Predicting panel attrition in longitudinal HRQoL surveys during the COVID-19 pandemic in the US.

Attrition in this longitudinal panel survey was not random. Besides commonly identified demographic factors that contribute to panel attrition, COVID-19 presented novel opportunities to address sample biases by correlating attrition with additional behavioral and HRQoL factors in a constantly evolving environment. While age, ethnicity, and survey difficulty consistently predicted attrition, other factors, such as COVID-19 experience, changes of employment, productivity, physical health, mental health, and financial situation impacted panel attrition during the pandemic at various degrees.