The Randomised Control Trial which was supported by Bill & Melinda Gates Foundation has been wound up and results are due to be published very soon by Innovations for Poverty Action (IPA). The intervention has been running for four years since October 2012, when the RCT started with the base line and randomisation of villages, which continued six months overtime until September 2013. Once this was completed by IPA, the actual ‘treatment’ could begin with Africa AHEAD and Ministry of Health mobilising the villages to form up Community Health Clubs. The training for trainers was done in November 2013 and 50 Classic Club were started up. This was followed by the intervention February to June 2014, during which time 50 villages were given the Classic CHC Training of 20 topics. The 50 Lite Villages were then trained and ran a shortened version of the CHC training, dubbed’ Lite’ in May and June 2014. The idea was that the two wings could be compared in order to understand if length of training and number of topics would impact on levels of hygiene behaviour change.
Seasonal mis-timing: The entire training of the community had to be conducted during the rainy season which doubtless would impact on membership numbers and attendance rates. Rusizi is one of the wettest areas of Rwanda with an annual rainfall of 1400 mm most of which pours down daily between September and May. Therefore the timing of the intervention could not have been worse.
Post training behaviour change: After this period of 20 weekly dialogue session, it was planned to reinforce key messages and encourage hygiene behaviour change through running competitions between the CHCs. This would encourage positive peer pressure to conform to general standards approved by the CHC and reinforced by public recognition throughout the district. However, according to the research protocol, after the six month training these villages had to be left at to their own devices with no further involvement of Africa AHEAD as it was felt that this would contaminate the true level of response.
Equity of treatment: The following year, 2015, was dedicated to ensuring that all 150 villages in the trial were brought up to the same level. This meant that the 50 Lite villages which had received only 2 months (instead of 6 months) and the 50 control villages (who had not even started a CHC) would all undergo ‘Classic’ training so they have all been given the same amount of information in participatory activities of 20 different topics. The project was due to be completed by November 2015. However, as all the activities had not been completed, the project was extended to the end of June 2016.
The initial findings of the RCT by IPA were shared with partners in May 2016. Results were disappointing compared to similar project in Zimbabwe where hygiene behaviour change was more evident. However, when the data was looked at in more detail and the variable of number of sessions attended was introduced, it appears that the properly mobilised CHCs had indeed responded as predicted and there were some changes in sanitation upgrading. These results will be published by IPA.
Attendance of CHC sessions affects behaviour change: Our own project records based on information collated from membership cards and registers show that the bulk of the Classic CHC had been shortchanged in the training due to the poor timing of the intervention and that this had impacted on the level of response. Unfortunately due to the late start, the training period was cut to 5 instead of 6 months, and therefore in most cases the training was not properly done. In fact only 6% of the Classic CHCs had met 20 or more times as expected. Although all the CHCs had completed the 20 topics, these had been concertina-ed into fewer meetings. 45% of the CHCs had met 17-19 times, and 39% had met between 9-16 times. Therefore this intervention was not a good example of a ‘Classic’ level of training as it should have been done.
Assumptions of the CHC approach: The model rests on the assumption that at least 20 sessions (ie physical meetings) are needed to provide enough reinforcement to induce people to chance. In addition, for these sessions to be effective they should be ‘participatory’ in method, so as to enable people to arrive at their own decision through analysis rather than blindly follow top-down directives. It stands to reason that if more than one topic is done in a 2 hour session, the chances are it will be rushed and members will not be able to discuss adequately. Too much information in one day will result in people forgetting a lot of it.
Way Forward: To get an accurate analysis of the effectiveness of the CHC Model more should to be done to complete the training and ensure better levels of response. All the CHCs from the three wings should join in the competitions at District and a final round of data needs to be collected for the endline.
Lessons Learnt: It was clear that the amount of time calculated both for the research and for the implementation was badly underestimated and led to mis-timing of the whole intervention. As a result of slavish adherence to the research protocol at the expense of the community, this was a less than perfect example of the true CHC Model, which is a flexible and people-led model. The RCT therefore may not deliver up a true representation of the power of this model to mobilise community and achieve hygiene behaviour change. It is obvious to most laymen that diarrhoea cannot be prevented after a few discussions and that to transform hygiene standards in the home may take a year or more for most households to achieve. The bar was set too high to succeed. This was due to applying standards achieved in Zimbabwe to Rwanda without adequate experience of the country. Whereas, the government appeared to be enabling through policy, in practical terms there were human factors which resulted in mismanagement which undermined the project in the field.
Additional Support: In recognition of the above factors the Gates Foundation has approved another six month budget of US$200,000 to complete all activities and allow the winding down of the project prior in order to meet all commitments to the community in the completion of the RCT. It will also allow the new monitoring website to be completed to ensure an effective tool for all NGOs and government to monitor CHCs in Rwanda. Future projects using CHC will benefit from this powerful tool which enables community self monitoring.