Li H, Kalra M, Zhu L, Ackermann DM, Taba M, Bonner C, Bell KJ. Communicating the Imperfect Diagnostic Accuracy of COVID-19 Rapid Antigen Self-Tests: An Online Randomized Experiment.
Med Decis Making 2024;
44:437-450. [PMID:
38651834 PMCID:
PMC11102651 DOI:
10.1177/0272989x241242131]
[Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [Abstract] [Key Words] [MESH Headings] [Grants] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Received: 07/29/2023] [Accepted: 02/26/2024] [Indexed: 04/25/2024]
Abstract
OBJECTIVE
To investigate the potential impacts of optimizing coronavirus disease 2019 (COVID-19) rapid antigen test (RAT) self-testing diagnostic accuracy information.
DESIGN
Online randomized experiment using hypothetical scenarios: in scenarios 1 to 3 (RAT result positive), the posttest probability was considered to be very high (likely true positives), and in scenarios 4 and 5 (RAT result negative), the posttest probability was considered to be moderately high (likely false negatives).
SETTING
December 12 to 22, 2022, during the mixed-variant Omicron wave in Australia.
PARTICIPANTS
Australian adults. Intervention: diagnostic accuracy of a COVID-19 self-RAT presented in a health literacy-sensitive way; usual care: diagnostic accuracy information provided by the manufacturer; control: no diagnostic accuracy information.
MAIN OUTCOME MEASURE
Intention to self-isolate.
RESULTS
A total of 226 participants were randomized (control n = 75, usual care n = 76, intervention n = 75). More participants in the intervention group correctly interpreted the meaning of the diagnostic accuracy information (P = 0.08 for understanding sensitivity, P < 0.001 for understanding specificity). The proportion who would self-isolate was similar across scenarios 1 to 3 (likely true positives). The proportion was higher in the intervention group than in the control for scenarios 4 and 5 (likely false negatives). These differences were not statistically significant. The largest potential effect was seen in scenario 5 (dinner party with confirmed cases, the person has symptoms, negative self-RAT result), with 63% of the intervention group and 49% of the control group indicating they would self-isolate (absolute difference 13.3%, 95% confidence interval: -2% to 30%, P = 0.10).
CONCLUSION
Health literacy sensitive formatting supported participant understanding and recall of diagnostic accuracy information. This may increase community intentions to self-isolate when there is a likely false-negative self-RAT result. Trial registration: Australia New Zealand Clinical Trial Registry (ACTRN12622001517763).
HIGHLIGHTS
Community-based diagnostic accuracy studies of COVID-19 self-RATs indicate substantially lower sensitivity (and higher risk of false-negative results) than the manufacturer-supplied information on most government public Web sites.This online randomized study found that a health literacy-sensitive presentation of the imperfect diagnostic accuracy COVID-19 self-RATs supported participant understanding and recall of diagnostic accuracy information.Health literacy-sensitive presentation may increase community intentions to self-isolate after a negative test result where the posttest probability is still moderately high (i.e., likely false-negative result).To prevent the onward spread of infection, efforts to improve communication about the high risk of false-negative results from COVID-19 self-RATs are urgently needed.
Collapse