The main outcome, the success rate, is defined by a composite of three endpoints : alive, not acutely malnourished per the definition applied at inclusion and not having an additional episode of AM throughout the 6- month observation period. All other children are classified as 'unsuccessful'.
The secondary major outcome, the recovery rate, is defined by reaching during the 6 month follow-up a MUAC>=125 and no oedema during two consecutive weeks, a minimum RUTF treatment period of 4 weeks and good clinical condition.
Sample size:
For the main objective, the expected success rate is 68% with a statistical power of 90%. For the secondary priority objective n°1, the expected recovery rate is 82% and for the secondary priority objective n°2, the expected recovery rate is 74% with a statistical power of 80%. For all objectives, the margin of non-inferiority set is 10% with a level of significance set at 1.25% unilaterally. An inflation of 5% to account for unexploitable data was also added. The number of randomized subjects required is:
- 568 participants to meet the main objective;
- 295 participants with severe acute malnutrition to meet the secondary priority objective n°1.
- 384 participants with MUAC<115mm at admission to meet the secondary priority objective n°2.
Data collection and monitoring :
A paper form will be completed by the trial nurses during the outpatient follow-up or at home visits. Data will be recorded by data entry agents supervised by a data manager using RedCAP software.
Data monitoring will be performed every week at each site by clinical trial monitors under the responsibility of the research activities manager, according to the recommendations of Good Clinical Practices. All data entered in the database will be checked for completeness and consistency. The methods of data entry, coding, control, validation and database freezing will be described in a "data management" guide. Before the implementation of the trial, a monitoring plan will be established.
Analysis :
Before the end of the inclusion period, a statistical analysis plan will be established.
The occurrence of the primary endpoint (success rate) will be compared between the OptiMA and ComPAS strategies to the standard protocol.
The occurrence of the primary secondary endpoint (recovery rate) will be compared between the OptiMA and ComPAS strategies to the standard protocol, for children randomized in the severe acute malnourished stratum and for children admitted with a MUAC<115mm.
These comparisons will be made by Intention To Treat (ITT) (including all randomized participants), and Per-Protocol (PP) (including only those participants who received the full randomized treatment strategy).
The primary analysis (success in the overall population regardless of the level of malnutrition) and the main secondary analysis (recovery rate in the "severely malnourished" stratum) in ITT and PP are non-inferiority analyses.
The OptiMA and ComPAS strategies will be deemed non-inferior to the standard strategy if the primary and main secondary analysis statistically demonstrate non-inferiority in both ITT and PP.
The primary analyses in terms of success and recovery will be performed on available data. In case of missing data, a sensitivity analysis will be performed using the maximum bias method. Missing data can be vital status if the child is absent at the last visit, and anthropometric data (weight, MUAC, height). In the case of missing height data, the last available height can be taken into consideration given the low variability of this value from one month to the next.
The probabilities of success, recovery and relapse in each of the strategies and the quantities of RUTF consumed per child to achieve recovery will serve to construct a cost-effectiveness model.
