Scientific writing is often considered a “necessary evil” to ascend the academic ladder. After all, the phrase “publish or perish” is still commonly used. Academic promotion, improving knowledge of a subject, establishing professional contacts, disseminating knowledge,1 and obtaining grant funding are a few reasons why physicians choose to publish.
Physicians may feel the pressure to publish, but scientific writing is a learned process and one in which few academic physicians have received any formal instruction.2–6 Although scientific writing often has a reputation for being dense or unreadable,7–11 we hope to assist physicians to develop a manuscript that is enjoyable to write and read.
We will assume that you have already completed a well-conducted scientific study and are ready to write up your findings for a peer-reviewed scientific journal. If you are new to scientific writing, it is helpful to have a mentor, particularly an experienced writer in your field who can guide you. If you are writing about a research project, you should have formed a hypothesis to your research question before analyzing any of the data. You will want to do a thorough literature search of your topic so that you are familiar with all published work. Writing an outline may seem elementary, but it can be an important tool to organize your thoughts.12 Think about what journals you may want to publish in and follow their publication guidelines as far as word count and paper structure. Choosing your journal early on will help you to write for the readers of a specific journal. You do not want to “write down to” your readers or include details that have been published previously, such as an established surgical technique. You also want to think about the conclusions of your paper before you begin to write—the “so what?” question. Why is this project novel? What makes it different from other published studies on this topic? Why is it clinically relevant? Table 1 gives a list of “dos” and “don'ts” for scientific manuscript writing.
WRITING THE ABSTRACT
Although the abstract is the first writing component of a scientific manuscript, it should actually be written last. You should check the format of the journal that you are submitting to, but a structured abstract generally has four sections: Background/Purpose, Methods, Results, and Conclusions. After the rest of the paper has been completed, you can select a sentence or two from the appropriate areas (Introduction, Methods, Results, and Discussion) to include in the aforementioned abstract sections. The abstract, however, is not the introduction and should not be cut and pasted from the introduction. An abstract that is identical to the introduction in any paper or grant will be summarily rejected because of the perceived lack of effort to construct a thoughtful synthesis of the materials in the paper. You should be concise and state the purpose usually in one sentence. The emphasis should be placed on the methods and results, which should each be written in three to four sentences. The conclusion can typically be written in two sentences—the first summarizes your findings and the second makes a conclusion.12 Only data contained in the paper should be included in the abstract—it should not contain any new data. Care should be taken in writing the abstract. Many physicians will not read past the abstract if it is not well-written. Similarly, many readers only need to scan through an abstract to determine whether an article is pertinent to their topic of interest.
WRITING THE INTRODUCTION
Before any writing begins, it is important to perform a thorough literature search. You should be familiar with all the recent advances in your field of study as well as important historical references. “Incomplete, inaccurate, or outdated review of the literature” is one of the common reasons for manuscript rejection.13–15
The Introduction should be written in the present tense.16 You can begin in the first paragraph by mentioning the most important references and stating the research problem. The sec-ond paragraph can elaborate on the magnitude of the problem and unresolved issues. The final paragraph describes the rationale for your study and should end with the purpose: a hypothesis of what you are expecting to find.12 Examples of elements to include in a well-written manuscript are shown in Table 2. Of course, before you begin writing or even analyzing your data, you should have developed a hypothesis. Your observed results may not match your expected results when you analyze your data, but we will discuss that in the Discussion section.
WRITING THE MATERIALS AND METHODS
The traditional sequence of a manuscript is Introduction, Methods, Results, and Discussion, which has been referred to by the acronym IMRAD. Pollock et al. have suggested drafting the manuscript in the sequence MRDAI.17 The Methods section should be written in enough detail that another researcher would be able to duplicate your study.12 In fact, the Methods section is most often responsible for outright rejection of a manuscript18 because the lack of detail is a common problem.19 This is the section where you want to state everything you did (controlling for biases, validating research tools) to increase the reliability of your results.
