Across the health professions, it is often assumed that medical students, residents, and faculty inherently absorb the knowledge on how to construct a successful journal manuscript. That is a fallacy. Crafting a clear and logical message that presents one’s data and conclusions can be incredibly challenging. Dr. Craig Newgard, Oregon Health & Sciences University (OHSU), shares his recipe for success in an itemized fashion. He also reviews this template in a recent podcast with the SAEM Research Learning Series.
[person name=”Craig Newgard, MD MPH” title=”OHSU Professor of EM” image=”https://www.aliem.com/wp-content/uploads/2017/11/Newgard.jpg” twitter=”newgardc”]
Craig is the Director of the Center for Policy and Research in EM at OHSU with over 140 peer-reviewed journal publications in the areas of EMS, trauma systems, field triage, and advanced statistics.[/person]
Instructions and Disclaimer by Dr. Newgard
I developed and crafted the following manuscript template to be used by authors to catalyze the writing process. The template is not specific to any one journal, but works for most. Writing can be highly individual, with different perspectives and approaches. The following approach is what I use and includes my embedded editorial perspectives with examples in italics. The “XX” areas are sections for you to fill in.
When you first start writing, don’t worry about getting it right the first time – type whatever ideas you have, even if in an outline format or short notes (both a version of free-writing). Organize these initial thoughts under the headers and subheaders below. The rule is many revisions and iterations. You can polish the text later.
Double-space all text for the abstract and core manuscript text.
- Keep the title succinct, representative of the study, and interesting.
- This might be the only information that someone sees when doing a literature search.
- Don’t answer the study question in the title – it is rarely so simple.
|Item||Description and/or Example|
|Author List||Author order is important and closely tracked in academic circles for a variety of purposes, including promotion and tenure. The first author should be the lead person who developed the study idea and is doing the lion’s share of the work and writing. The second author is sometimes reserved for the biostatistician or a very involved co-author who doesn’t quite meet criteria for senior author. The last author is the “senior author” and should be the person supervising the group, providing guidance and key insights, often the content expert or senior person working the most behind the scenes to make the project happen.|
|Word Count||The target is typically 3,000 – 3,500 words, although this differs by journal.|
|Additional Presentation Information|
|Address for Correspondence||[Your name]|
Department of Emergency Medicine
Center for Policy and Research in Emergency Medicine
Oregon Health & Science University
3181 SW Sam Jackson Park Road, mail code CR-114
Portland, Oregon 97239-3098
Phone (503) 494-1668
|Funding||If the project was supported by grant funding, the funder likely has specific language to use in this section – check with them.|
- Clean, succinct, clear writing is critical to a paper being accepted for publication.
- Attention to detail and organization is very important.
- Make sure your study objectives are clearly answered in the Results, picked apart and interpreted in the Discussion, and cleanly summarized in the Conclusions.
- If you list certain topics or outcomes in a specific order in the study objective, keep this order throughout the Results, Discussion, Conclusions, and Tables/Figures.
- Anything that makes your paper easier to read increases the chances for acceptance (and readers following and remembering what you say).
|Item||Description and/or Example|
|Conclusions||Be sure to stick to conclusions directly supported by the results.|
Generally write 3 paragraphs, with the total page count no more than 1.5 double-spaced pages. Save the detailed literature exploration for the Discussion. The Introduction builds your case for doing the project. The following subheaders come from Annals of Emergency Medicine and are a good way to organize your Introduction. For most journals, you do not need to include these subheaders as text in the manuscript.
