Four common fatal mistakes when writing a paper: No. 4

My many years of editing experience have led me to identify four very common and unfortunately fatal mistakes that scientists and physicians make when they write their paper. These mistakes are so serious that you risk immediate rejection if you make even just one of them. Here I will describe one of these mistakes and show how you can avoid it.

FATAL MISTAKE 4: Not clearly describing patient selection and comparator groups 

Example: monocentric retrospective study comparing a nutrition program to standard care after radical surgery for advanced gastric cancer.

Methods: “This cohort study included patients with advanced non-metastatic gastric cancer who underwent radical surgery in 2014-2015.”

Results: “Patient characteristics are shown in Table 1. The nutrition program improved variables 1, 2, and 3 but not variables 4, 5, and 6.”

This paucity of information leaves the reader with many questions:

• Were the patients ALL CONSECUTIVE patients (or randomly selected, or a convenience series?)

• What were ALL the inclusion and exclusion criteria?

• How many patients were excluded, and for which reason? How did the nutrition-program and standard-care patients compare in terms of numbers excluded for certain criteria?

• How did the nutrition-program patients compare to the standard-care patients in terms of baseline and perioperative variables?

Why is this a serious error?

Because it is essential to disclose the information that will show if there is patient selection bias that could affect the interpretation and generalizability of your results.

How can you avoid this mistake?

By consulting the relevant biomedical study reporting guideline. In this example, STROBE would be appropriate. A randomized controlled trial would require CONSORT. The relevant guideline for your study can be found here. I have also constructed a simplified STROBE guideline that contains relevant CONSORT elements.