Systematic reviews of clinical laboratory studies: pilot risk of bias tool developed by consensus

Abstract

Introduction

Some research studies aim to elucidate pathophysiology by examining blood or tissue markers in relation to clinical findings. In COVID-19, this has led specialists to promote treatment options based on single studies without systematic appraisal and critical summaries of the data. As we could not identify any published tools for this purpose, we developed a pilot risk of bias tool by consensus, and report here on our approach.

Methods

Using an expert consultative consensus process, a panel of five topic experts were guided through a set of iterative steps to develop questions intended to elicit information about the study methods and reporting in clinical laboratory studies. The team piloted the tool in three clinical laboratory studies, and then applied it formally as a component in assessing a hypothesis about mechanisms in the post-COVID-19 condition as part of a Cochrane review.

Results

The pilot tool assessed study quality and bias across three domains applicable to comparative and single-arm clinical laboratory studies: collection and handling of samples, experimental methods, and reporting of the results. In the Cochrane review, the tool identified substantive risk of bias in the included clinical laboratory studies.

Conclusion

The plethora of COVID-19 research has highlighted the need for formal methods to systematically appraise clinical laboratory studies related to disease pathology. This tool provides a systematic approach to appraise the validity of these studies. Our process may guide others in the development of appraisal tools in areas where they are needed. Given the relationship between clinical laboratory studies and the development of medical treatments, further development of this risk of bias tool is important for evidence-based healthcare and research.