A feasibility and pilot study of the effects of Rojiroti microfinance on the health and nutrition of children under five in Bihar, northern India

Every year, 2.2 million children die of malnutrition. Many more have their life chances restricted because of the effects of malnutrition on their health and intellect. Different programmes have been tried to combat malnutrition in children. These include distributing food directly; giving children micronutrients (eg vitamins and zinc); employing community workers to give nutritional advice; and health interventions which protect children against the consequences of malnutrition (eg immunisation against measles). Research by bodies such as the World Health Organisation (WHO) has shown that these strategies can lead to better nourished children. However, when these approaches are used in a large scale (eg a whole country) they sometimes fail to reach those most in need (those living in extreme poverty). To help the greatest number of children, we must choose approaches which are cost effective and which can be sustained in the long term.
We are a group of community workers, scientists, economists and doctors. Our partners are the Centre for Promoting Sustainable Livelihood (CPSL) - which is a non-governmental organisation in India: a UK charity (Rojiroti UK); Patna Medical College India; and two UK academic partners (the University of Nottingham and the Liverpool School of Tropical Medicine). Our team have experience of helping very poor people in rural areas of India to fight poverty. Our community workers encourage people in poor hamlets or "tolas" in northern India to form self help groups (SHGs). Most SHGs are formed by women. They are asked to save a little money regularly, initially Rs2.5 (3 pence) per week. If they save regularly, their savings entitle them to a loan. These loans start small - Rs50 (50 pence). Women in the group can receive external loans (from CPSL) of Rs500 (£5) after 3 months and Rs3,000 (£30) after 6 months (if credit is good). Loans may be used for emergencies (eg medical expenses); to allow women access existing government support schemes (eg money to travel to subsidised food shops); or to invest in livestock or agricultural equipment. Using these loans helps women avoid local money lenders (who charge much higher interest) and avoid emergency sales of their property (at knock down prices). We call this programme Rojiroti (which translates as "Livelihood").

We want to see if we can test whether Rojiroti improves children's health, using a test called a "cluster randomised controlled trial". We don't yet have all the information we need to design such a trial so we will first do a feasibility and pilot study. We will start by checking the feasibility of things like consent, weighing and measuring children, collecting data accurately and keeping track of participants in the trial. We will run a small version of the trial (called a pilot study) so that we have some preliminary data on children's nutrition. We think a key measure of the effectiveness of Rojiroti will be the children's weight corrected for height. This is called the weight for height Z score (WHZ). However, we will also record other measures of nutrition. We need preliminary data to work out how many tolas we will need for the definitive trial to test the effect (if any) of Rojiroti on children's nutrition. In our pilot study, half the tolas will get Rojiroti immediately and half will get it after 18 months. Which group gets Rojiroti immediately will be decided at random. We will weigh and measure children at the beginning and again after 18 months. We will compare the two groups to see if Rojiroti makes a difference to children's nutrition. We do not expect to be able to show a difference in our pilot study. However, once we have done the pilot study, we will know how many tolas we would need for a full trial which would answer the question. This means we can work out how many staff we will need and what the trial will cost. We can then make a funding application which will have all the necessary details.

Every year 2.2 million children die of malnutrition and conventional strategies fail to reach those most in need. We will investigate a complementary approach - Rojiroti microfinance. This innovative scheme, for the very poor, has been delivered by the Centre for Promoting Sustainable Livelihood, in Northern India, since 2001. There are now over 50,000 members (mostly women). SHG members take out small loans to cope with emergencies, avoiding high interest lenders and forced asset sales. There is, as yet, no robust evidence that Rojiroti microfinance benefits child nutrition.

We will evaluate the feasibility of a cluster randomised trial of Rojiroti microfinance, amongst very poor women in Bihar, India. The unit of randomisation will be the "tola" (around 100 people). We will measure nutrition in children under 5 in the intervention and control tolas at baseline, 9 and 18 months. The intervention group will get immediate Rojiroti microfinance and the delayed group will wait 18 months. The weight for height Z score (WHZ) is the likely future primary outcome. However, we will collect data for other nutritional indices. We will use the measured effect size and summary statistics to calculate the sample size and power of a definitive trial. We will evaluate other outcomes, such as: <5 mortality, immunisation and indebtedness. We will collect preliminary qualitative data to investigate the mechanism of any effect.

We will enroll 60 tolas in 2 phases. Phase 1 (20 tolas) has begun, funded by the DfID GPAF. Consent and baseline measurements have been completed. Tolas are grouped in pairs of similar size and are >15km apart (to avoid "viral" spread of SHGs). Consent is requested at SHG level and the discussion is video recorded. Randomisation occurs after consent. Nutrition is assessed in all under fives in the tola, whether or not their mothers are in a SHG. The full feasibility and pilot study will take 2 years. We will then apply for funding for a definitive trial.

Who will benefit from this research?
A. The clinical trials community
B. Women living in extreme poverty and their families
C. Users of the raw data: Bihar State Government; Epidemiological Researchers; Non Governmental Organisations working to alleviate poverty and malnutrition in the region
D. Rojiroti UK and the Centre for Promoting Sustainable Livelihood (CPSL)
E. Our group of collaborators. The MRC and partners who may fund future rounds of the Joint Global Health Trials scheme.

How will they benefit?
A. As described under "Academic Beneficiaries", the clinical trials community have already benefitted from the novel approach to consent, described in our published protocol (Trials 2014; 15(1): 298). It is challenging to obtain valid consent from women living in extreme poverty, many of whom will be illiterate. We hope that this approach will be useful to field investigators, evaluating complex interventions in poor communities.
B. We believe the model we have proposed emphasises partnership with trial participants. Our pilot data from 550 women in 20 tolas show that 36% of women reported forced asset sales in the previous 18 months and just 2% are allowed to travel outside of the tola without their husband's permission. Observational data show that participation in Rojiroti microfinance means women are less likely to be forced to sell assets and a greater proportion are able to travel without their husband's permission. So, irrespective of the outcome for child nutrition, participation in Rojiroti, through the proposed study, will result in female empowerment and in the reduction of poverty.
C. There are no current data on nutritional indices such as weight for height Z score (WHZ), stratified by age - for children in Bihar. We will collect these data and make the raw dataset available through the Dryad web repository. http://datadryad.org We have previously done this in other research published by our group (BMJ 2012; 344: d7373).
D. Both the nutritional and qualitative data will be used by Rojiroti UK and the (CPSL) to understand better the communities where Rojiroti microfinance is offered.
E. These data will allow us to design a future definitive trial which will have a robust sample size and adequate statistical power. The MRC and partners will benefit from a future high quality, well designed, grant application to the Joint Global Health Trials scheme or its successor. The definitive trial will be generalisable to similar populations in the region. Rojiroti microfinance is already operating at scale. Currently there are over 50,000 Rojiroti members (almost all women) in 4,100 self help groups in Northern India. If effective in improving child nutrition, the approach can be scaled up further to reach beneficiaries in poor communities throughout the region.