Recruitment into antibody prevalence studies: a randomized trial of postcards vs. letters as invitations

Abstract

Abstract Background In a potential epidemic of an emerging infection, representative population-based serologic studies are required to determine the extent of immunity to the infectious agent, either from natural infection or vaccination. Recruitment strategies need to optimize response rates. Methods Within a seroepidemiologic study to determine the true burden of SARS-CoV2 infection in two Bay Area counties, we evaluated whether letter (L) or postcard (P) invitations with reminders were more effective at recruiting participant households. Using geographic, probability-based sampling, 9,999 representative addresses, split between Santa Clara and Solano counties, were randomized to receive an initial invitation (L or P); a randomized reminder mailing sent two weeks later to all non-respondents created four mailing type groups (L/L, L/P, P/L, P/P). Interested households provided contact information via survey to perform blood spot collection at home for testing and then receive SARS-CoV2 serology results. Comparison of demographics among respondents and non-respondents used census tract data. Results Receiving any reminder mailing increased household response rates from 4.2% to between 8–13% depending on mailing combination. Response rates from two letters were 71% higher than from two postcards (13.2% vs. 7.7%, OR = 1.83 [95% CI: 1.5–2.2]). Respondents were older, more educated and more likely white than non-respondents. Compared to Solano county, Santa Clara county had different demographics and increased household response rates (L/L: 15.7% vs 10.7%; P/P: 9.2% vs. 6.1%; p < 0.0001); the effect of mailing types, however, was the same (L/L vs. P/P: Santa Clara: OR = 1.83 [95% CI: 1.4–2.3]; Solano: OR = 1.84 [95% CI:1.4–2.5]). Conclusion Letters, as both invitations and reminders, are more effective recruitment tools than postcards and should be considered when seeking a representative population-based sample for serological testing.