Table 4 Quality assessment criteria for selected studies

Author, year, reference

A) Selection bias

Information bias

Confounding

Total # Yes

Overall quality: strong, 10 yes and + ; moderate, 6–9 yes; weak, 1–5 yes

1) Did 80–100% of selected individuals agree to participate?

2) Were all participants selected or recruited from the same population, subject or unit of study, or similar populations, including the same period?

3) Selection of participants into the study, or into the analysis, not based on participant characteristics observed after the start of the exposure or exposure window being studied

4) Does 80–100% of participants completed the study?

5) About the risk of bias arising from the measurement of outcomes, measurement or ascertainment of the outcome has not differed between exposure groups or levels of exposure

6) Does the measured exposure (including changes over time) well-characterize the exposure metric specified to be of interest in this study?

7) Were exposure have been assessed more than once over time of observation?

8) Assessment of the outcome has not been influenced by knowledge of participants’ exposure history?

9) Have the main potential confounders been measured?

10) Is the percentage of confounding factors controlled at least equal to 60–79%?

11) Were confounding factors that were controlled for, and for which control was necessary, measured validly and reliably by the variables available in this study?

12) Have the main potential confounders or confounders been statistically adjusted for their impact on the relationship between exposure and outcomes?

13) Did the authors use an analysis method that was appropriate to control for time-varying confounding?