The peer-review process instills rigor into the publication of scientific manuscripts. Editing the manuscript in response to the reviewer’s comments adds clarity to the paper. Reviewers often ask me what is expected of them, whether their feedback was useful. This editorial and the following one will provide some ideas on how to approach a review. At the Journal of Swine Health and Production, each manuscript is reviewed by two or three people with their own areas of expertise. Together, we expect they will identify the strengths and weaknesses of the manuscript. What follows is a more general approach to reviewing the introduction and methods sections of manuscripts.
It takes substantial time to thoroughly review a manuscript. Personally, I write comments on the paper as I conduct my review and put question marks beside confusing sections.
In the introduction, the author must provide sufficient background information to understand what is already known about the subject matter and to justify why this research was necessary. Reviewers determine whether the authors have missed key published manuscripts from the literature. The introduction ends with the objective of the study. The objective(s) must be clearly stated and focused so that the reviewer understands the purpose and the scope of the study. The reviewer can then determine whether or not the study design is appropriate to fulfill the purpose. Pertinent questions to ask when reviewing the purpose are “What is the population to which the results will apply?” “Does this topic fit within the mandate of the Journal of Swine Health and Production?” “Is this applied research of interest to pig or pork production?” Manuscripts focused on disease and production issues from anywhere in the world are acceptable. We do not typically publish basic research work with no link to pig or pork production. Next, the reviewer asks “Do we already know the answer to this question or will this study provide more information for our readers?” All research must be validated by more than one research team, in more than one institution, in more than one environment. This is a judgment call. If sufficient research papers indicate that gilts have lower litter size than sows, we do not need to publish that information again. However, it is important to determine whether there is a novel aspect to the manuscript before it is rejected.
The materials and methods (methods) must have sufficient detail so that someone could reproduce the study. If when you finish reading the methods you do not think you know exactly what the researcher did – something is missing. If the study is very complicated, would a timeline figure help the reader? As you read this section, write down your questions about exactly how it was done. What details are missing? If I have a question in the methods that is later answered in the results, that sentence is in the wrong place. Reviewing this section of the manuscript in a thorough manner is key. If there are “fatal” flaws in the scientific study design, the manuscript may have to be rejected. Was there an animal care committee evaluation? Do you believe the author can achieve the purpose with the methods outlined? What type of study was conducted and did the authors follow the accepted standards for this type of study? Are the animals (farms) used in the study representative of pigs in commercial farms? Would this make the pigs atypical or would it bias the results? Is there anything about how the animals were managed that could bias the results? As an example, excluding all small or ill animals from a study means that the results apply only to healthy, good-sized pigs. This would not cause you to reject the manuscript, but will caution you when you evaluate the implications of the work. It will help you to determine how applicable the results are to commercial swine. How were the comparison group or control animals selected and managed? Are there clear definitions for all variables? For example, did the authors use farrowing rate or adjusted farrowing rate? If they analyzed feed conversion, did they include dead or culled animals? Did they control for factors that might affect the association between the treatment and the outcome? As an example, if the outcome is litter size and the treatment is breeding method, did the researchers collect information on parity and age of semen? Further, what factors, other than the one of interest to the authors, could explain the association found in the study? How were the factors measured? Are the laboratory tests described in sufficient detail to reproduce? Were the case and control groups evaluated equally? What was the bias in further follow-up – such as postmortem evaluation? Was the data collected worthy of analysis? Did the authors apply the appropriate analytic tests? For every comparison, the author must make it clear how the analysis was conducted. It is not sufficient to state that linear regression was used. What were the dependent (outcome) variables, what were the independent variables (covariates), and was the analysis done in a uni- or multi-variate manner? What was the unit of analysis – pig, pen, side of barn? If there are areas of the methods that you feel unqualified to judge, write that down too. The editor will know you expect another reviewer to evaluate that portion of the manuscript. When necessary, Dr Zvonimir Poljak will evaluate the statistical analyses of submitted manuscripts.
Once you have finished the introduction and methods section, ask yourself again, “Do I understand the purpose? Do the methods allow the author to address the purpose? Do I have sufficient information to repeat the study in exactly the same manner as the author?”
-- Cate Dewey, DVM, MSc, PhD