While some program outcomes based on similar data have been published previously [1], limitations included small sample size and possible reference of seasonality. This study seeks to circumvent these limitations by nutritional outcomes of a larger set of children across seasons, considering the same intervention over a similar period of time (4 to 6 months).

AAM is being collaboratively implemented by Public Health Resource Society (PHRS), Ekjut, Child In Need Institute, Chaupal, and IDEA, in seven blocks spread across the Indian states of Bihar, Chhattisgarh, Jharkhand and Odisha. The selection of areas was further based on vulnerability-mapping, and analysis through a formal baseline survey in the program areas confirmed high prevalence of wasting and stunting. An ethical approval for the entire program, including anthropometry, data analysis and research, was obtained from the Institutional Ethics Committees of PHRS and Ekjut. The main interventions of the program were:

In the AAM crèches, every child’s weight was recorded monthly and height once in six months besides at the time of admission. In April 2013, it was decided that for the purpose of uniformity across crèches, heights would be measured at entry point and in the months of May and November each year for all children, irrespective of their date of admission. The outcomes of 45 SAM children were reported early in the program as part of a cohort of 587 children in all categories [1]. These 45 children are also included in the current analysis.

The data were recorded using a customized software for the program and, the data analysis was carried out using Stata (version 13). Z-scores were calculated using the growth standards provided by WHO [4]. The usual WHO cut-offs for implausibility were applied.

Out of the 179 included children (100 males), 86% belonged to scheduled tribe and 5% to scheduled caste categories. Most fathers (81%) and mothers (57%) worked as casual agricultural or non-agricultural labourers. Two-thirds held Below Poverty Line /Antodyaya cards and 93% lived in kutchha houses; 70% of the mothers had no schooling at all.

The average age of the children at the time of first measurement was 14.5 months. For 129 children, the first measurement coincided with their time of admission to the crèche. For the rest, average duration of stay of the children before the time of their first available and valid weight and height measurement was 1.9 months. For 63 children, the gap between the two measurements was 4 months, for 48 children the gap was 5 months, and for 68 children the gap was 6 months. All the children were referred to nutritional rehabilitation centers through the Anganwadi worker / ASHA as per government protocols. Only 30 of them were ever admitted there.

Table I presents the nutritional status of the 179 SAM children after a period of 4 to 6 months. During this period, about three-fourths of the children improved their status from SAM to normal or moderate malnutritional. The mean Z-score for these children increased significantly (P<0.001) from -3.63 (0.51) to -2.3 (1.07). Table I also presents the change in wasting status based on the first and last anthropometric measurements of these children in the program. Only 23 (13%) children continued to remain in the category of SAM at the end of the analysis period. Of these 23 children, the duration between the two measurements for 15 children was between 4 to 6 months. 65 children graduated from the programme or were unavailable for measurement after 4-6 months. The outcomes of these children are not statistically different from the overall 179.

TABLE I	Change in Weight-for-Height Status of Children With Severe Acute Malnutrition

The improvement in Z-scores was statistically significant (P<0.001) in both seasons in which the first measurement of children was taken. The change in the Z-scores for these two groups did not show any significant difference between the two seasons (Table II).

TABLE II Weight-for-Height Status in Children with SAM 

This village-level based study documented that almost three-fourth of the children with SAM show improvement in their WHZ category over a period of 4-6 months, when managed in a comprehensive community-based program.

Most other observational studies on community-based management of SAM in India have reported the nutritional outcomes of a smaller number of children [5-7]. Primarily facility-based interventions with outreach facilities to community, such as the Darbhanga model [2] have been able to cover far larger numbers with distinctly better nutritional outcomes. However, the Darbhanga model was challenged by very high dropout rates, no intervention for prevention or early detection, and questionable sustainability. In the recently published results of a rigorous randomized control trial [8] comparing the efficacy of three different feeding regimens for SAM, the recovery rate in the most efficacious arm was also only 57%, further dwindling to 17% (29% remaining SAM) 16 weeks after the end of the treatment [8]. In our study, gains continued with time with proportion of children with SAM dropping with increasing duration of intervention.

The limitation of this study is that these results are based on program data from the crèche component of the comprehensive program; the outcomes represent the net and ‘best case’ outcomes of all three strategies of AAM and not just the crèche program. The project was located in medically underserved areas and the referrals to the government health system were not sufficient to meet the many immediate medical needs of the children in crèches. The program itself was not able to offer direct medical services or facility-based care and this was a limiting factor in nutritional outcomes and overall well-being of attending children. Every crèche could not be provided with stadiometers/infantometers. The quality of data collected by supervisors was monitored through a protocol of quality checks following rigorous capacity-building. However, as with most field and facility programs on malnutrition, we did not formally analyze inter- and intra-observer reliability [1].

This study describes the observations of the AAM program, across seasonal influences, on nutritional outcomes in a community setting, and confirms the findings of other studies. The findings of this study reinforce the need for daycare programs for children that can accommodate all the requirements of a comprehensive community-based program for malnutrition.

Acknowledgements: This study is based on the collaborative project "Community-based Action on Malnutrition among Children Under Three Years: A multi-strategy intervention" run by the following organizations: CINI, Chaupal, Ekjut, IDEA and Public Health Resource Society. We are grateful for the leadership of Dr. Prasanta Tripathy, Dr. Ganapathy Murugan, Dr. Suranjeen Prasad Pallipamula, Ms. Jayeeta Chowdhury, Ranjan K Panda, Sulakshana Nandi, Dr. Audrey Prost, Haldhar Mahto, Gangaram Paikra and Digvijay Kumar. We would also like to thank Biswanath Dasgupta for his insightful comments on the paper. We thank all collaborating organizations for agreeing to use the Project MIS data for this study.

Contributors: VP: conceptualized and led the study. All authors participated in data analysis, manuscript writing and approved the final draft.

Funding: Tata Social Welfare Trust; Competing interests: None stated.