Thiazolidinedione use and atrial fibrillation in diabetic patients: a meta-analysis

Table 1 Quality assessments of included studies

Study, year	Study type	Randomisation Method	Blinding	Eligibility criteria reported	Study Population representative of normal practice	Method of follow-up properly defined	Equal follow-up between groups	Was loss to follow-up reported or explained	Prospective recruitment	Consecutive recruitment
PROactive, 2005 [12]	RCT	Randomised permuted blocks	Double	Yes	Yes	Yes	Yes	Yes	Yes	Yes
Anglade, 2007 [13]	Case control	NA	NA	Yes	Yes	Yes	Yes	No loss to follow-up	No	Yes
RECORD, 2009 [14]	RCT	Randompermuted blocks	None	Yes	Yes	Yes	Yes	Yes	Yes	Yes
Gu, 2011 [15]	Cohort	NA	NA	Yes	Yes	Yes	Yes	No loss to follow-up	Yes	Yes
Chao, 2012 [16]	Case control	NA	NA	Yes	Yes	Yes	Yes	No loss to follow-up	No	Yes
Liu, 2014 [17]	RCT	Computer	Double	Yes	Yes	Yes	Yes	No loss to follow-up	Yes	Yes
Pallisgaard, 2016 [18]	Cohort	NA	NA	Yes	Yes	Yes	Yes	No loss to follow-up	Yes	Yes

ISSN: 1471-2261