nutritional rehabilitation in this vulnerable population. Authors’ Note CC and MV formulated the research question and designed the study. MV, NGH, and ABD conducted the study. CC performed the analysis and wrote the first draft of the manuscript. All authors read and commented on drafts and approved the final version of the manuscript. No donor had any part in the study design, analysis or writing of this manuscript. The ANRS 12279 MAGGSEN Cohort Study Group Study sites: (1) Centre Hospitalier National d’Enfants Albert Royer (Dakar, Senegal): Haby Sy Signate (investigator); Aminata Diack Mbaye, Helene Mbodj, Ndeye Fatou Diallo and Ndeye Rama Diagne (study clinicians and coinvestigators); Aissata Ba, Aicha Dia (study clinicians); Ndeye Ngone Have, Astou Dieye (social workers); Henriette Faye (midwife); Oumy Fall (nurse); Rosine Sonko (laboratory technician); and Natalia Laschina (psychologist). (2) Synergie pour l’Enfance/Centre Hospitalier Roi Baudouin (Guediawaye, Senegal): Ngagne Mbaye (investigator); Baly Ouattara and Maty Diouf (study clinicians and coinvestigators); Abdou Kader Niang (nurse); Alhadji Bassine Diom, Adama Ndour (social workers); and Bigue Badiane (laboratory technician). Supporting institutions: Agence Nationale de Recherches sur le SIDA et les h´epatites virales, France: Brigitte Bazin (Sponsor Representative). Study Coordination: Institut de Recherche pour le D´eveloppement (IRD), Montpellier, France: Cecile Cames (study coordinator), Philippe Msellati (investigator), Lea Pascal (study analyst), Amady Ndiaye (study technician), Suzanne Izard (data manager), Caroline DesclauxSall (doctoral fellow), Ousseynou Ndiaye and Binta Seck (biologists), and Ibrahima Diallo (study monitor, until June 2013). Acknowledgments The authors gratefully acknowledge all the participating children and their caregivers. They also acknowledge the United Nations Child’s Fund and the World Food Programme country offices in Senegal for the provision of Plumpy Nut and Plumpy Sup, respectively, within the study. They sincerely thank David Masson (Sidaction) for his critical review of the manuscript and his valuable advice in the interpretation of the results. Declaration of Conflicting Interests The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article. Funding The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: Financial support to the MAGGSEN Cohort was provided by the Agence Nationale de Recherche sur le SIDA et les h´epatites virales (ANRS) (grant number: 12279).