The Methods section should be written in the past tense.16 To avoid any confusion by the readers, you may want to start by describing the type of study you performed: randomized, prospective, retrospective, case-control, and so on. If it is a study involving human subjects, you should state that you obtained institutional review board approval and include the subject inclusion and exclusion criteria. Include how and from where subjects were recruited, as well as randomization and blinding procedures. Including patient characteristics, such as disease stage or severity and comorbid conditions, can help the readers to determine whether the findings of the article are applicable to their patients of interest.16 It may be helpful to include a diagram of the number of subjects recruited, how many were excluded, losses to follow-up or withdrawal, and your final sample size. Figure 1 shows an example of a similar diagram from a meta-analysis. Details of the sample are included in the Results, but most journals will require that the previously stated information is included in the Methods. Likewise, preliminary experiments or pilot studies can be included in the Methods section if they helped you to arrive at the methods used in your study.2 If equipment was used, the equipment manufacturer, model, and calibration methods should be included. It may also be helpful to include a timeline that shows how and when different aspects of the study protocol took place.12 The final paragraph of the Methods section should describe the statistical analysis. You may need assistance writing this paragraph from your statistician if you did not perform the analysis yourself. You should include the tests used, the p value that determined statistical significance, and whether an a priori power analysis was conducted to decide your necessary sample size. The power and sample size calculation is often a neglected area of scientific presentation and should always be performed before conducting a study.
WRITING THE RESULTS
In writing the Results section, it is important to “only state the facts!” The Results section is not a place to include citations or your interpretation of the data. “Make your point with data, not arguments.”20 The Results section should be written in the past tense.12 You should begin by describing the study sample demographic data, which can be done in a simple table. The data reported in a table should not be repeated in the text. Furthermore, tables and figures should stand on their own. You should include a title, legend, and labels for the axes. The readers should be able to determine what the table is about by only looking at it and not having to read any text. Any percentages should include raw numbers so that the readers are not misled by a large percentage (such as 25 percent) that only came from one of four subjects.
It should be noted that the word “significance” should only be used to describe statistical significance. You should avoid using significant as a synonym for importance. General phrases, such as “showed a trend toward” when results are not statistically significant, often tend to signal a poor study design and should be avoided.2 Also, the word “data” is plural. Thus, it is correct to state, “Our data are ” rather than “Our data is.”
WRITING THE DISCUSSION
You should begin writing the Discussion by discussing the major findings and relating them to your hypothesis. Did you reject your null hypothesis with your findings or were you unable to reject it? Although being unable to reject your null hypothesis (for example, seeing no significant change between your treatment groups) may seem like a “crash and burn” situation, you can still publish your paper. It has been stated that “[m]anuscripts describing studies with negative findings are especially tough to get accepted by medical journals, with publication rates generally less than one-third that of manuscripts describing studies with positive findings.”21 You should describe whether you conducted an a priori power analysis to determine your necessary sample size to avoid a beta error (concluding that there is no difference between treatment groups when there is in fact a difference). Chung et al. found that 82 percent of negative studies in a major hand surgery journal had inadequate statistical power (probability to detect a true difference if a difference is present) to support their conclusions.22
Previous studies have shown that physicians, particularly those with no formal education in epidemiology and biostatistics, have a limited ability to interpret study results.23–25 More than 58 percent of medical residents, however, use statistical information in the published literature in forming opinions or when making medical decisions.26 Thus, it is important in the Discussion to spell out the meaning of your statistical findings without appearing condescending to your readers. When describing your study, past tense should be used, but present tense should be used for established knowledge from other investigators.16
In addition to the meaning of your data, you want to discuss the importance. You should compare and contrast your results to those found in the literature. Care should be taken not to criticize other published work. Similar findings will strengthen your results, but you should still point out how your study differs from previous similar studies. When explaining the study results, it is important to consider all possible explanations rather than just those that fit your preconceived biases. You should, however, avoid coming to conclusions that are not supported by your data. Some of the most common criticisms of the Discussion section by editors' and reviewers' involve “coming to erroneous or unsupported conclusions, drawing conclusions disproportionate to the results, uncritically accepting statistical results, and interpreting the findings in a manner not concordant with data reported.”18 Unless you are absolutely sure that it is true, you should avoid statements such as, “This is the first study to demonstrate .” After all, there are few studies that change the course of scientific progress, but some authors are overly enthusiastic in advertising their study as a seminal contribution.19 It is also important to discuss the clinical relevance of your findings and how patients or physicians may benefit from them.27
In the Discussion section, it is essential to address the limitations and strengths of your study. We have stated them in this order, rather than as strengths and limitations, because it is nice to leave the reader on a good note when he or she finishes reading your paper. Thus, acknowledge your study's limitations first. One limitation may include a lack of generalizability, which often happens when trying to obtain a homogeneous sample. You may use your study's limitations to make suggestions for future research. If not, you should still include suggestions for further research, usually in your concluding section. This last paragraph or last few sentences can also be used to propose ideas for changes to medical practice.27
REMEMBERING STYLE AND GRACE
Throughout the writing and editing process, it is important to remember the style and grace that is needed in writing any good manuscript. First of all, do not make the readers and reviewers work unnecessarily! When reading through your paper, try to keep the reviewers in mind. When a question is left unanswered or is not answered until the Discussion, the readers can get frustrated and give up on your paper entirely. Make it easy on them by explaining early on (usually in the Methods) why you did what you did—whether it was by choice or not. On a similar note, consistently use the same word to describe the same thing to provide continuity and avoid confusion.28 Of course, it is also helpful to have someone else read your manuscript—a colleague and maybe even a layperson. A different set of eyes and a different perspective can point out areas that need clarity.
