Use of Nonresponse Adjustment Factors for the Social Determinants and Health Equity Module in the Behavioral Risk...

The 2022 Behavioral Risk Factor Surveillance System introduced an optional module called social determinants and health equity to address health-related social needs, thereby supporting studies of social determinants of health. There were nonignorable nonresponses to the social determinants and health equity module.

Nonresponse referred to an interviewee who completed the Behavioral Risk Factor Surveillance System core survey but chose not to answer any of the social determinants and health equity module questions. The study included 2022 Behavioral Risk Factor Surveillance System participants from 39 states; Washington, DC; and 2 territories. The sequential multiple imputation approach was used to impute the household income. Multivariable logistic regression was used to estimate the propensity of not answering any questions in social determinants and health equity. A nonresponse adjustment factor was developed for each state using a propensity score estimated from the logistic regression model.

Florida, California, and New Jersey had the highest nonresponse rates at 29.3%, 26.4%, and 25.6%, respectively. After excluding the outlier Puerto Rico, the median value of nonresponse adjustment factors for states ranged from 1.09 in Idaho to 1.67 in Florida.

The nonresponse adjustment will mitigate the nonresponse bias in the analysis of social determinants and health equity data. The adjustment factor developed by the authors will be useful for analysts from various states, programs, and institutions studying social determinants and health equity.