William R. Brown, M.D., Editor
Professor Emeritus, University of Colorado School of Medicine
Denver CO, USA
Note from the editor: This primer is not a comprehensive treatise on the writing of a medical research paper; several excellent references of that kind are listed below. This is intended to be a brief and concise aid to good writing. It is based on my extensive experience in writing and editing research papers (>150) and being a reviewer and editorial board member of several prestigious journals.
- Doing good research is one thing. Getting research papers published is another—and sometimes much more difficult.
- Bad writing is a common barrier to getting a paper published. Papers are rejected often and for many reasons: language problems, poor organization, lack of clarity, unnecessary length (wordiness), exaggerated conclusions, redundancy, and many more.
- These problems can be avoided by giving attention to certain principles and details.
Ⅰ. Common principles of good science writing
- Be succinct and concise. Use words sparingly—and don’t use unnecessary words at all.
- Write from the viewpoint of a reader, who doesn’t want to waste time.
- Use short, declarative sentences freely. No more than 22 words per sentence is a useful rule of thumb.
- Avoid repetition and redundancy.
- Avoid statements, such as, “Ours is the first study . . .” or “We are the first to
show . . .” Rather, say, “To our knowledge, we are the first to show . . .” - Avoid excessive use of the passive voice (but not when it sounds or reads better than the active voice): “Our technician incubated the cells for 24 h.” Better: “The cells were incubated for 24 h.”
- Avoid use of “there is,” “there are,” “there were,” such as, “There were 10 patients enrolled in our study.” Better: “Ten patients were enrolled in our study.” Even though this is passive, it sounds better and is more concise.
- Avoid “-sion” or “-tion.” For example, “The surgeons made the decision.” Better: “The surgeons decided.”
- Avoid pompous words, such as “utilize,” “symptomatology,” and “etiology.” Use “use,” “symptoms,” and “cause” instead.
- Avoid unnecessary or excessive use of abbreviations and acronyms. Unless a term is unusually long and complex do not abbreviate it if it will be appear fewer than four times in the document.
- Consider making a list of abbreviations at the beginning of your paper, especially if you’ve used numerous and unfamiliar abbreviations.
- Avoid value judgments, such as, “our impressive results” or “our striking results.” Let the reader decide the merits of your findings.
- Ask an objective colleague or friend to review your paper before you submit it.
- If you are not a native-English speaker, ask one or more to edit your paper.
- When you think you’ve finished writing the paper, you haven’t. Set it aside for several days (or weeks), pick it up, and re-read it. You’ll be surprised at the blunders you’ve made and revisions you want to make—in particular, shortening it*.
- Freely consult sources on writing and editing:
- Zeiger M. Essentials of Writing Biomedical Research Papers. 2nd ed. New York, NY: McGraw Hill; 2000.
- American Medical Writers Association. Workshops and self-education
- AMA Manual of Style
- The Chicago Manual of Style
- The McGraw-Hill Handbook of English Grammar and Usage
- International Committee of Medical Journal Editors
- Huth EJ. Writing and Publishing in Medicine. 3rd ed. Baltimore, MD: Lippincott, Williams & Wilkins; 1998.
ⅠⅠ. The content of a medical research paper
a. Title
- The title should be concise, specific and contain key points of the work.
- Subtitles may be used to expand on the title, but should not contain key elements of the study as a supplement to the title. For example,
Original: Clinicopathological features, pattern of lymph node metastasis and prognostic factors in patients with EGJ carcinoma after total gastrectomy: a retrospective study on 123 patients in China.
Better: A Retrospective Study of Chinese Patients after Total Gastrectomy for Treatment of Esophagogastric Junction Cancer.
- Avoid or don’t use abbreviations, especially uncommon ones.
- Capitalize the first letter of each major word.
b. Abstract
- Check and abide by the journal’s word limit, which often is 250 words or less.
- Follow the journal’s guidelines for style (unstructured or structured). If structured, note the subheadings, e.g., objectives, methods, results, conclusions.
- State objective(s) clearly.
- Make conclusions respond to objectives.
- Make conclusions agree with conclusions stated at end of the paper’s discussion.
- Do not overstate conclusions and significance of results or findings.
- Avoid abbreviations. Many journals do not allow abbreviations in abstracts.
- Avoid unnecessary experimental detail.
