Write up
Frozen
Writing a manuscript is a challenging task, but easy to master.
Introduction, methods, results, discussion are the main components.
Start with the results section.
Copy forward the methods section and make edits/ changes as appropriate to your paper.
Introduction
Keep it 1 to 1.5 pages long with double space between the lines.
All final studies have the same patient sample, same intervention and therefore the same goals.
So, start writing introduction while you are collecting your data.
Keep tab of appropriate references/ citations along the way.
Methods
Methods comprise of search strategy, study selection, data extraction, outcomes assessed, definitions, and statistics.
You will find pretty much similar lay-out in any meta-analysis paper.
These sections most often will address the basic requirements on the PRISMA & MOOSE checklists.
Edit and change details pertaining to your paper.
Results
Start with search results (summarize it in a PRISMA flow diagram) followed by population characteristics.
Talk about risk of bias assessment in one paragraph.
Present the main meta-analysis outcomes, followed by subgroups analyses and meta-regression.
Pay attention to avoid clutter and reader fatigue by referring the reader to a table or figure and avoid redundant information.
Validation
Important feature to highlight in any meta-analysis is the heterogeneity and trend of the pooled results.
A high heterogeneity does not always mean the study is bad.
In my papers, I always dedicate a paragraph on 'validation of meta-analysis results' and talk about sensitivity analysis, heterogeneity, publication bias.
Occasionally, you can also mention the quality of evidence based on GRADE rating.
Discussion
Keep it 2 to 2.5 pages long.
Avoid too much theoretical rhetoric, which is usually the norm for a review paper but not for a meta-analysis.
Collect points to highlight from primary studies while you are at your data collection stage.
Open your paragraph with reporting the primary outcomes, plan for a paragraph comparing your paper to current papers in literature, a strength/ limitations para and finally the conclusion.
Pay particular attention not to cheerlead your results and do not highlight statistical significance.
Conclusion
Must be short & sweet.
Only present findings or recommendations that is supported by your study data and avoid deviating into too much lateral info.
Always comment on future studies or research warranted or not.
Manuscript
The final product should have the opening title page (also have it as a separate file, as some journals might ask it to be uploaded as a separate file).
Author contributions, emails, disclosures, ethics statements.
Abbreviations, abstract.
Manuscript body that includes all items discussed above.
Legend.
Use page-breaks (cntrl+enter or cmmnd+return) between file components, tables, fgures, references are better to be on separate pages with page-breaks between them.
A supplementary file that typically describes the risk of bias assessment table, literature search strategy, forest plots, funnel plot, PRISMA checklist & MOOSE checklist.
write up video part1
Lets start that manuscript!
write up video part2
Start that introduction section while you are collecting your data
write up video part3
Go ahead and ask for my past write-up template, just edit it to fit your paper
write up video part4
Finish up, finalize, add those extra information from your 'remarks' column
Start
Key is to start, just start, start somewhere, start anywhere but start.
Final manuscript must be reviewed at least 3 times.
Ask friends and team members to take a look and report errors.
It will be intimidating at the beginning, you will doubt yourself and your study along the way, others will always help fuel that doubt, but don't stop and believe me the final product will always be a masterpiece.
When its all done, pat on your back, you have now created something beautiful, now the next step is to publish.
