James Marcus, Tiago Maia, Paul Williams, Renata Yokota, Sofie Arnetorp, Timothy A. Herring, Sudhir Venkatesan, Philip A. Powell, Johan L. Severens, James W. Varni, Sylvia Taylor, Marieke Krol, John E. Ware
Abstract Background Many immunocompromised individuals continue to practice physical distancing behaviors to avoid severe COVID-19 outcomes. To measure the intensity of these physical distancing behaviors, we aimed to develop and validate a standardized metric across age groups: the Physical Distancing Scale for COVID-19 Avoidance (PDS-C19©). Methodology PDS-C19 development involved content scoping, initial draft PDS-C19 design phase, and exploratory testing (via asynchronous online forums). Psychometric properties (structural validity, reliability, and construct validity) of the PDS-C19 were evaluated using a random sample of immunocompromised and non-immunocompromised participants ( n = 1059) of the EAGLE Study, including adults and children aged 0.5–17 years from the US and UK. Adult caregivers of children aged ≤ 12 years completed a proxy version of the PDS-C19 on their behalf, in addition to providing self-reported responses. Results Findings from the two asynchronous forums ( n = 23 and n = 22) indicated that the items identified in the scoping phase were relevant and well-understood. As part of the structural validity analysis, item-response distributions, inter-item correlations, and exploratory factor analyses supported a nine-item unidimensional intensity scale. Confirmatory factor analyses reduced the PDS-C19 to seven items (two items were dropped due to high residual correlations or differential item functioning by age). Internal consistency was high (ω = 0.97); convergent validity was good, with high correlations ( r max =0.74) with clinical outcome assessments of the ability to participate in activities and with worry. The PDS-C19 had good known-groups validity, with scores distinguishable by types of activities, level of worry, and number of close contacts. Conclusion The PDS-C19 had acceptable structural and construct validity and high internal consistency, demonstrating its value as a measure of physical distancing to avoid COVID-19 in immunocompromised and non-immunocompromised individuals.