A few grammatical considerations will assist your readers. Use transitions for flow. “Transitions let the reader know how each sentence relates to the story and how parts of sentences are related.”28 You should also limit the use of passive language. Although scientists tend to use it because they think that it is objective and do not want to indicate who is conducting a certain action,28 it becomes quite boring to continually read, “The sutures were removed” and “The data were analyzed.”
Other grammatical points include writing with precision. Instead of saying that you observed a change, say that you measured an increase or a decrease.28 Along this point, you should avoid wordiness and using long or unfamiliar words when a commonly used shorter one will convey the same message.28 Words such as marked, revealed, and demonstrated are overused and have lost their intended meaning.29 Although it is common for writers, especially inexperienced ones, to try to use “flowery” language, it is best to remember that “less is more.” You can delete unnecessary adjectives and adverbs, such as fundamentally, very, and great. You should also examine your writing for wordy phrases such as “it is often the case that”28 or “in our opinion.” If it is not your opinion, it should have a citation attached to it. The words “while” and “since” have primary temporal definitions, but “while” is often used as a synonym for “although” or “whereas” and “since” is commonly used instead of “because.”28 Be mindful of these subtleties that can slow down your readers and cause them to misinterpret what you are trying to say.
Well-written paragraphs usually start with making a point in the first sentence and then developing that point throughout the paragraph. Focusing on a single major point in each paragraph allows the readers to follow the author's train of thought.28
Although the focus of this article has been how to write a scientific manuscript, we have decided to include a section on grant writing. After you have successfully written and published a few papers, you may hope to get funding to conduct your own research and compete in a different arena. An in-depth article on grant writing was written by Chung and Shauver,30 but we will touch on the basics. In manuscript writing, you have already conducted the research, whereas in grant writing, you have probably conducted preliminary research but want to obtain funding to study questions by expanding on that preliminary research. Of course, style and grace are also important in grant writing. The general formats of manuscripts and grants are similar. In manuscripts, you have an Introduction, Methods, Results, and Discussion. Likewise, in a National Institutes of Health–formatted grant proposal, you have Specific Aims and Research Strategy (incorporating the Significance, Innovation, Approach, and Preliminary Studies/Progress Report).31 The Introduction corresponds to the Specific Aims, where the goals and hypotheses are presented. The research question is the most important section of the grant.32,33 The impact (one of the new National Institutes of Health review criteria) of your proposed research on the research field should be spelled out for the reviewers in the Specific Aims. The Methods section corresponds to the Approach section. New National Institutes of Health page limitations have reduced the overall size of the grant proposal, encouraging researchers to be less detail-oriented in describing their methods. The Results section corresponds to Preliminary Studies/Progress Report. Although a grant does not have a Discussion section per se, your “so what?” question is extremely important to answer to the significance of your proposal. The Innovation section should describe whether your proposal introduces novel concepts, approaches, or methodologies.34
New National Institutes of Health review criteria consist of an overall impact score that reflects reviewers' “assessment of the likelihood for the project to exert a sustained, powerful influence on the research field(s) involved.”34 The five core review criteria remain the same as before but have been prioritized as Significance, Investigators, Innovation, Approach, and Environment.34 Guidelines published by the National Institutes of Health regarding how to prepare for grant writing are similar to those for manuscript writing, including finding a mentor, preparing an outline, and soliciting feedback from colleagues.35 As described previously for manuscripts, it is best to make the grant reviewers' lives easy. One way this can be done is by separating each of the review criteria into identifiable sections. You can also get a feel for the questions reviewers might ask, depending on their research interests, by identifying the review committee that your proposal will be directed to and seeing who on that committee will most likely review your proposal. Although you should search the literature to obtain background and see if similar studies have been conducted, you should also search the National Institutes of Health Research Portfolio Online Reporting Tools (RePORT) to see if similar studies have been funded.36
