STFM Blog

Discussion/Conclusion 
The discussion section of a paper restates what the paper found and then goes on to explain how these findings fit into the existing literature and how they impact the field. We also expect the authors to document how the study answered their original research question. Occasionally, we see a paper with an interesting finding, but that finding is not related to its a priori research question. This is something to review.  

One of the biggest mistakes we see in discussion sections is when authors overstate the importance or significance of their findings. For example, a study of 12 residents doing a rotation to learn about social determinants of health may find a significant change in their behavior. Still, it doesn’t provide evidence to add such a rotation to all residencies. Results of such a small study cannot make such an impact. We also expect reviewers to conduct a limited literature search themselves. If the authors claim that their paper is the first to do something, then the reviewer has the responsibility to double-check that fact. Reviewers should search PubMed and one other index (Scopus, CINAHL, ERIC, etc.) for verification. Reviewers should not use AI for this task.

We also expect authors to include limitations to the study. Most papers could have had larger sample sizes or more diverse samples. The reviewer wants to see that the authors have considered the study’s limitations and have commented on how the results remain valid despite them. The end of the discussion section should include a summary of the study, a repeat of the key results, and a statement about future research. Do the authors suggest repeating the study with more participants, at several different residencies, or broadening the sample to include medical students? Reviewers should ensure that the authors provide limitations, a summary of findings, and have placed their findings in the context of the current literature. When authors have not done this, they can rewrite the discussion using the direction provided in the article by H.G. Welch (https://cancer.dartmouth.edu/sites/default/files/2019-05/papertrail.pdf)

Tables and Figures 
Using figurative methods to explain results (and sometimes methods) can be a great way to help readers understand your study. As such, we expect reviewers to carefully review all tables and figures. Are they easy to understand? Do the numbers add up? (This is more common than you would think.) Do they enhance understanding of the results? If the tables and figures are confusing or unclear, the editors want to know. Because if you, as the reviewer (someone who is spending a lot of time with the paper), don’t understand, then the reader won’t either. As the reviewer, you expect Table 1 to describe the demographics or other characteristics of the study participants. Then you expect to see more tables with the study outcomes, including statistical calculations to demonstrate significance. We would like reviewers to comment on whether a figure or graph would help explain the results. 

All tables should stand on their own. To relate the table to the text, the authors should refer to the table after a summary sentence, e.g., “Patient demographics are included in Table 1.” The only other information from a table to be included in the text is the most significant finding, such as “Although we conducted our study in Salt Lake City, UT, 96% of our respondents identify as Black.” (This is significant because Salt Lake City, UT, has a notoriously small Black population (about 2.5%). If you find that the authors are repeating everything from the table in the text, you should encourage the authors to choose the table or the text.

References   
As stated above, reviewers should do a brief literature search about the topic covered in the paper. It is excellent to suggest papers for the authors to consider if they are not included in the reference list. As many reviewers are experts in the field, they can also suggest documents that they have authored. Other specific aspects of the reference list to review include the type of papers (i.e., meta-analyses or systematic reviews vs. small non-randomized trials). Are the documents referenced old or current? Do you know of any studies that are not included? One reason you may be asked to serve as a reviewer is that you have worked in a similar area, so you would be familiar with the literature. Do the authors cite their own papers over other publications? In this age of AI, we also suggest checking the DOIs of the articles in the reference list to ensure they are valid. Often, real DOIs are reported with references that do not exist. Reference checking is an important quality control step, and reviewers should hand-check references for accuracy. (Please do not use AI for it; AI may lie).

How Peer Review Can Help You:
Working as a peer reviewer can help you become a better writer. You can see how other people structure a paper or even how they describe methods, and can take away from that ways that you want to do it in your own work (or not). Being a peer reviewer will also help you stay up to date with the science in your field. To be a highly regarded family medicine academician, you will need to have a sense of what journals are publishing. Thirdly, being a high-quality peer reviewer can advance your career. Editors are often leaders and will invite skilled peer reviewers to be on Editorial Boards or become part of the editorial team. It is a great way to build your reputation.

Peer reviewing can also give you ideas and help you further your scholarship. JR once reviewed a paper for the New England Journal of Medicine (NEJM) that described hypothetical diversity interventions. This inspired him to submit an article to the NEJM on verified diversity interventions that was eventually published.

We hope that these blogs have helped support your work as a peer reviewer. Please visit the reviewer page on Family Medicine’s website for other resources or to sign up (Family Medicine).