. 2022 Sep 8;17(9):e0273733. doi: 10.1371/journal.pone.0273733

Table 1. Characteristics of included studies.

M0, induction BCG therapy only; M1, 1-year BCG maintenance therapy; M2, BCG maintenance therapy for more than 1 year.

Category	Study	Methods	Study design	BCG strain	No. of Patients	Mean/median follow-up period	Recurrence	Quality assessment (SIGN)
M0 vs. M1	Mohamed et al. 2020 [13]	M0	Prospective	Pasteur	27	40 months	15	2+
	Mohamed et al. 2020 [13]	M1	Prospective	Pasteur	26	35 months	5	2+
	Yoo et al. 2012 [14]	M0	Retrospective	OncoTICE	34	16.5 months	16^a	1+
	Yoo et al. 2012 [14]	M1	Retrospective	OncoTICE	92	43 months	21^a	1+
	Okamura et al. 2011 [15]	M0	Retrospective	Tokyo	27	66 months	13^b	1+
	Okamura et al. 2011 [15]	M1	Retrospective	Tokyo	48	102 months	8^b	1+
	Koga et al. 2010 [16]	M0	Randomized controlled	Tokyo	27	28.7 months	7^a	2+
	Koga et al. 2010 [16]	M1	Randomized controlled	Tokyo	24	26.5 months	1^a	2+
	Akaza et al. 1995 [17]	M0	Randomized controlled	Tokyo	55	42 months	20	1+
	Akaza et al. 1995 [17]	M1	Randomized controlled	Tokyo	52	48 months	22	1+
M0 vs. M2	Miyake et al. 2021 [18]	M0	Retrospective	Tokyo or Connaught	874	48 months	175^a	2+
	Miyake et al. 2021 [18]	M2	Retrospective	Tokyo or Connaught	405	48 months	41^a	2+
	Koguchi et al. 2020 [19]	M0	Retrospective	Tokyo	40	36.2 months	14	2+
	Koguchi et al. 2020 [19]	M2	Retrospective	Tokyo	38	36.2 months	5	2+
	Joshua et al. 2019 [20]	M0	Retrospective	Not addressed	40	Not addressed	8^a	1+
	Joshua et al. 2019 [20]	M2	Retrospective	Not addressed	61	Not addressed	7^a	1+
	Yuk et al. 2018 [21]	M0	Retrospective	Not addressed	29	63 months	14	2+
	Yuk et al. 2018 [21]	M2	Retrospective	Not addressed	26	63 months	5	2+
	Nakai et al. 2016 [22]	M0	Randomized controlled	Connaught	42	51 months	9^b	2+
	Nakai et al. 2016 [22]	M2	Randomized controlled	Connaught	46	51 months	9^b	2+
	Martínez-Piñeiro et al. 2015 [23]	M0	Randomized controlled	Connaught	195	103 months	80^b	2+
	Martínez-Piñeiro et al. 2015 [23]	M2	Randomized controlled	Connaught	202	102 months	68^b	2+
	Muto et al. 2013 [24]	M0	Retrospective	Connaught	64	42.3±33.1 months	23^b	1+
	Muto et al. 2013 [24]	M2	Retrospective	Connaught	40	51.1±34.6 months	6^b	1+
	Hinotsu et al. 2011 [25]	M0	Randomized controlled	Connaught	42	Not addressed	14^a	1+
	Hinotsu et al. 2011 [25]	M2	Randomized controlled	Connaught	41	Not addressed	5^a	1+
	Palou et al. 2001 [26]	M0	Randomized controlled	Connaught	61	77.8 months	16	2+
	Palou et al. 2001 [26]	M2	Randomized controlled	Connaught	65	77.8 months	10	2+
	Lamm et al. 2000 [27]	M0	Randomized controlled	Connaught	192	Not addressed	113^b	2+
	Lamm et al. 2000 [27]	M2	Randomized controlled	Connaught	192	Not addressed	77^b	2+
	Badalament et al. 1987 [28]	M0	Randomized controlled	Pasteur	46	22 months	3	1+
	Badalament et al. 1987 [28]	M2	Randomized controlled	Pasteur	47	22 months	6	1+
	Hudson et al. 1987 [29]	M0	Randomized controlled	Pasteur	21	17.2 months	6	1+
	Hudson et al. 1987 [29]	M2	Randomized controlled	Pasteur	21	17.2 months	5	1+
M1 vs. M2	Gupta et al. 2020 [30]	M1	Randomized controlled	Moscow	38	Not addressed	5^b	2+
	Gupta et al. 2020 [30]	M2	Randomized controlled	Moscow	40	Not addressed	6^b	2+
	Oddens et al. 2013 [31]	M1	Randomized controlled	OncoTICE	339	7.1 years	145	2+
	Oddens et al. 2013 [31]	M2	Randomized controlled	OncoTICE	338	7.1 years	131	2+

^aDuring the 2-year follow-up;

^bduring the 5-year follow-up.

The quality assessment was indicated by Scottish Intercollegiate Guidelines Network (SIGN) checklist. 1+ means well-conducted RCT with a low risk of bias. 1- means RCT with a high risk of bias. 2+ means well-conducted cohort studies with a low risk of bias. 2- means cohort studies with a high risk of bias.