There are a few other general reminders to keep in mind. Follow the journal guidelines as far as formatting, line numbering, word limits, figures, and citations. This will save you the time and hassle of having the editor send the paper back to you before it can be sent out for review. Aesthetics is another item to keep in mind. Before submitting (especially when submitting online), preview your paper to be sure that you do not have a heading at the end of a page without text underneath it. If you have large blocks of text, it might be a good idea to add one or more subheadings. Make sure that your font is the same throughout and that you have not overused italics. Some reviewers do not like to see italics because the overuse of italics may signal to the reviewers that the authors consider the materials presented to be too complex for the reviewers to understand. Spell out abbreviations the first time you use them but do not try to avoid word count limitations by bombarding the readers with multiple abbreviations. Nonuniversal abbreviations force the readers to remember their meanings and substitute the full phrase each time they appear.29 An abbreviation should be used often enough in the paper, preferably more than 10 times, so that the readers do not forget the meaning.37
Of course, proofread, proofread, proofread, and use spell check! As previously mentioned, allow someone else to proofread, too. You should also look through your final manuscript and notice your citations. Are you continually citing the same author? If so, this can show bias. Make sure that you are citing a wide range of authors and not reiterating someone else's ideas.
If you have conducted a randomized, controlled clinical trial, you should follow the Consolidated Standards of Reporting Trials (CONSORT) statement.38 This statement requires authors of a randomized, controlled clinical trial to fill out a checklist of the key information that should be included.16 Figure 2 has a general manuscript checklist of items to include.
Although you should be optimistic, you should also remember that there is a chance that your manuscript will be rejected by the journal you initially submit to. The acceptance rate of clinical research–based manuscripts submitted to major biomedical journals is 30 to 40 percent.39 The acceptance rate for journals such as the New England Journal of Medicine and JAMA is less than 10 percent.16 Even if your manuscript is not accepted, the comments by the reviewers will most likely help you to identify your weaknesses and improve your paper. If your manuscript is accepted with revisions (major or minor), then congratulations! You may, however, still have plenty of work to do. In revising, draft a response letter with each reviewer's comments typed out and explain how you have addressed them. Also state where in the paper the revision can be found (i.e., Methods section, paragraph 2, line 2). Remember to be polite; if you are choosing not to make a suggested revision, you should give a reason for not doing so. Although it is not necessary to do every suggested edit, your manuscript will not be looked upon kindly if you choose to make a rebuttal for all or the majority of suggestions.
Although scientific writing can be a long and tedious process, your writing ability will continuously improve. Remember, “Only the researcher who is competent in the art of written communication can play an active and effective role in contributing to science.”40
This work was supported in part by a Midcareer Investigator Award in Patient-Oriented Research (K24 AR053120) from the National Institute of Arthritis and Musculoskeletal and Skin Diseases (to Dr. Kevin C. Chung).
1. Schein M, Farndon JR, Fingerhut A. Why should a surgeon publish? Br J Surg.
2. Eastwood S, Derish PA, Berger MS. Biomedical publication for neurosurgery residents: A program and guide. Neurosurgery
3. Miedzinski L, Davis P, Al-Shurafa H, Morrison J. A Canadian faculty of medicine and dentistry's survey of career development needs. Med Educ.
