Overview of Pre–COVID-19 Pandemic Sample (NHANES 2017– March 2020)

Background

NHANES 2019–2020 field operations were suspended in mid-March 2020 due to the COVID-19 pandemic. Data collection was completed in only 18 of the planned 30 primary sampling units (PSUs), and the resulting partial cycle was not nationally representative on its own. To enable valid national estimates for the pre-pandemic period, the National Center for Health Statistics (NCHS) combined the partial 2019–March 2020 data with the previously released 2017–2018 NHANES cycle.

The resulting NHANES 2017–March 2020 pre-pandemic data files represent a 3.2-year period and were constructed using additional PSU-level weighting adjustments to account for incomplete data collection and differences in sample design across cycles. These adjustments recalibrated the combined sample to align with the 2015–2018 NHANES sample design strata and national demographic benchmarks.

Comprehensive documentation describing sample design changes, weighting methodology, nonresponse bias assessments, quality evaluations, new phlebotomy weights, and analytic considerations is available in the NCHS analytic guidance and accompanying technical reports here: NHANES Analytic Guidance and Brief Overview for the 2017-March 2020 Pre-Pandemic Data Files

Impact on the National Exposure Report

NHANES data serve as the foundation for the National Exposure Report (NER) data tables. The construction of the 2017–March 2020 pre-pandemic files have important implications for how NER biomonitoring data and derived concentration measurements from this period should be analyzed and interpreted.

Key considerations for the National Exposure Report include:

• NER estimates for this period must use the combined 2017–March 2020 pre-pandemic survey weights, which are calibrated to produce nationally representative estimates for the full pre-pandemic period.
• Subgroup trend comparisons with earlier NHANES cycles should be interpreted cautiously; the PSU-level weight adjustments were optimized for overall national estimates rather than specific demographic subgroups.
• Variances for many estimates are reduced due to the larger combined sample size, although increased variability in weights may lead to higher variance for some subgroups or analytes.
• The pre-pandemic files represent a longer time span than standard 2-year NHANES cycles; analysts combining these data with other cycles must appropriately rescale survey weights.
• Despite these complexities, the NHANES 2017–March 2020 pre-pandemic data provide the most reliable nationally representative measurements of chemical concentrations available for the period immediately preceding the COVID-19 pandemic. These measurements serve as a critical baseline for inferring exposures and interpreting subsequent pre-pandemic findings in the National Exposure Report.