The results of the RCT in Rusizi have now come to light and although there has been no formal presentation in Rwanda the paper was delivered in USA in October 2016 by Innovations for Poverty Action, at the Water Institute Conference in North Carolina.
In December, 2015, Africa AHEAD hosted a workshop with district stakeholders to feed back on the project data collected during the implementation period. The 12 Environmental Health Officers brain stormed in two groups to assess the impact of the CHCs. This debate provided much insight as to how the intervention had been compromised. This is vital information which should be disseminated to enable lessons to be learnt for future CHC projects not only in Rwanda but elsewhere. The issues raised also cast some doubt on the use of a Randomised Control Trial in the field where variables (such as timing, weather, and people interest and behaviour) cannot be adequately controlled as they can in a scientific laboratory .
The two groups examined the Project Monitoring Data (PMD) which used 50 observable indicators in the household Inventory. See paper. They concluded that the reason there was such a discrepancy between the PMD and findings of the RCT, was that the end line of the RCT had been conducted prematurely before the communities had finished their training properly and had time to improve their facilities with concrete action. By contrast the PMD was collected a year later than the RCT, which enabled sufficient time for behaviour change to take place. By this stage the implementors were unfettered by the protocol controls of the RCT which had forbaden access to communities following the training period, even through at this stage is is normal for communities to need reinforcement.
It was agreed although the results of the RCT had been unimpressive, the project was still in its infancy at that stage, and it therefore could not have been expected to have achieved the overambitious target of reducing diarrhoea in such a short time. Furthermore, some noted that the research questions were actually not properly aligned with the objectives of the intervention. That nutritional standard of the children could have improved in such a short time (4-6 months) was put down to the inexperience of the research team. The 24 sessions in the CBEHPP training focuses mainly on hygiene conditions in the home, with only a cursory introduction (2 theoretical sessions) on the meaning of a balanced diet. For stunting to be alleviated requires a proper livelihoods programme with a practical agricultural training in food production and storage. It was also irrelevant to report on the lack of improved water source when no water component was included in the project.
In addition, the RCT definition of ‘latrine improvements’ focused on the building structure (floor, roof, walls) rather than the ‘covering of the squat hole’ which was the major drive in CBEHPP, given that in Rwanda latrine structures were already in existence in over 90% households in most villages before the training began. Therefore, the challenge in the intervention was to persuade households to add well-fitting covers to their existing latrines, so as to stop fecal-oral transmission by flies leaving the latrine. The RCT measurement missed this important nuance completely!
It was further noted by EHOs that the the start-up of the CHCs had been delayed by six months due to the base line survey conducted by the researchers which was only completed in September rather than June. This delay pushed the training of CHCs into February and as a result the 24 session intervention had been rushed in order to be completed on schedule by end of June. There was no flexibility from the research team to adjust to seasonal constraints. Project records showed that only 2 CHCs out of 50 CHCs had actually met the required 24 times, due to this curtailment of the intervention. Thus, the ‘treatment’ that had been measured was bound to deliver a false reading as it could not be considered a ‘Classic’ training. Furthermore, it had been conducted in the heavy rains which had negatively affected response rate.
Finally, both groups mentioned how the EHOs had been hampered by lack of transport to visit the field and little follow up had been conducted as is usual in a CHC programme to keep the momentum up after the training and ensure information is translated into better hygiene facilities and practices. The promise of competitions had never been fulfilled as time had run out for the intervention group and the final year was spent in bringing the control groups up to the same level of training as the ‘Classic’. The Gates Foundation agreed to additional funding to complete all promised activities and funds were finally received in October 2016. The competitions are being held in the next month March 2017, after which the programme will close down.
It was also acknowledged by local leadership that with the shake up in the Ministry of Health due to the removal of the Head of Child Health and the Minister of Health at National Level during the very time of the intervention, as well as the removal of the Mayor of Rusizi, the government programme had been temporarily derailed. The Community Based Environmental Health Promotion Programme (CBEHPP) had been negatively effected and it had been a challenge to keep the programme going at District Level in 2015/16.
However, 2017 promises to be a new start. With a new Mayor, District Head of Health in Rusizi, and a new Minister of Health there is a great hope for the future of CBEHPP in Rwanda. However, the Gates Foundation is not interested in any further support for what they consider a failed intervention, and their priorities have shifted to urban sanitation. Other donors are being invited to take up this worthwhile endeavour to scale up the CHCs throughout Rusizi as requested by the Mayor.