Keywords

These keywords were added by machine and not by the authors. This process is experimental and the keywords may be updated as the learning algorithm improves.

In questions of science, the authority of a thousand is not worth the humble reasoning of a single individual.

Italian physicist and philosopher Galileo Galilei (1564–1642)

Galileo, a leader in the Scientific Revolution that followed the hyper-religiosity and doctrinaire thinking of the Middle Ages, was a champion of reasoning. Ideally, of course, reasoning is based on observed facts—data—and medicine science’s best format for data presentation and analysis today is the report of a clinical study.

In Epidemiology is the report of a study that examined the risk of fatal myocardial infarction in persons exposed to aircraft noise. (Huss et al. Epidemiology. 2010;21:829). Following analysis of 4.6 million individuals over 5 years, the authors concluded that there is, indeed, a dose-related association between aircraft noise and death from myocardial infarction, independent of possible confounding factors such as particulate matter air pollution and socioeconomic status of the municipality.

According to a report published in Pediatrics, the use of probiotics, specifically Lactobacillus reuteri, can reduce daily crying episodes in colicky babies, a conclusion based on a randomized, double-blind, placebo-controlled trial involving 50 infants (Savino et al. Pediatrics. 2011;126;e526).

Are there actually more fatal medication errors in July, when new medical residents arrive on hospital floors, or is this simply an “urban legend?” Actually there is some truth to the belief, according to a study of 62 million death certificates recorded over 27 years, with a July mortality spike that Phillips and Barker conclude “results at least partly from changes associated with the arrival of new medical residents” (Phillips et al. J Gen Intern Med. 2010;25:774).

A report in The Lancet describing a 20 year follow-up of five randomized trials to assess the long-term effect of aspirin on colorectal cancer incidence and mortality presents this conclusion: “Aspirin taken for several years at doses of at least 75 mg daily reduced long-term incidence and mortality due to colorectal cancer” (Rothwell et al. The Lancet. 2010;376(9754):1741).

In the Journal of the American Medical Association (JAMA) is a report of a study intended “to determine whether integrating smoking cessation treatment into mental health care for veterans with posttraumatic stress disorder (PTSD) improves long-term abstinence rates.” A randomized controlled trial involving 943 smokers with military-related PTSD revealed that the answer to the research question is: Yes (McFall et al. JAMA. 2010;304:2485).

Preparing the report of original research is arguably the most challenging undertaking in medical writing. It is not necessarily that writing the report is so complicated, because the model is prescribed. In a sense, you need only fill in the blanks. The challenging part is that one needs to have completed a research study and have the resulting data available. What follows assumes that you have completed that task, and are composing your report for publication.

Because there is a prescribed model, writing the report of original research has one advantage over other types of medical writing. You will not need to dream up a concept and structure for the article. In previous chapters I discussed diverse ways to approach the review article, editorial, book chapter, and other models of medical publication. For the report of original research, there is only one model, called IMRAD. This acronym stands for the major parts of a research report: Introduction, Methods, Results, and Discussion. The IMRAD format is what editors are accustomed to reviewing. It is what clinicians and scientists are used to reading. Deviating from this format risks summary rejection. I describe the IMRAD format below.

The rigid format for the report of original research highlights the fact that research reports are written to be published and cited, more than they are written to be read. They are intended to be repositories of scientific data rather than literary gems. They just happen to be in prose. Day [1] has summarized this viewpoint very well: “Some of my old-fashioned colleagues think that scientific papers should be literature, that the style and flair of an author should be clearly evident, and that variations in style encourage the interest of the reader. I disagree. I think scientists should indeed be interested in reading literature, and perhaps even in writing literature, but the communication of research results is a more prosaic procedure.”

With that said, I still plead with authors, even those composing research papers, to construct paragraphs thoughtfully, avoid long and convoluted sentences, select words carefully, avoid the use of jargon, and express their ideas as clearly as possible.

Thinking About the Research Report

For the clinical investigator, getting research results published can be the difference between professional success and failure in the academic medicine arena. It can determine whether or not one gets the big grant, or if one receives tenure. Entire academic careers have been built on a single groundbreaking research study, carefully reported in a prestigious journal. Whether you are a patient care physician or a research-track academic faculty member, whether you have done a practice-based research project or a randomized clinical trial, remember that your research is not completed until the results are reported in print.

