Being a successful expert reviewer =/= being so as an expert writer.
Why?
➡️ Difficult to be self-critical
➡️ No "formal" training in writing
How to improve then (besides accumulating years of exp😅)?
Here are the Top 10 Tips from @radiology_rsna Scientific Style Guide!
3/
1⃣ : Formulate a specific, hypothesis-driven study purpose.
Don't be too general! Go straight to the point... and don't forget to close the loop by addressing the purpose in your conclusion.
4/
2⃣ : Material & Methods Essentials, part 1
1st paragraph must include:
➡️ IRB approval
➡️ Patient (written/verbal) informed consent (or waiver), and HIPPAA approval (in US only).
➡️ Registry name & number (if appl.)
➡️ Animals care committee approval (if appl.)
5/
3⃣ : M&M Essentials, part 2 (still in paragraph 1):
Specify study type (prospective or retrospective).
Don't forget to name the prospectively enrolled patients as "study participants" and not "patients"!
6/
4⃣ : M&M Essentials, part 3!
Give details regarding patient selection (incl./excl. criteria & sampling technique), # of invovled centers, study dates, and power calculation (and/or justification of sample size).
7/
5⃣ : Results!
Table 1 should be your patient demographic & clinical characteristics.
Figure 1 should be your patient incl./excl. flow diagram.
And both should be summarized in the 1st paragraph of Results.
8/
6⃣ : For AI papers only 🧠💻
Mention how your algorithm can be accessed in your M&M section.
Also give the unique identifier for revision of the code used in your work.
9/
7⃣ : Do NOT provide p-values alone without mentioning the corresponding comparison values.
We need both to appreciate statistical AND clinical significance (especially in the era of Big Data).
10/
8⃣ : If there are readers in your study, don't forget to include a reproducibility subsection in M&M and Results.
Don't go below 3 readers. And avoid consensus reads.
11/
9⃣ : Wherever there is a percentage, include the corresponding numerator & denominator and unit of measure.
If these percentages (proportions) are part of your main results, also include confidence intervals.
12/
🔟 : Consult a biostatistician BEFORE conducting your study; FOR your results; and BEFORE submitting.
Don't even bother trying without one in your team.
Did you know? Biostatisticians are the most commonly added co-authors in authorship change requests for @radiology_rsna.
13/end of 🧵
That's it!
I hope you have enjoyed & learned anything useful from this short #tweetorial. Please share!
May your future @radiology_rsna papers (scientifically) rock!
Interested in learning more? Scan below ⬇️ to access the whole Radiology Scientific Style Guide👍
• • •
Missing some Tweet in this thread? You can try to
force a refresh