Table 4.

Domains of risk of bias in RoB-SPEO.

	Domain	Description of bias	Existing tools with such a domain
1	Bias in selection of participants into the study	Bias in selection of participants into the study (commonly called selection bias) is the bias due to systematic differences between the characteristics of the study sample (defined as the sample of individuals participating in the study) and those of the target population (defined as the population for which the authors of the study sought to assess exposure) (Porta, 2014).	ROBINS-I, Risk of bias (RoB) instrument for non-randomized studies of exposures, Navigation Guide RoB tool, OHAT RoB tool
2	Bias due to a lack of blinding of study personnel	Bias due to a lack of blinding of study personnel (commonly called performance bias) is the bias that arises due to a lack of blinding of exposure assessors and other study personnel to relevant participant characteristics (e.g. disease status) that leads to exposure assessment that differs depending on participant characteristics.	Navigation Guide RoB tool, OHAT RoB tool
3	Bias due to exposure misclassification	Bias due to exposure misclassification is “erroneous [and systematic] classification of an individual, a value, or an attribute into a [exposure] category other than that to which it should be assigned”, leading to under- or over estimation of prevalence of exposure status (or level) (Porta, 2014).	ROBINS-I, RoB instrument for non-randomized studies of exposures, Navigation Guide RoB tool, OHAT RoB tool
4	Bias due to incomplete exposure data	Bias due to incomplete exposure data is the biases that arises from exposure data missing in a way that the exposure assessment is differential by exposure status (or level) in the target population (i.e. not random).	None, but existing tools do have domains on bias due to missing data in general or missing outcome data
5	Bias due to selective reporting of exposures	Bias due to selective exposure reporting is the systematic difference arising from selective reporting of exposures or exposure categories.	None, but existing tools do have domains on bias due to selective reporting in general or selective reporting of outcomes
6	Bias due to conflict of interest	Bias due to conflicts of interest is the bias introduced if financial and other interests influence the design, conduct, data collection, analysis and/or reporting of a study (Woodruff and Sutton, 2014).	Navigation Guide RoB tool
7	Bias due to differences in numerator and denominator	Bias due to differences in numerator and denominator is the bias that arises when there is a mismatch of definition and/or counting of persons contributing to the numerator and the denominator in the ratio used to estimate prevalence (Williams et al., 2006).	None
8	Other bias	Other bias is any other bias specific to a particular study rather than applicable to all studies.	ROBINS-I, RoB instrument for non-randomized studies of exposures, Navigation Guide RoB tool, OHAT RoB tool