The five papers I cited above all had research questions that provoked my interest, and perhaps yours. Although I had never thought much about the issue before I read the title, I wanted to know if aircraft noise might be associated with death from myocardial infarction. I was intrigued by the well-designed study of the use of probiotics in colicky babies. The finding of a July morality spike confirms one of my long-held, but previously evidence-deficient, beliefs regarding the risks of unseasoned residents caring for unsuspecting patients. We clinicians need to know if it is worthwhile to advise the use of low dose aspirin for the primary prevention of colorectal cancer in our patients. And as one who teaches residents seeing smokers with mental health issues, I was intrigued by the possibility of combining therapy for these problems. For this reason, I briefly summarized the study outcomes, even though this is a book about writing and not clinical science.

I mention the above—about my interest in the research questions that prompted the studies—because you may be tempted to stretch the definition of research too far. As a surgeon, you may consider reporting the findings of your last 200 cases of lumbar laminectomy or laparoscopic cholecystectomy. If you are an internist, you may believe that your colleagues are keenly interested in how you treated 100 consecutive patients with congestive heart failure. Such studies do not set out to answer a clinical question and generally do not have anything important to say. They may qualify as quality improvement efforts, but are not likely to result in a publishable research paper.

A research report describes your research, whether it involves humans, rats or a meta-analysis of previously published studies. In general: You generate a research question, and then collect data to answer the question. “Mining” tons of data to find something, anything that has statistical significance is not good research, and a paper describing such a method will be evident to the informed reviewer.

When planning a research report, be aware that competition for publication space in leading refereed journals is intense, and research papers are typically rewritten several times before final acceptance. Ultimately, your clinical research paper will be judged by its impact on your specialty and on the greater body of medical knowledge, as evidenced by its citation in other research articles, review papers, and textbooks.

Just before launching into a discussion of the how to write a report of hypothesis-based research based on the quantitative scientific model, I want to acknowledge another branch of research that also is reported in the literature—qualitative research. This type of research, which medical science has borrowed from our colleagues in sociology and anthropology does not generate piles of data. Instead, we find terms such as focus groups, studying stories, and mixed methods research [2]. The IMRAD model, describe below, does not readily lend itself to reports of qualitative research, which tend to use more innovative styles, prompted by the nature of the results being reported. Perhaps that will be a good chapter for the next edition of this book. For now, we will examine how to report the results of traditional quantitative research.

The Expanded IMRAD Model

The IMRAD model of research reports has evolved over generations of scientific publications. It has at its core four elements:

  • Introduction: Why is the topic important, what prior research has been done, and what question did you set out to answer?

  • Methods: Who were your subjects and what did you do to them? How did you analyze the data?

  • Results: What did you find out?

  • Discussion: What do your findings mean?

These four items are the foundation of IMRAD. Research papers, however, have more than just the four main components, and I am going to present an expanded IMRAD model. Keep in mind the four key elements as we explore the IMRAD and more, beginning with selection of the title for the report.

Title

The title is the “label” for the paper. The title must tell, more or less, what was studied. An early question the writer must answer is this: Should I reveal my conclusion in the title?

One of the studies cited at the beginning of the chapter is titled “Long-term effect of aspirin on colorectal cancer incidence and mortality: 20-year follow-up of five randomized trials.” If I read only the title and had no other background knowledge, I might misunderstand the authors’ intent and take away the message that aspirin actually increases the risk of colorectal cancer incidence and mortality. In fact, the study showed just the opposite. Therefore, I believe that a better title would be “Long-term use of aspirin reduces colorectal cancer incidence and mortality: 20-year follow-up of five randomized trials.” Might the title of the JAMA article about tobacco cessation have been more helpfully titled: “Integrating tobacco cessation into mental health care for posttraumatic stress disorder: a randomized controlled trial showing greater prolonged abstinence?”

Whimster [3] writes: “I believe that readers need a verb in the title, such as a newspaper headline usually has, and that to be meaningful it should convey the message, as in: ‘Rickettsial endocarditis is not a rare complication of congenital heart disease in dental practice: a report of five cases.’”

Sometimes we authors dream up witty titles. Consider the following title: “It’s b-a-a-a-a-a-a-ck again, or how to live with the new APA manual: reprise for Edition 6” (Baggs et al. Res Nurs Health. 2009;32:477). Generally, editors counsel against using clever phrases in titles, and rightfully so.

Note how often colons show up in article titles. They allow progression from the general to the specific, all in an integrated title phrase. See the papers cited above. One example, cited above as one of the chapter’s five index papers, is the title “A July spike in fatal medication errors: a possible effect of new medical residents.” The authors discuss the general problem and then the data sources. The reader has a better idea of the article’s contents than if only the first phrase were listed.

