Background
Malnourished children usually look very thin or wasted and they have a high risk of death and illness. Treating severely malnourished children in hospitals is not always desirable or practical in rural settings, and home-based treatment may be better. Home-based treatment can be food prepared by a caregiver (such as flour porridge or energy- and nutrient-dense locally available foods), or ready-to-use therapeutic food (RUTF) provided by a clinic. RUTF is usually made according to a standard, energy-rich composition defined by the World Health Organization (WHO). Typically, the ingredients for standard RUTF include milk powder, sugar, peanut butter, vegetable oil, vitamins and minerals; but ingredients vary depending on local availability, cost and acceptability. Benefits of RUTF include a long shelf life without refrigeration and they require no preparation. This is an update of our previous review, where definite conclusions about the effects of RUTF could not be drawn from the four studies that were available at that time.
Review question
We assessed standard RUTF compared to an alternative dietary approach (e.g. flour porridge or locally available foods) and examined whether smaller amounts and different formulations of RUTF can achieve similar health outcomes in severely malnourished children aged between six months and five years. The main health outcomes that we investigated were recovery from severe malnutrition, deterioration or relapse, death and the rate of weight gain.
Included study characteristics
We searched databases for studies up to the October 2018, and found 15 studies with 7976 children. Eight studies were conducted in Malawi, four in India, and one apiece in Kenya, Zambia, and Cambodia. One small study included only children infected with HIV, another study analysed children with and without HIV separately for the main outcome (recovery), while the other studies included children who were not infected with HIV or who were untested. Overall, we judged six studies to be at high risk of bias, three studies to be at unclear risk of bias, and six studies to be at low risk of bias. (With 'risk of bias', we mean the extent to which the methods used in a study enable it to determine the truth.) All the studies lasted between 8 and 16 weeks. Only five studies followed up children after the study (for a maximum of six months), and generally reported on a limited number of outcomes.
Of our 15 included studies, six were linked to funding or donations from industry, one did not report the source of funding, and eight studies reported funding where sponsors did not include industry.
Key findings
Compared to alternative dietary approaches, standard RUTF probably improves recovery (moderate-quality evidence) and may increase the rate of weight gain slightly (low-quality evidence), but the effects on relapse and death are unknown (very low-quality evidence). With 'quality of evidence' we mean how confident we are that the particular finding represents the true effect. For example, 'very low-quality' means we are very uncertain about the finding, 'low-quality evidence' means the future research is very likely to change the finding, 'moderate-quality evidence' means that future studies may change this finding, and 'high-quality evidence' means that it is unlikely that future studies will change the finding.
Standard RUTF meeting total daily nutritional requirements may improve recovery and relapse compared to a similar RUTF given supplementary to the usual diet (low-quality evidence), but for death and the rate of weight gain, the effects are not known (very low-quality evidence).
When comparing RUTFs of different formulations, it makes little or no difference for recovery whether a standard or alternative formulation RUTF is used (high-quality evidence). For relapse, using standard RUTF decreases relapse (high-quality evidence). It probably makes little or no difference to death (moderate-quality evidence) and to the rate of weight gain (low-quality evidence) whether standard or alternative formulation RUTF is used.
Well-designed, randomised controlled trials (experimental studies where participants meeting the inclusion criteria have an equal chance of being allocated to any of the intervention or control groups) in which analyses have been performed separately for children with and without HIV, and that also measure and report on diarrhoea occurrence, are needed.
Compared to alternative dietary approaches, standard RUTF probably improves recovery and may increase rate of weight gain slightly, but the effects on relapse and mortality are unknown. Standard RUTF meeting total daily nutritional requirements may improve recovery and relapse compared to a similar RUTF given as a supplement to the usual diet, but the effects on mortality and rate of weight gain are not clear. When comparing RUTFs with different formulations, the current evidence does not favour a particular formulation, except for relapse, which is reduced with standard RUTF. Well-designed, adequately powered, pragmatic RCTs with standardised outcome measures, stratified by HIV status, and that include diarrhoea as an outcome, are needed.
Management of severe acute malnutrition (SAM) in children comprises two potential phases: stabilisation and rehabilitation. During the initial stabilisation phase, children receive treatment for dehydration, electrolyte imbalances, intercurrent infections and other complications. In the rehabilitation phase (applicable to children presenting with uncomplicated SAM or those with complicated SAM after complications have been resolved), catch-up growth is the main focus and the recommended energy and protein requirements are much higher. In-hospital rehabilitation of children with SAM is not always desirable or practical - especially in rural settings - and home-based care can offer a better solution. Ready-to-use therapeutic food (RUTF) is a widely used option for home-based rehabilitation, but the findings of our previous review were inconclusive.