4. Pololi L, Knight S, Dunn K. Facilitating scholarly writing in academic medicine. J Gen Intern Med.
5. Neacy K, Stern S, Kim H, Dronen S. Resident perception of academic skills training and impact on academic career choice. Acad Emerg Med.
6. Broaddus VC, Feigal DW Jr. Starting an academic career: A survey of junior academic pulmonary physicians. Chest
7. Anonymous. In pursuit of comprehension. Nature.
8. Anonymous. How experts communicate. Nature Neurosci.
9. O'Donnell M. Evidence-based illiteracy: Time to rescue “the literature.” Lancet.
10. Tobin MJ. Compliance (COMmunicate PLease wIth Less Abbreviations, Noun Clusters, and Exclusiveness). Am J Respir Crit Care Med.
11. Knight J. Scientific literacy: Clear as mud. Nature.
12. Branson RD. Anatomy of a research paper. Respir Care.
13. Bordage G. Reasons reviewers reject and accept manuscripts: The strengths and weaknesses in medical education reports. Acad Med.
14. Samet JM. Dear author: Advice from a retiring editor. Am J Epidemiol.
15. DeBehnke DJ, Kline JA, Shih RD. Research fundamentals: Choosing an appropriate journal, manuscript preparation, and interactions with editors. Acad Emerg Med.
16. Pakes GE. Writing manuscripts describing clinical trials: A guide for pharmacotherapeutic researchers. Ann Pharmacother.
17. Pollock A, Evans M, Wiggin M, Balch C. Writing your first scientific paper. In: Troidl HE, ed. Principles and Practice of Research: Strategies for Surgical Investigators.
New York: Springer-Verlag; 1991 :391–403.
18. Byrne D. Publishing Your Medical Research Paper: What They Don't Teach in Medical School.
Baltimore: Lippincott Williams & Wilkins; 1998.
19. Pierson DJ. The top 10 reasons why manuscripts are not accepted for publication. Respir Care.
20. Alexandrov AV. How to write a research paper. Cerebrovasc Dis.
21. Dickersin K, Min YI. Publication bias: The problem that won't go away. Ann N Y Acad Sci.
22. Chung KC, Kalliainen LK, Hayward RA. Type II (beta) errors in the hand literature: The importance of power. J Hand Surg.
23. Berwick DM, Fineberg HV, Weinstein MC. When doctors meet numbers. Am J Med.
24. Wulff HR, Andersen B, Brandenhoff P, Guttler F. What do doctors know about statistics? Stat Med.
25. Weiss ST, Samet JM. An assessment of physician knowledge of epidemiology and biostatistics. J Med Educ.
26. Windish DM, Huot SJ, Green ML. Medicine residents' understanding of the biostatistics and results in the medical literature. JAMA.
27. Hess DR. How to write an effective discussion. Respir Care.
28. Derish P, Eastwood S. A clarity clinic for surgical writing. J Surg Res.
29. DeBakey L, DeBakey S. The case report: II. Style and form. Int J Cardiol.
30. Chung KC, Shauver MJ. Fundamental principles of writing a successful grant proposal. J Hand Surg.
32. Inouye SK, Fiellin DA. An evidence-based guide to writing grant proposals for clinical research. Ann Intern Med.
33. Falk GW. Turning an idea into a grant. Gastrointest Endosc.
37. Zeiger M. Essentials of Writing Biomedical Research Papers.
2nd ed. New York: McGraw-Hill; 1999.
38. Begg C, Cho M, Eastwood S, et al. Improving the quality of reporting of randomized controlled trials. The CONSORT statement. JAMA.
39. St. James J, Spiro H. Writing and Speaking for Excellence: A Guide for Physicians.
Boston: Jones and Bartlett Publishers; 1996.
40. Ebel H, Bliefert C, Russey W. The Art of Scientific Writing: From Student Reports to Professional Publications in Chemistry and Related Fields.
Weinheim, Germany: VCH; 1987.
PRS Mission Statement
The goal of Plastic and Reconstructive Surgery® is to inform readers about significant developments in all areas related to reconstructive and cosmetic surgery. Significant papers on any aspect of plastic surgery—original clinical or laboratory research, operative procedures, comprehensive reviews, cosmetic surgery—as well as selected ideas and innovations, letters, case reports, and announcements of educational courses, meetings, and symposia are invited for publication.