Skip to main content
. Author manuscript; available in PMC: 2023 Mar 1.
Published in final edited form as: J Clin Epidemiol. 2021 Nov 3;143:186–196. doi: 10.1016/j.jclinepi.2021.10.023

Table 1.

Synthesis of harms: Challenges and opportunities

Series paper Description
Qureshi R, Mayo-Wilson E, Li T. Summaries of harms in systematic reviews are unreliable Paper 1: An introduction to research on harms.[1] This paper describes key aspects of assessing harms. First, it identifies the various ways that harms are classified, including the terminology used to describe grouping of harms and different data dimensions such as rarity, severity, relatedness, and timing. Second, it explores different methods for assessing harms in clinical research. Third, it describes the challenges associated with analyzing, reporting, and synthesizing harms.
Qureshi R, Mayo-Wilson E, Rittiphairoj T, McAdams-DeMarco M, Guallar E, Li T. Summaries of harms in systematic reviews are unreliable Paper 2: Methods used to assess harms are neglected in systematic reviews of gabapentin.[2] Part 1 of an overview of gabapentin systematic reviews that serves as a methodologic study for how harms are synthesized. This paper examines the methods used in systematic reviews and meta-analyses to assess harms across all stages of the review process and compared these with recommendations for assessing harms in reviews. We found review methods are focused on addressing questions of benefit and the tokenistic inclusion of harms in reviews means that the methods may not meet the needs to produce a valid synthesis of harms.
Qureshi R, Mayo-Wilson E, Rittiphairoj T, McAdams-DeMarco M, Guallar E, Li T. Summaries of harms in systematic reviews are unreliable Paper 3: Given the same data sources, systematic reviews of gabapentin have different results for harms.[3] Part 2 of an overview of gabapentin systematic reviews that serves as a methodologic study for how harms are synthesized. This paper evaluates the results for harms across these reviews, with a particular interest in the consistency of results given similar supporting data. We found significant challenges in the selection of harms to assess and report, largely driven by reviewer preferences as opposed to standardized approaches, leading to different summaries of harms across reviews.