On a technical basis, the instructions for authors may prescribe a word or character limit for the title. Also, I believe that titles should not contain acronyms or abbreviations, no matter how widespread the author and editor consider their use.

Authors

The chief issues in authorship of a research report are generally twofold: (1) Who is an author? (2) How shall the authors be listed?

As discussed in Chap. 5, everyone who contributed substantially to a research project and preparation of the report should be listed as author. Furthermore, each author should have participated sufficiently in the work to take public responsibility for appropriate portions of the content [4].

What about adding author names of those who have contributed very little? Strasburger [5] described the problem: “Fiction is written by one individual; medical articles may be written by committee. There is no such thing as ‘author inflation’ in fiction, simply because there is no need for it. Medical writers must publish or perish, academically. Fiction writers must publish or perish, existentially.” Despite the need to avoid perishing, it is inappropriate to have your name listed if you have not met the criteria listed in the previous paragraph. You must not become an “author inflation” perpetrator.

No department chair or research director should insist on being named as an author unless there has been a significant contribution to the study and to writing the paper. Authorship listing by administrative fiat is academic malpractice. Adding the name of a prestigious senior faculty member as the final entry on a long author list might help get the paper a better review, but including the well-known name implies that person’s active participation in the project. Gratuitous addition of an author name is ethically inappropriate.

The order in which authors are listed on a research report should be decided very early in the process, generally during one of the first meetings of the research group planning the study. Changes in the rank order can be made later if contributions of individuals to the project do not turn out to be what was originally planned.

The first author should logically be the one who has done most of the work on the study being reported. Generally this is the person who led the research team and who has created the early drafts of the paper. From then on, authors should be listed according to how much they contributed to the study and the report. As one whose last name begins with a letter toward the end of the alphabet, I have never considered alphabetical listings of names to be fair to the Taylors, Washingtons, and Zells of the world.

A quirk of citation listing holds that when the paper is used as a reference in other studies, if your paper has seven or more authors, only the first three are named and the rest will join the et al. army of obscurity.

Abstract

The abstract is an author-generated synopsis of the paper. Many believe that the final version of the abstract should be the last item written, since only then will you know exactly what is in the paper that you are summarizing. When writing an abstract, select each word as if your paper’s being read depended on it, and jettison verbal clutter ruthlessly.

In general I have always taught that the abstract should mirror the IMRAD structure of the paper. That is, the paper’s introduction, methods, results, and discussion (conclusions) should each be presented in a sentence or two, and many good abstracts have exactly four short paragraphs. According to the Uniform Requirements, “The abstract should provide the context or background for the study and should state the study’s purpose, basic procedures (selection of study subjects or laboratory animals, observational and analytical methods), main findings (giving specific effect sizes and their statistical significance, if possible), principal conclusions, and funding sources. It should emphasize new and important aspects of the study or observations.” [4]

The current trend is for journals to require structured abstracts [6]. This means that information in the abstract is presented according to specific headings that differ a little with each journal. All structured abstracts will include the four key components of the IMRAD model, although synonyms for these headings may be used, including some variations such as “Context,” and “Main Outcome Measures.” Some journals prefer abstracts with full sentences; others encourage the use of phrases. Here, from the Archives of Surgery, is what we might find in a well-written structured abstract for a paper titled: “Risk Factors for Lymphedema in a Prospective Breast Cancer Survivorship Study” (Kwan et al. Arch Surg. 2010;145:1055 ):

  • Objective: To determine the incidence of breast cancer-related lymphedema (BCRL) during the early survivorship period as well as demographic, lifestyle, and clinical factors associated with BCRL development.

  • Design: The Pathways Study, a prospective cohort study of breast cancer survivors with a mean follow-up time of 20.9 months.

  • Setting: Kaiser Permanente Northern California medical care program.

  • Participants: We studied 997 women diagnosed from January 9, 2006 through October 15, 2007, with primary invasive breast cancer and who were at least 21 years of age at diagnosis, had no history of any cancer, and spoke English, Spanish, Cantonese, or Mandarin.

  • Main Outcome Measure: Clinical indication for BCRL as determined from outpatient or hospitalization diagnostic codes, outpatient procedural codes, and durable medical equipment orders.

Note that so far, only one section—Participants—contains a complete sentence. Later in the abstract, under Results and Conclusions, paragraphs become longer and the style changes from phrases to complete sentences.

