Data collection

Beyond Borders: Insights from challenges in collecting nutrition data from United States household surveys

Low, middle, and high-income countries (LMIC) face similar challenges when collecting population-level data to inform nutrition policies & programs. Earlier this year, the American Society of Nutrition’s (ASN) Committee on Advocacy and Science Policy published a white paper highlighting the challenges facing the National Health Examination & Nutrition Survey (NHANES) in the United States.

Like the Demographic and Health Surveys (DHS) and other large-scale household surveys in LMIC, the NHANES is the primary nationally representative data source in the United States for key nutrition indicators, including anthropometry, dietary intake, and micronutrient deficiencies. 

Looking across the many challenges identified by the ASN Committee, at least three themes emerge that are relevant to LMIC.

    1. The need to adapt survey content and collection to reflect a changing society

Standardization and representativeness are the primary goals of most national surveys, including the NHANES and the DHS. A changing society presents challenges to both. Survey designers are reluctant to change core questionnaires to maintain standardization and allow for comparisons across time. Adding new content can increase survey costs and respondent fatigue. However, this reluctance has costs for public health leaders who need new information. 

The NHANES has addressed emerging topics by incorporating modules and oversampling specific populations in individual survey cycles. However, it fails to capture data on a growing elderly and immigrant population. Similarly, there are rapidly changing dietary and lifestyle patterns across LMIC. Some countries, like South Africa, have included men and modules on non-communicable diseases in their DHS; however, many countries lack data on these critical issues. Creative approaches and coordinated advocacy efforts, like the global one led by DataDENT around the DHS-8 core questionnaire and national efforts, are needed to identify and prioritize missing information.

Furthermore, representativeness is threatened when people refuse to participate in a survey. The NHANES response rates have dramatically declined over the last two decades, indicating a growing challenge. Response rates for household interview modules dropped from 82% in 1999-2000 to 61% in 2015-2016 by 2017-2020[1] it was only 51%. Response rates for the physical assessment and biomarker modules dropped even more, from 76% to 47% during the same period. While statistical methods maintain representativeness, low response rates limit their effectiveness.  

Response rates to the DHS and other surveys across LMIC are much higher than the NHANES, and these contexts could share lessons about community mobilization with the United States. However, as norms evolve in LMIC, it is naïve to assume that response rates will remain high without investment in new strategies to motivate the population. 

    1. The need to invest in measurement research and methods of innovation

The ASN Committee called for a “dedicated innovation program” for the NHANES survey that will allow for rigorous testing of new data collection devices and methods to ensure they are valid and can be effectively implemented in the survey. They also cited a need for research on how data from alternative sources can be used alongside the NHANES to meet decision-maker needs.

DataDENT supports household survey innovation in LMIC by conducting research to improve methods to collect and analyze indicators of nutrition coverage in the DHS and other household surveys. Other data collection innovation investments in LMIC include Improve, INDDEX, Global Diet Quality Project, and USAID Advancing Nutrition’s work on anemia assessment. However, similar to the United States, more investment in methods development and research around data use is needed.

    1. Political commitment, funding & multisector / cross-institution coordination

The ASN Committee called out low political commitment for the NHANES demonstrated by stagnant funding and a failure to renew the National Nutrition Monitoring and Related Research Program Act of 1990, a 10-year piece of legislation that mandated collecting and reporting key nutrition data in the United States. They also highlighted the complexities of coordinating across government sectors with different mandates, including the Center for Health Statistics and the US Department of Agriculture. COVID-19 revealed these vulnerabilities and systemic weaknesses in the United States’ public health data infrastructure. As a result, the US government has now made the modernization of data systems a public health priority.

Many LMIC face challenges with multisector coordination for nutrition data and mobilizing domestic resources for data collection. Data collection priorities often remain “donor driven” and insufficient to address the need. Our landscaping of Official Development Assistance for nutrition data and information systems showed that between 2017 and 2019 donor investment in data and information systems remained stagnant, falling below the 5% benchmark, and was not directed to public institutions. The recommendations for data commitments developed by DataDENT and partners for the 2021 Tokyo Nutrition for Growth (N4G) Summit stress the importance of domestic investment in nutrition data. 

The way forward

Encouraging leaders to invest in data value chains alongside policy and program implementation is challenging given constrained resources and competing policy priorities in high, middle and low income countries.  It is essential that data investments are informed by the needs of data users – who are the starting and ending link in the data value chain. For the NHANES in the United States, the ASN Committee has called for the National Academies of Sciences, Engineering, and Medicine (NASEM) to conduct a study to establish an actionable framework for the future of the NHANES. In Nigeria and Ethiopia DataDENT is collaborating with government partners to carry out formative research to understand the data needs and priorities of policy makers and program leadership across sectors and administrative levels and to use findings to inform data strategies and stakeholder alliances. There is also a need to share insights across contexts. The Global South has lessons to share with the US Public Health system around community mobilization to improve response rates in household surveys; while the US and other contexts have developed innovations that can be taken up in LMIC.

[1] Modified to reflect the “pre-COVID” period

Leave a Reply

%d bloggers like this: