Table 3.

Risk of bias in studies included in the systematic review on detection of HB surface antigen

Author	Patient selection	Bias	Index test	Bias	Reference standard	Bias	Flow and timing	Bias
	Was a case control design avoided? Consecutive or random sample of patients? Inappropriate exclusions?		Blinded to reference standard Could the conduct or interpretation of the index test have introduced bias?		Blinded to index? Could the reference standard have introduced bias?		There is an appropriate interval between the index test and reference standard? All patients receive the same reference standard and are included in the analysis?
Alidjinou	NR, but no case control design	UR	Not blinded, interpretation unbiased	LR	Not blinded, interpretation unbiased	LR	NR	UR
Boa-Sorte	No case control design, consecutive recruitment	LR	blinded	LR	blinded	LR	Same reference standard, all patients included in analysis	LR
Brown	Not reported	UR	Not reported	UR	Not reported	UR	Not reported	UR
Farzadegan	Only cases, no consecutive sampling	HR	Not blinded, interpretation unbiased	LR	Not blinded, interpretation unbiased	LR	NR	UR
Farghaly	Case control design	HR	Not blinded, interpretation unbiased	LR	Not blinded, interpretation unbiased	LR	NR	UR
Forbi	No case control design	LR	Not blinded, interpretation unbiased	UR	Not blinded, interpretation unbiased	UR	Sampling not reported, same reference standard	UR
Gruner	NR	UR	Not blinded, NR	UR	NR	UR	NR	UR
Halfon	NR, probably case control design	UR	Not blinded, NR	UR	Not blinded, NR	UR	NR	UR
Kania	Consecutive recruitment	LR	Not blinded, interpretation unbiased	UR	Not blinded, interpretation unbiased	UR	Sampling reported	LR
Khan	No case control design	LR	Not reported, unclear whether blinded	UR	Not reported, unclear whether blinded	UR	Sampling not reported	UR
Lee	Consecutive recruitment	LR	Not blinded, interpretation unbiased	LR	Not blinded, interpretation unbiased	LR	Sampling reported, same reference standard	LR
Lukacs	Case control design	HR	Not blinded, interpretation unbiased	LR	Not blinded, interpretation unbiased	LR	Sampling not reported, same reference standard	UR
Mayer	NR, probably case control	UR	Not reported, interpretation unbiased	LR	Not reported, interpretation unbiased	LR	Sampling not reported	UR
Mendy	Case control design	HR	Not blinded, interpretation unbiased	LR	Not blinded, interpretation unbiased	LR	Sampling reported	LR
Mohamed	Case control design	HR	Not blinded, interpretation unbiased	LR	Not blinded, interpretation unbiased	LR	NR	UR
Mossner	Sampling from high-risk and low risk groups	HR	Not blinded, interpretation unbiased	UR	Not blinded, interpretation unbiased	UR	Sampling reported, same reference standard, all patients included in analysis	LR
Nielsen	Only cases	HR	Not reported	UR	Not reported	UR	Sampling not reported	UR
Parkinson	Only cases	HR	Not blinded, interpretation unbiased	LR	Not blinded, interpretation unbiased	LR	Sampling not reported, same reference standard	UR
Ross	Sampling not reported, probable case control design	HR	Not blinded, interpretation unbiased	LR	Not blinded, interpretation unbiased	LR	Flow reported	LR
Villa	Case control design	HR	NR	UR	NR	UR	NR	UR
Villar	Case control design	HR	Bias possible, as selective samples by OD values	HR	Not blinded, interpretation unbiased	LR	Sampling reported	LR
Zhuang	No case control design	LR	Not blinded, interpretation unbiased	LR	Not blinded, interpretation unbiased	LR	Sampling not reported	UR
Zoulek	Unclear, but no case control design, probably random or successive	LR	Not blinded, interpretation unbiased	LR	Not blinded, interpretation unbiased	LR	Sampling not reported, same reference standard	LR

Abbreviations: LR: low risk, HR: high risk, UR: unknown risk, NR: not reported; shaded: low risk of bias