To assess the effects of home-based RUTF used during the rehabilitation phase of SAM in children aged between six months and five years on recovery, relapse, mortality and rate of weight gain.
We searched the following databases in October 2018: CENTRAL, MEDLINE, Embase, six other databases and three trials registers. We ran separate searches for cost-effectiveness studies, contacted researchers and healthcare professionals in the field, and checked bibliographies of included studies and relevant reviews.
Randomised controlled trials (RCTs) and quasi-RCTs, where children aged between six months and five years with SAM were, during the rehabilitation phase, treated at home with RUTF compared to an alternative dietary approach, or with different regimens and formulations of RUTF compared to each other. We assessed recovery, deterioration or relapse and mortality as primary outcomes; and rate of weight gain, time to recovery, anthropometrical changes, cognitive development and function, adverse outcomes and acceptability as secondary outcomes.
We screened for eligible studies, extracted data and assessed risk of bias of those included, independently and in duplicate. Where data allowed, we performed a random-effects meta-analysis using Review Manager 5, and investigated substantial heterogeneity through subgroup and sensitivity analyses. For the main outcomes, we evaluated the quality of the evidence using GRADE, and presented results in a 'Summary of findings' table per comparison.
We included 15 eligible studies (n = 7976; effective sample size = 6630), four of which were cluster trials. Eight studies were conducted in Malawi, four in India, and one apiece in Kenya, Zambia, and Cambodia. Six studies received funding or donations from industry whereas eight did not, and one study did not report the funding source.
The overall risk of bias was high for six studies, unclear for three studies, and low for six studies. Among the 14 studies that contributed to meta-analyses, none (n = 5), some (n = 5) or all (n = 4) children were stabilised in hospital prior to commencement of the study. One small study included only children known to be HIV-infected, another study stratified the analysis for 'recovery' according to HIV status, while the remaining studies included HIV-uninfected or untested children. Across all studies, the intervention lasted between 8 and 16 weeks. Only five studies followed up children postintervention (maximum of six months), and generally reported on a limited number of outcomes.
We found seven studies with 2261 children comparing home-based RUTF meeting the World Health Organization (WHO) recommendations for nutritional composition (referred to in this review as standard RUTF) with an alternative dietary approach (effective sample size = 1964). RUTF probably improves recovery (risk ratio (RR) 1.33; 95% confidence interval (CI) 1.16 to 1.54; 6 studies, 1852 children; moderate-quality evidence), and may increase the rate of weight gain slightly (mean difference (MD) 1.12 g/kg/day, 95% CI 0.27 to 1.96; 4 studies, 1450 children; low-quality evidence), but we do not know the effects on relapse (RR 0.55, 95% CI 0.30 to 1.01; 4 studies, 1505 children; very low-quality evidence) and mortality (RR 1.05, 95% CI 0.51 to 2.16; 4 studies, 1505 children; very low-quality evidence).
Two quasi-randomised cluster trials compared standard, home-based RUTF meeting total daily nutritional requirements with a similar RUTF but given as a supplement to the usual diet (213 children; effective sample size = 210). Meta-analysis showed that standard RUTF meeting total daily nutritional requirements may improve recovery (RR 1.41, 95% CI 1.19 to 1.68; low-quality evidence) and reduce relapse (RR 0.11, 95% CI 0.01 to 0.85; low-quality evidence), but the effects are unknown for mortality (RR 1.36, 95% CI 0.46 to 4.04; very low-quality evidence) and rate of weight gain (MD 1.21 g/kg/day, 95% CI - 0.74 to 3.16; very low-quality evidence).
Eight studies randomised 5502 children (effective sample size = 4456) and compared standard home-based RUTF with RUTFs of alternative formulations (e.g. using locally available ingredients, containing less or no milk powder, containing specific fatty acids, or with added pre- and probiotics). For recovery, it made little or no difference whether standard or alternative formulation RUTF was used (RR 1.03, 95% CI 0.99 to 1.08; 6 studies, 4188 children; high-quality evidence). Standard RUTF decreases relapse (RR 0.84, 95% CI 0.72 to 0.98; 6 studies, 4188 children; high-quality evidence). However, it probably makes little or no difference to mortality (RR 1.00, 95% CI 0.80 to 1.24; 7 studies, 4309 children; moderate-quality evidence) and may make little or no difference to the rate of weight gain (MD 0.11 g/kg/day, 95% CI −0.32 to 0.54; 6 studies, 3807 children; low-quality evidence) whether standard or alternative formulation RUTF is used.