. 2020 Aug 18;2020(8):CD013699. doi: 10.1002/14651858.CD013699

1. Summary of ROBINS‐I signalling questions for included cohort studies.

Author	Confounding	Selection of participants	Classification of interventions	Deviations from intended interventions	Missing data	Measurement of outcomes	Selection of reported result	Overall risk of bias
Danquah 2019	No	No	No	No	Probably Yes^a	Probably Yes^b	No	Moderate
Helmich 2017	No	Yes^c	No	No	Probably Yes^a	No	No	Moderate
Ha 2016	No	No	No	No	Probably Yes^a	Probably Yes^b	No	Moderate
Leecaster 2016	No	Yes^d	No	Yes^e	Probably Yes^a	Probably Yes^b	No	Serious
Mastrandrea 2015	No	Yes^d	No	Yes^e	Probably Yes^a	Probably Yes^b	No	Serious
Smieszek 2014	No	Yes^d	No	Yes^e	Probably Yes^a	Probably Yes^b	No	Serious

^aMissing data risk will always be a serious threat in contact tracing as the true number of close contacts is rarely known. ^bParticipants knew they would have to report the number of close contacts throughout the study. As a result, they could possibly report differently than they would have had they not known. ^cStudy of successive patients over only two months, very few cases. ^dAs students were volunteers, their desire to participate may have biased their results. ^eNot all participants reported close contacts on a daily basis, and not all of those wearing devices self‐reported contacts.