The instructions to authors for the Journal of the American Medical Association state, “Reports of original data should include an abstract of no more than 300 words using the ­following headings: Context, Objective, Design, Setting, Patients (or Participants), Interventions (include only if there are any), Main Outcome Measure(s), Results, and Conclusions. For brevity, parts of the abstract may be written as phrases rather than complete sentences.” [7] The JAMA Instruction for Authors goes on to tell what should be presented in each section of the abstract.

The tight word limitation and the many topics to be covered serve to get the important data into tightly written abstracts, but at the expense of some very complicated, number-laden, and almost incomprehensible sentences in the Results section.

On a technical basis, descriptions of work that has been done (Methods and Results) should be written in the past tense. An explanation of what you think (Discussion) is written in present tense, often with a phrase such as, “We conclude. . . .” In the spirit of intellectual honesty, the abstract must never contain a conclusion that is not supported by what is in the body of the paper.

Clinical Trial Registration

For all clinical trials, the name of the trial registry, registration number, and URL of the registry must be included [4, 7].

Key Words

In some instances you will be asked to identify Key Words. Key words can be what keep your report from being lost in the information jungle. They are part of the retrievability process that can contribute to the number of times your paper will be cited. In the instructions to authors, many journals request that you submit three to ten key words or short phrases. These will “assist indexers in cross-indexing the article and may be published with the abstract. Terms from the Medical Subject Headings (MeSH) list of Index Medicus should be used; if suitable MeSH terms are not available for recently introduced terms, present terms may be used” [4] (see Chap. 1 and Appendix 1).

Introduction

Finally we arrive at the “I” in IMRAD. The introduction should identify the problem you set out to solve. In a sense, it describes the context of the study. In general terms, the introduction should cover three areas:

  • Problem statement: What is the general nature of the problem that merits valuable journal space and the reader’s attention?

  • Background and work to date: What are the most pertinent, timely published studies that relate to the problem?

  • The research question: What is the specific, focused question that you set out to answer? If you have a formally stated hypothesis, here is where it should be presented.

The Problem

The Introduction classically opens with a broadly stated and virtually unassailable generalization about the problem. The Introduction to the paper on aspirin and colorectal cancer begins, “Colorectal cancer is the second most common cancer in developed countries. . . ”

Good Introductions are often written as a three-level “inverted pyramid.” The broadest statement comes first. Here is an example: Primary care clinicians encounter many patients with headaches. Next comes a more narrows statement, e.g., The clinician treating a headache patient is always aware that, in a few individuals, the cephalgia might be the tip-off to a life-threatening disease. This might be followed by an even more specific statement setting out the purpose of the study: This study examines patient records to identify symptoms and signs that might identify persons most at risk for life-threatening causes of head pain who should be referred for diagnostic imaging.

Background

Describe the key work that has been done on the topic to date. Do not present an exhaustive literature review dating back to the Renaissance. Be very selective and include only articles that have a direct bearing on your research question.

Research Question

State clearly the question you are trying to answer. One focused question is usually better than many. The question may be stated as a query or perhaps as a hypothesis, but often is phrased as a statement of intent: In the study of aircraft noise and fatal heart attacks, the research question is stated: “We examined residential exposure to aircraft noise and mortality from myocardial infarction, taking air pollution into account.”

To inform the reader as to what your study is all about, it is vital that you articulate the research question clearly in the introduction. I wish that writers of research reports would all do so and would frame their research questions as direct queries or even as hypotheses. For example, the authors of the paper on integrating tobacco cessation into mental health care for posttraumatic stress disorder state an explicit hypothesis: “Our multisite randomized controlled­ effectiveness trial hypothesized that integrating smoking cessation treatment into mental health care would improve long-term smoking abstinence rates in veterans with PSTD compared with referral for specialized cessation treatment.” However, most authors are less explicit, and I have learned to be content with somewhat vaguely stated research questions stated as: “The purpose of this study was to investigate both early and late dumping syndromes after gastrectomy for gastric cancer in 2 high-volume centers in Japan” (Mine et al. J Am Coll Surg. 2010;211:628).

Technical Issues in Writing the Introduction

When writing your introduction, use the present tense when describing the general nature of the problem and the background work. Then the research question, if presented as a statement, is usually in past tense, as in the examples above.

