Search

This paper summarizes the Risk of Bias of Individual Studies in Systematic Reviews of Health Care Interventions revised by the Agency for Healthcare Research and Quality (AHRQ) and introduces how to use Revman software make risk of bias graph or risk of bias summary. AHRQ tool can be used to evaluate following study designs: RCTs, cohort study, case-control study (including nested case-control), case series study and cross-sectional study. The tool evaluates the risk of bias of individual studies from selection bias, performance bias, attrition bias, detection bias and reporting bias. Each of the bias domains contains different items, and each item is available for the assessment of one or more study designs. It is worth noting that the appropriate items should be selected for evaluation different study designs instead of using all items to directly assess the risk of bias. AHRQ tool can be used to evaluate risk of bias individual studies when systematic reviews of health care interventions is including different study designs. Moreover, the tool items are relatively easy to understand and the assessment process is not complicated. AHRQ recommends the use of high, medium and low risk classification methods to assess the overall risk of bias of individual studies. However, AHRQ gives no recommendations on how to determine the overall bias grade. It is expected that future research will give corresponding recommendations.

In this last paper of the series about risk of bias assessment, we introduce the application of risk of bias assessment results. Risk of bias assessment is one of the key steps in the assessment of quality of evidence. The risk of bias assessment results could be the "diagnosis" of individual studies, which helps decision making related to the inclusion and exclusion of individual studies, as well as the data analysis in the systematic review process. This paper focuses on how to incorporate risk of bias assessment results in the GRADE assessment for quality of evidence, including the principles and the tips for the application.

This paper introduces the tools related to Quality In Prognosis Studies (QUIPS) to assess the risk of bias in studies of prognostic factors and the relevant points of assessment and to illustrate the application of QUIPS in published prognostic research. The QUIPS tool identified 6 important areas to consider when evaluating validity and bias in studies of prognostic factors including participation, attrition, measurement on prognostic factors, outcomes, confounding factors, statistical analysis and reporting. It also provided a new method for evaluation on bias in the areas of prognostic research.

This paper introduced the Revised Tool for the Quality Assessment of Diagnostic Accuracy Studies (QUADAS-2), including the development and comparison with the original QUADAS, and illustrated the application of QUADAS-2 in a published paper related to the study on diagnostic accuracy which was included in systematic review and Meta-analysis. QUADAS-2 presented considerable improvement over the original tool. Confused items that included in QUADAS had disappeared and the quality assessment of the original study replaced by the rating of risk on bias and applicability. This was implemented through the description on the four main domains with minimal overlapping and answering the signal questions in each domain. The risk of bias and applicability with 'high','low' or 'unclear' was in line with the risk of bias assessment of intervention studies in Cochrane, so to replace the total score of quality assessment in QUADAS. Meanwhile, QUADAS-2 was also applicable to assess the diagnostic accuracy studies in which follow-up without prognosis was involved in golden standard. It was useful to assess the overall methodological quality of the study despite more time consuming than the original QUADAS. However, QUADAS-2 needs to be modified to apply in comparative studies on diagnostic accuracy and we hope the users would follow the updates and give their feedbacks on line.

This paper summaries the Risk Of Bias In Non-randomized Studies-of Interventions (ROBINS-I), a tool for evaluating risk of bias about Non-randomized Studies-of Interventions (NRSI), and introduces the application of ROBINS-I in a published NRSI. According to the characteristics of NRSI, evaluation field and signaling question were designed in ROBINS-I to provide essential information about risk of bias for NRSI included in systematic reviews. ROBINS-I is the tool in assessment of risk of bias in observational studies and quasi-randomised studies. Although the tool has been used in practice to some extent, but it still needs further improvement. Attention should be paid to its update and progress.

This paper introduces the revised cochrane risk of bias tool for individually randomized, parallel group trials (RoB2.0), compates RoB2.0 and the previous version (RoB1.0). And illustrates the application of RoB2.0 for a published clinical trial. As a comprehensive tool, RoB2.0 provides more information on the risk of bias for evidence synthesis. RoB2.0 is still under development and it is suggested that the users should follow the updates of the developers in the future.

This paper reviews the concept of risk of bias, followed by demonstrating why assessment of risk of bias in systematic reviews should be different from that of quality of evidence, methodological quality, reporting quality, precision, and external validity. We also discuss the recent development of tools for risk of bias assessment, the problems with the tools themselves, and the challenges in using these tools. This review may help systematic reviewers understand risk of bias assessment and the use of assessment tools.

Objective: To analyze methodological bias and variation of systematic reviews on diagnostic test accuracy (DTA)., Methods: Meta-analyses on DTA were identified through an electronic search through databases as Medline, Embase and Cochrane between 1 January 2008 and 31 December 2012. Results from Meta-analyses on 10 primary studies were included. Pairs of reviewers worked independently to extract the related data of interest, together with those original data of the primary studies. Mixed linear model was used to investigate the direction and strength of the association among the 14 studies, featuring on estimates of the diagnostic accuracy., Results: A total of 23 papers on Meta-analyses with 550 primary studies were included. Results from mixed linear model showed that significant low estimates of diagnostic accuracy in studies unsatisfying " the reference standard would likely to correctly classify the target condition" [relative diagnostic odds ration (RDOR) =0.018 6, 95% CI: 0.001 0-0.358 5]. Studies whose reference standard were not independent of the index test produced significantly higher estimates of diagnostic accuracy (RDOR= 2.396 6, 95% CI:1.242 8-4.622 7)., Conclusion: Messages as " Is the reference standard likely to correctly classify the target condition?" and " Was the reference standard independent of the index test", were the origin of the methodological bias and variation of systematic reviews on diagnostic test accuracy.