Table 2.

Risk of bias assessment.

Study	Sequence generation	Baseline characteristics	Blinding (intervention)	Incomplete outcome data	Selective reporting	Other bias
Amorim et al.23	Unclear	Low	Unclear	Unclear	Low	Low
Burke et al.24	Unclear	Unclear	Unclear	Unclear	Low	Low
Chijiwa et al.25	Unclear	Unclear	Unclear	Low	Low	Low
Deak et al.26	Unclear	Unclear	Unclear	Low	Low	Low
Henry et al.27	Unclear	Unclear	Unclear	Low	Low	Low
Mahmoud et al.28	Unclear	Unclear	Unclear	Unclear	Low	Low
Majidi et al.29	Unclear	Unclear	Unclear	Unclear	Low	High
McKim et al.30	Unclear	Unclear	Unclear	Low	Low	High
Molina-Hernandez et al.31	Unclear	Unclear	Unclear	Low	Low	Low
Molina-Hernandez et al.32	Unclear	Unclear	Unclear	Low	Low	Low
Nagpal et al.33	Unclear	Unclear	Unclear	Low	Low	Low
Rinwa & Kumar34	Unclear	Unclear	Unclear	Low	Low	High*
Saravi et al.35	Unclear	Unclear	Unclear	Low	Low	Low
Saravi et al.36	Unclear	Unclear	Unclear	Unclear	Low	Low
Singh et al.37	Unclear	Unclear	Unclear	Unclear	Low	Low
Singh & Goel38	Unclear	Unclear	Unclear	Low	Low	High*
Tong et al.39	Unclear	Unclear	Unclear	Low	Low	Low
Wang et al.40	Unclear	Unclear	Unclear	Unclear	Low	Low
Wong et al.41	Low	Unclear	High	Low	Low	Low
Xu et al.42	Unclear	Unclear	Unclear	Low	Low	High
Zheng et al.43	Unclear	Unclear	Unclear	Unclear	Low	Low
Zheng et al.44	Unclear	Unclear	Unclear	Low	Low	Low

Risk of bias relating to Allocation concealment, Random housing, Blinding (assessment), and Random outcome assessment was unclear for all included studies, and as such these domains have been omitted from the table. *High risk of bias only for the treatment groups receiving an intervention in addition to minocycline. For the treatment groups receiving only minocycline, risk of bias was low.