- Do not cut and paste exact sentences from the body of the paper, but summarize or restate only the important points.
c. Introduction
- Clearly and concisely state the question to be addressed by your study, and why it is being asked.
- Review briefly the existing state of knowledge and cite relevant literature references. If appropriate, state what is not known.
- State the method(s) or experimental approach(es) used to answer the question, including whether the study was retrospective or prospective.
- State the novelty or uniqueness of the research.
- The answer to the question should not be given in the introduction.
- Keep the introduction brief and focused.
d. Materials and Methods
- This section should give enough detail to allow the reader to evaluate the work fully and repeat the experiments exactly, but it should be as brief as possible.
- Materials
o Chemicals, reagents, drugs
o Experimental materials, animals, human subjects (much data on human subjects can be given in tables—and should not be repeated in the Material and Methods section). - Methods
o Study design
o Order in which the studies were conducted
o How studies were conducted
o Rationale for the studies
o IRB approval and consent of subjects
o For randomized trials and systematic reviews know and adhere to PRISMA (Preferred Reporting Items for Specific Reviews and Meta-Analyses).
o For randomized control trials know and use the CONSORT flow diagram. - Cite references of relevant published methods.
- Analysis of data
o State statistical methods used, and cite references for methods that are not well known.
o State the P value considered significant.
o Reference computer program(s) used.
e. Results
- The Results section should contain only results pertinent to the question asked in the Introduction.
- But the section should include results whether or not they support the hypothesis being tested.
- Organize the results from the most important (usually the question or hypothesis of the study) to the least important.
- The section may include experiments that were not initially planned but were determined on the basis of results obtained from planned experiments.
- Experimental AND control results should be stated.
- Most data should be presented in tables or figures—and not repeated verbatim in the Results text.
- Report statistical analyses. Where appropriate, give 95% confidence intervals as well as P values.
- When writing the paper, consider writing this section first; the results are the groundwork for the entire paper.
f. Discussion
- Begin the discussion with a concise statement of the study’s most important results, especially those that respond (whether positively or negatively) to the question(s) asked initially.
- Don’t begin with a review of others’ results and publications; these items can diminish the impact of your results and should come later.
- Don’t bury your results amidst excessive detail of other studies.
- Don’t reiterate your results in detail. Focus on their significance.
- Do discuss the significance and importance of your results—without exaggeration or excessive exuberance.
- Do compare your results to existing knowledge as indicated.
- Explain conflicting results. Defend your results when needed.
- Give credit fairly to others and yourself.
- Explain the newness or novelty of your study.
- Don’t write an exhaustive review of the literature, only that which is related to the present study.
- Mention the limitations of your study.
- Suggest future studies that may be indicated on the basis of your findings.
- Conclude with a clear and concise summary of your results and their significance. Tie the conclusion into your originally stated goals.
- Make the concluding paragraph or statement(s) agree with those of the abstract.
- Do not make the discussion unnecessarily long. Keep it concise and focused.
g. Figures and Tables
- Figures and tables should relate most of the results of the paper and, often, details of the material and methods used.
- The reader should be able to grasp the main message of the paper by referring to the tables and figures, without much reference to the text.
- Table and figures should be designed clearly and concisely, with informative titles and legends and neatly aligned columns and rows.
- Micrographs should be clear, with appropriate size and dimensions.
- Figures should be easy to read, with legible lettering.
- The same data should not be presented in tables AND figures.
h. References
- A research article is not a review article, even though authors often fail to make the distinction and include too many references,
- References cited in a research article should be only those that have direct relevance to the current research.
- The reference list should be as short as needed to accomplish the relevancy goal.
- Articles that have not been accepted for publication should not be included in the reference list.
- Every reference in the reference list must be cited in the text, and every reference in the text must be included in the reference list.
- Check the reference style required by the journal to which you will submit.
- A commonly used style is the so-called Vancouver style and is available on the Internet: http://www.hsr.it/biblio/uniform.html.
- References that are cited in the text numerically should be listed in numerically in the reference list.
- References from the Internet can be cited as: Author. Article title. Available from: url:http://Internet address or World Wide Web address.
- An excellent online resource for references is Dr Able Scribe PhD, AMA Style Stat at http://www.docstyles.com/amastat.htm#Sec40.
i. Indexing terms (key words)
- Many journals require that you provide a list of indexing terms that readers can use in searching for your article.
