Efficacy of a Green Banana–Mixed Diet in the Management of Persistent Diarrhea: Protocol for an Open-Labeled, Randomized Controlled Trial

Background Diarrhea is the second-leading cause of death in children under 5 years of age. In low- and middle-income countries, 3%-20% of acute diarrheal episodes become persistent diarrhea (PD) (ie, duration ≥14 days), which results in 36%-56% of all diarrheal deaths. In Bangladesh, PD causes >25% of diarrhea-related deaths. Commensal gut microbiota dysbiosis is increasingly recognized in the pathogenesis of PD. Hospital-based management of PD requires a hospital stay, which increases the risk of infection and hospital costs. The higher cost of treatment and high case-fatality rates reiterate PD as an important public health problem. At the International Centre for Diarrhoeal Disease Research, Bangladesh (icddr,b), for the last two decades, a consensus-based guideline has been followed for PD. Observation has revealed that green banana helps in the resolution of diarrhea. However, no larger prospective study has been conducted to evaluate the efficacy of green banana in the management of PD among children older than 6 months of age. Objective Our objective is to assess the efficacy of full-strength rice suji (semolina) with and without green banana compared to three-quarter-strength rice suji in the management of PD in children aged 6-36 months at the Dhaka Hospital of the icddr,b. Methods This open-labeled, randomized controlled study aims to enroll a total of 145 children with PD who have not been improving on a diet of milk suji. Children will be randomized into three different diet-specific groups: full-strength rice suji containing green banana, full-strength rice suji alone, and three-quarter-strength rice suji. The primary outcome is the percentage of children who recovered from diarrhea by day 5. Results Recruitment and data collection began in December 2017 and were completed in November 2019. Results are expected by April 2020. Conclusions This study is expected to provide insights into the incorporation of green banana into the dietary management of PD. This would be the first study to investigate the role of microbiota and metabolomics in the pathogenesis of PD. Trial Registration ClinicalTrials.gov NCT03366740; https://clinicaltrials.gov/ct2/show/NCT03366740 International Registered Report Identifier (IRRID) DERR1-10.2196/15759


Response to comment from Prof C A Kawser on the research protocol
Sir, Thank you so much for your explicit review of the protocol. We have discussed among the investigators. Here we have added our response to individual comment after each comment italic font and also revised the protocol accordingly. RRC:  Not seen it properly in the section "icddr,b Strategic Priority/ Initiative (SP 2015-8)"  Thank you for the suggestion. We have changed the strategic priority accordingly. (page no 1)  Is it necessary to link the identity of the individual?  During data collection, there will be an identifier link that will help to retrieve any missing data until analysis. But all the data will be kept confidential and the report/manuscript/article will be published without the name or any sorts of identification of the children studied  But may request to tick 'yes' for 'internal publication' as well as 'sharing with' GOB and NGO  "Yes" is marked to the options as per your suggestion Protocol itself:  It seems there are three 'intervention' arms without a control. In the background section, protocol suggests 'milk-suji' followed by 'rice-suji' (no mention of strength) as a dietary intervention for PD but that intervention is not taken as controls.  At icddr,b,conventionally ¾ th strength rice suji is used for the management of PD in > 6 months to 60 months old children who did not improve with milk suji. And that is the control diet.
 ¾ strength rice-suji has calorie conc. of 57 Kcal/100ml which is quite low and needs justification, with reference, of its use.  We agree with your comment. However, the children usually suffer from osmotic diarrhea in this stage. Diet with sufficient caloric density is often unsuccessful. Considering these facts we are proposing to raise the calorie by using full strength diet in the trial. This is described in details on page 20, paragraph 2 in the revised protocol.
 The main objective is 'efficacy' of green banana in full-strength rice-suji compared to other two arms. But what are the 'outcome' variables that will be taken as 'efficacy'?  Outcome variable "Duration of diarrhea" will be taken as efficacy  There are three other secondary objectives. Is it necessary to have the first one, is it not well described! Regarding the third, please mention specific outcome variables to be compared. We appreciate your suggestion and these are addressed in the objective/outcome section of the revised protocol (page 16 ).
 There was a publication by G H Rabbani on the same topic, but 'background' section did not elaborate on it to justify the 'repetition' of the study. In sample size calculation the same study was referred while selecting the sample size.  The study conducted by Rabbani et al was a good study where the study enrolled only male children who were well nourished without any comorbid condition and also involved a small sample size. According to your suggestion, it was discussed in the background section.
 In background section it was claimed that existing dietary intervention causing good outcome in 95% of cases. Then what are the changes in outcome is expected!!!?? If expected then that should be depicted in sample size by reducing 'bad outcome' of 5% to lower value. In that case sample size calculation will be totally different and possibly much larger than what is contemplated.  The study described 95% success rate was actually the cumulative cure rate even after using pregestimil and was a retrospective descriptive study. We think the precise information on the percentage of the >6 months to 60 months old children improved from start to end of using 3/4 th strength rice suji is difficult to get from that study. So we have now not using that information in the revised protocol.
 At 'background' few points were termed as 'prognostic' indicator, but are they not considered as 'risk factor'?  Yes, these are risk factors. These are changed according to your suggestion  A long section was described in 'first person', but needs to be expressed in third person. The section described the existing management protocol but did not identify the lacunae or problem question/s that drives the present study.  According to your suggestion, the entire section was revised with a description of problems for which the study was planned.
 Fig-3 is not properly shown.  It was reproduced from an article, as it was, it lacks some information. We have planned not to use it and deleted in the revised protocol.
 Why a second exclusion criterion is kept? Other co-infections are to be attended. What is the chance of positivity for cholera in such cases?  Around 10% of patients with persistent diarrhea had growth of organisms in their stool and required antibiotics. If in one arm, more patients have growth than other groups, then it may produce problems during analysis.
 Why -5 SD is taken an exclusion criteria, not -3 or -4? Needs justification.  We have plans to include children with PD and SAM. However, if a child has extreme malnutrition (Z score -5 or less), s/he needs closer monitoring, feeding through naso-gastric tube, so excluded from the study.
 Please justify the volume of feed to be provided in the study. If a child is not breastfed, all the children will get <70 Kcal/kg/d. For partially breastfed, this amount will be much lower. It needs careful reconsideration/justification.