Taking the nutrition conversation to the district-level in India: Data challenges are likely to constrain multisectoral action for nutrition

Cross-posted from POSHAN

India, land to millions of undernourished children, has a decentralized governance landscape that requires translation of national policy guidance to state guidelines and resources, and finally, district-level coordination and action to deliver for nutrition.  However, little is known about how to motivate and support district leadership and key departmental actors to fulfil their roles in providing leadership for, and strengthening delivery of various government programs that together have the potential to close gaps in the immediate and underlying determinants of nutrition. 

Several national and state programs abound that address causes such as food security, women’s education, poverty, water and sanitation.  Working well, and working together, these programs have tremendous potential to ensure that enabling conditions for better nutrition are created within communities and households.  The execution of these programs, however, is dependent on effective leadership at the district-level.

With civil society partners in Madhya Pradesh, Odisha and Uttar Pradesh, POSHAN  is, therefore, testing the potential for data-driven nutrition-sensitization workshops at the district-level to build a greater understanding for the level and nature of undernutrition at the district-level, the status of immediate, underlying and basic causes of undernutrition.

District nutrition profiles draw on diverse sources of data to compile a set of indicators on the state of nutrition and its cross-sectoral determinants.  The profiles are intended to be conversation-starters at the district level, and to enable discussions about why undernutrition levels are high, and what factors, at multiple levels, might need to be addressed to improve nutrition.

To date, POSHAN has developed 14 district nutrition profiles, for districts in Odisha, Uttar Pradesh, Madhya Pradesh and Jharkhand.  In developing these profiles, we were struck by several challenges in finding recent and reliable data sources for the diverse drivers of undernutrition at district level. These included:

• Diversity of sectors from which data must be sourced: The data had to reflect the different sectors that influence nutrition like food security, water and sanitation, economic status and women’s issues. This required using various datasets and identifying nutrition-relevant indicators in them. District-level nutritional status data, for instance, was available only in the DLHS-2, HUNGaMA and NIN reports; data on immediate factors like breastfeeding and immunization were available in DLHS-2 and DLHS-3 reports; data on underlying factors like water, sanitation and food security had to be drawn from the Census reports or computed using the unit-level National Sample Survey (NSS) 68th survey. Finally, Planning Commission estimates and State Economic Survey reports had to be reviewed to compile data on basic indicators like poverty and district domestic product.
• Temporal issues:  Most of the data are from different reports and often, this meant that the years during which the data was collected varied. The temporal diversity in the data made it difficult to compare nutrition data at the district and state level or even different types of indicators for each district.
• Indicator definitions:  While all indicators were defined as they appear in global guidelines, delving into various sources for district-level data and comparing them with each other meant that some of these definitions had to be altered slightly to the data available. Vitamin A supplementation, for instance, used data for 9-59 month children in one official report and data for 12-23 month children in another .
• Sampling differences: Some of the data sources provided only rural data and used smaller samples. This made it difficult to compare indicators using these data sources with data available from national level surveys.
• Data skills:  Some data, e.g., on food security and diet diversity, require the use of unit-level data from large, complex data sources such as National Sample Survey Organization (NSSO) data.  Others are less challenging e.g., WASH indicators and access to services, which can be almost directly obtained from the census reports.

There were some pleasant surprises too. In Madhya Pradesh, a nutrition survey in 2010, by the National Institute of Nutrition, provided nutrition and health data for all districts. The latest census data and recent NSS 68th round provided reliable and reasonably up-to-date WASH and food security indicators that helped understand the underlying causes of undernutrition fairly well.  Many data sets and reports are available free, either on the internet or by requesting the data.

Notwithstanding the challenges noted above, we developed a set of district nutrition profiles in early 2014 (see examples below) that are now being used by POSHAN partners in the following ways:

1)     In Uttar Pradesh, Vatsalya, brought together district level officials in a meeting convened at the level of the District Development Officer, to discuss the state of nutrition in Lucknow district and identify areas for immediate action.

2)     In Odisha, the Public Health Resource Network convened a state-level meeting to present the first set of district profiles and launch a process for district-level meetings.

3)     In Madhya Pradesh, Vikas Samvad, is in the process of conducting district workshops in Shivpuri, Khandwa and Balaghat districts and using the district profiles to discuss the state of nutrition.

4)     Using profiles from Kandhamal, Dumki and Paschimi Singhbhum districts in the states of Orissa and Jharkhand, PHRN, together with PRADAN, used the district nutrition profiles in combination with a 2-3 day qualitative field assessment to identify critical areas for PRADAN to begin engaging in supporting actions for nutrition.

POSHAN believes in the power of data, and the power of data-driven conversations to mobilize for nutrition and we are excited to be supporting such a diverse set of knowledge mobilization partners in convening nutrition conversations using these district nutrition profiles.   There is more to come on this topic –more work on harmonizing indicators in the district nutrition profiles, building learning around linking the data profiles with meaningful conversations, and honing in on key district-specific actions based on nutrition-focused conversations.

Stay tuned!