- Often, the terms are to be chosen from the medical subjects heading (MeSH), which is listed annually in the January issue of Index Medicus.
- Use terms that are as specific to your subject as possible. For example: omeprazole rather than proton pump inhibitor.
- Some journals request that you use indexing terms that are not used in the title.
*Benjamin Franklin, 1750: “I have already made this paper too long, for which I must crave pardon, not having now time to make it shorter.”
- Medical personnel often use ‘fancy’ talk, and it should be discouraged. Examples: “ology” refers to the study of something. Thus, etiology, strictly is the study of cause and is not the cause. So, “The etiology of diabetes is ……” is bad. Write, the “cause of diabetes is …….”
- Similarly, “The patient’s symptomatology was/were headache, fever, and nausea” should be written as “The patient’s symptoms were headache, fever, and nausea.” And “the methodology used in this paper were . . .” should be “The methods used in this paper . . .”
- Rather than “the findings were as follows” write, “The findings were” (colon should not be used at all).
- “As described previously” usually is redundant. “As described” is adequate since “described” already is past tense.
- “A total of” usually is unnecessary: “A total of 39 patients” should be, “Thirty-nine patients.”
- “Suffer from” should be avoided. Not the “patients suffered from diabetes,” rather, “The patients had diabetes.” “Patients were diagnosed with . . ” should be “The diagnosis of XXXX was made in the patients” or “The patients had the diagnosis of XXXX.”
- “Utilize” is overused; usually it should be “used.”
- “Compared with” or “compared to” often is used incorrectly. “A was bigger compared to B” should be “A was bigger than B.”
- “In terms of” should be avoided. “In terms of complications, diabetes has more than has gout.” Change to “Diabetes has more complications than has gout.”
- Example of wordiness: “Measurements of blood glucose were conducted,” should be “blood glucose was measured,” or “the concentration of blood glucose was measured.”
- “The patient has had 3 surgeries on his/her hip.” This is incorrect usage, which has become common in medical circles as well as in lay. Surgery is a discipline; operation is the procedure. “The patient has had three operations on his/her hip” is correct. Use “operation” instead of “surgery” most often. For example: “Finished the surgery” is better written as, “finished the operation.”
- Put a space between numerical value and unit, such as 23 g; not 23g.
- Patients should not be referred to as “cases.” For example, “A case of hepatitis was treated” should be, “A patient with hepatitis was treated.” It is permissible, though, to refer to cases when not specifically referring to patients, such as “Our series consisted of 100 cases of hepatitis.”
- “Data” is plural, although nowadays it often is used as singular. I prefer the plural usage, such as “The data were reviewed.”
- Pay attention to the conventions for italicizing and capitalizing the names of microorganisms, such as bacteria and fungi.
- “Criteria” is plural and “criterion” singular.
- “Different” should not be used when “various” is intended. For example, “The drug induced many different reactions” should be the “drug induced various reactions.”
- Avoid “etc.” It says nothing.
- “Patients that have . . .” is better as “Patients who have . . .”
- You can get rid of many “there is” or “there are” like this: “There is a growing number of births.” Replace with, “The number of births is increasing.” “There were 15 studies that assessed . . .” Replace with, “Fifteen studies assessed . . .”
- Use of “by” is tricky. Example: “The publication biases were evaluated by Begg and Egger methods. Begg and Egger did not do the evaluation. Better: “The publication biases were evaluated with Begg and Egger methods.” Or, “The publication biases were evaluated by use of Begg and Egger methods.”
- Usually the generic name of drugs is preferred over the trade name. (The trade name is always capitalized and the generic name is not.)
- Traditionally, et al. has a period after al. because it is an abbreviation for et alia. However, the period is increasingly omitted in modern style guides.
- Hyphens are needed for terms such as 12 mm trocar. 12-mm trocar is correct.
- Expressions such as “. . . results in a reduction,” which are used ubiquitously, can be shortened to “. . . reduces.”
- “Comparable” and “similar.” “Comparable” means that things can be compared. “Similar” means they are about the same. Example: Outcomes with the two treatments were similar–not comparable.
- Avoid the double qualifier: “Findings suggest that A may be larger than B.” Better: “Findings suggest that A is larger than B.”