Table 4 Design features used for assessment of risk of bias
Domaina | Low risk of bias | High risk of bias |
|---|---|---|
Sampling | Consecutive patient recruitment or administrative database with random sampling | Non-consecutive patient recruitment or administrative database with non-random sampling |
Thromboprophylaxis documentation | Reporting of patients’ thromboprophylaxis | No reporting of patients’ thromboprophylaxis |
Source of information | Prospective data collection by study investigators Retrospective duplicate chart reviews with good documentation of agreement between reviewers | Retrospective duplicate chart reviews without documentation of agreement between reviewers Administrative database information |
Recruitment years | Studies with the majority of patient recruitment years 2010 or after | Studies with the majority of patient recruitment years 2009 or before |
Specification of length of follow-up | Studies that clearly define the time period of follow-up (up to 3 months) | Studies that do not clearly define the time period of follow-up |
Study type | International multicenter; Multicenter in one country; Single center, not single surgeon | Single surgeon series |
Overall risk of biasa | No high risk of bias domains: Very low risk of bias One high risk of bias domain: Low risk of bias Two high risk of bias domains: Moderate risk of bias Three or more high risk of bias domains: High risk of bias | |