- Goals of the investigation / study objective(s)
The following subheaders are helpful in organizing the methods and guiding the reader. I include them in every manuscript for clarity.
|Item||Description and/or Example|
|Study Setting||Provide additional relevant detail about the study setting. Why were these sites or hospitals chosen? How do their systems work? Why were these dates selected? There should be enough detail for the reader to compare the setting to their own and decide if the results apply to their patient population.|
|Patient Population||Describe inclusion and exclusion criteria, justify this patient sample, and the inclusion criteria.|
|Variables||List and describe all variables you collected and used for this analysis. There should be no variables included in the Results that are not listed here and under the Outcomes section.|
|Outcomes||List both the primary and secondary outcomes. Describe the rationale for using these outcomes, include supporting references. Why were these outcomes selected? Why are they important?|
|Statistical Analysis||Include sample size calculation, if performed. Describe any models used, how they were built and variables selected, plus model fit metrics. Mention subgroups or strata, if relevant. Describe how missing values were handled. Provide enough information for someone else to replicate the results. Use citations for your methods whenever possible. Consider justifying the use of certain statistical strategies (including citations) if multiple options exist.|
List your statistical software.
The Results can generally be summarized in 4-5 paragraphs.
- General description of the sample and % primary and secondary outcomes. Refer to Table 1 for sample characteristics. Minimize redundancy between results presented in text versus results provided in the Table.
- Main results: as above, minimize redundancy between tables/figures and Results text. Highlight key findings, as needed.
- Additional results
- Results from the sensitivity analyses and model fit (if multivariable analysis done)
- Of the XX injured patients transported by EMS during the 3-year period, XX had matched hospital records and formed the primary sample for analysis. XX (XX%) had ISS ≥ 16 and XX (XX%) died during their hospital stay. Characteristics of the primary sample are demonstrated in Table 1.
The Discussion section generally has as many as 4 paragraphs.
- First paragraph: Summarize the study findings briefly, and set the stage for the rest of the discussion.
- Remaining paragraphs: Place your results in context of existing literature on the topic, and describe why they are important. While the Results section needs to be objective and factual, the Discussion is your chance to interpret your findings, put them in context of current literature, and make them applicable. Why did this study matter? Why are the results important? How might they be used? The Discussion should not simply be a review of the literature. Focus on studies directly related to your project. Consider implications for policy or changes in clinical care.
- In this multi-region study, we demonstrate XXXX.
This typically has 2-3 paragraphs. This is your chance to be open and transparent about study limitations. Show the reviewers (and readers) that you understand the limits of the data and analysis. However, you don’t want to kill your manuscript here by focusing too heavily on the limitations. Find some balance.
This is traditionally a brief section with only 1-2 sentences. Stick to conclusions that are directly supported by your results, and highlight those with a potential impact on policy or patient care.
Start on a separate page, immediately following the core manuscript text. For most journals, they want references listed in the order cited in the manuscript. Single-space or double-space based on the journal’s preferences. Use a reference manager (e.g., EndNote).
Tables / Figures
Most journals cap the total number of tables and figures to around 4-5 (check the Author Instructions). Figure out what is most important to present and how best to spend your page space. Each table/figure starts on a new page.
How you set up Table 1 is up to you, but it should adequately describe relevant characteristics of the primary sample and potentially include key comparison groups (e.g., in a RCT) or strata as additional columns. If the sample size is not included in the table, put it in parentheses after the title. Every table and figure should be able to “stand alone” (i.e., make sense if it was removed and pasted into another document). Many journals no longer want p values for these tables – only include them if useful.
- Table 1. Demographics/characteristics of the sample (n = 89,441)
People like pictures, so give them what they want. Figures are also quite helpful in conveying information and relationships between the data. Try to find ways to translate your key findings into graphic format and maximize the amount of information available in the figure without confusing people. Make them clear and informative. When possible, represent the findings in graphic format, rather than a table (except for Table 1). Attention to these details will make your paper more marketable and therefore more likely to be accepted. It will also create a work where the key findings are easy to consume and more likely to be remembered.
An appendix is not required, but can be a useful tool. If you have a really long, complicated Methods section, consider moving some of the information here and referencing the Appendix in the text. If you have too many tables and figures, but feel they are all important (or want them available to reviewers and readers), consider moving some to the Appendix. Many journals keep appendices as online material to save page space. If you have an Appendix, it needs to be cited in the core text, something like:
- Results from additional stratified analyses are included in the on-line Appendix.