The uniform requirements advise, “Do not include data or conclusions from the work being reported.” [4] Not everyone agrees with this stance. Both Day [1] and Whimster [3] advocate stating the conclusions early in the article, and not keeping­ the reader in suspense, as you would with a whodunit mystery novel. The best spot for the important implication for translation of your findings to clinical practice may be in the introduction. In this area where controversy exists, use your best judgment, based on the data you are presenting.

Methods

The Methods section, sometimes called Participants and Methods or perhaps Methods and Materials, should describe a logical experimental approach. Because this section presents a number of topics, subheadings are often used. In the Methods section of the article mentioned above about probiotics and infant colic, the authors used the following headings: Subjects, Study Objectives and Outcomes, Study Design and Sample Collection, Analysis of Bacterial Groups by FISH, Analysis of Fecal L. reuteri DSM 17 938, Analysis of Fecal Ammonia, and Statistical Analysis. The Methods section of the study on aircraft noise and fatal MI has five headings: Study Population, Outcomes, Exposure to Aircraft Noise, Exposure to Air Pollution, and Statistical Analysis.

Fundamentally, this section needs to describe the subjects, what you did to them, and what statistical methods you used. After writing the first draft of the Methods section, ask yourself whether what you are presenting allows reproducibility. That is, could a trained investigator in your field replicate your study, given the information you have provided?

Methods should not include numerical data, which should be presented in the Results section.

Subjects

Describe the subjects studied, including age, gender, and other important characteristics that may be pertinent to the study. Uniform Requirements recommends that when authors use such variables as race or ethnicity, they “should define how they measured these variables and justify their relevance.” [4]

State also whether any potential subjects were excluded and why they were excluded, if there is a meaningful reason. For example, in the study of probiotics and infant colic, subjects with chronic illness or gastrointestinal disorders were excluded.

Method

Here you describe what was actually done to the subjects. Also, if appropriate, describe any data collection tools, such as survey instruments. If apparatus was used, identify the item and manufacturer. Be sure to identify all drugs by generic name; adding the trade name is optional, but is useful for the practicing clinician. Be sure to include medication doses and routes of administration.

Statistics

Describe the statistical methods used, “with enough detail to enable a knowledgeable reader with access to the original data to verify the reported results.” [4] This generally means identifying specific tests used. In the study of aircraft noise and fatal MI, the description of the statistics used begins, “We analyzed the association between aircraft noise and cardiovascular mortality using Cox proportional hazard models, with age as the underlying timescale. . . ”

In an effort to make this book more useful for those involved in research studies, I have added a lexicon explaining Commonly Encountered Methodologic and Statistical Terms. (See Appendix 5) In the meantime, here are some thoughts, several taken directly from the International Committee of Medical Journal Editors (ICMJE) Instruction for Authors [4], about the use of statistics in presenting reports of clinical research studies:

  • Avoid relying solely on statistical hypotheses testing such as P values, which fail to convey important information about effect size [4].

  • When data allow, present quantifiable findings with appropriate indicators of measurement error or uncertainty, such as confidence intervals [4].

  • When describing the statistics you employed in analyzing data, specify the computer software used [4].

  • Don’t hesitate to seek help with statistics. It is easy to get lost in the unfamiliar forest of statistical analysis. A study by Strasak et al. found, “Five of 31 papers for the New England Journal of Medicine (NEJM) (16.1%) contained usage of wrong or suboptimal statistical tests, either because of incompatibility of test with examined data, inappropriate use of parametric methods, or use of an inappropriate statistical test for the scientific hypothesis under investigation.” [8]

  • Finally, as I was reminded by a reviewer commenting on the first edition of this book, you should never allow a pharmaceutical company or other research sponsor to do the statistical analysis of your data [9].

Table 11.1 lists some publications that may be helpful for the statistically challenged.

Table 11.1. Sources of statistical information for medical writers
Full size table

Stick with the approach that the informed reader should be able to replicate your research methods used and the statistical analysis. Provide all data that were analyzed, and show the outcomes of the statistical methods that you described under Methods.

Results

What did you discover? Describe your findings in a logical sequence and do so fully, yet succinctly. To support my plea for readability in research reports, I like the image created by Alexandrov [10]: “Make data presentation so clear and simple that a tired person riding late on an airplane can take your manuscript and get the message at first reading.” (As a physician, I find this a formidable challenge, given that with the imperfect pressurization of aircraft cabins, there is a measureable decrement blood oxygen saturation, and hence in cerebral oxygenation, at 35,000 ft elevation).

I have sometimes said, only partly in jest, that the ideal Results section has a single sentence, “The results are presented in Table 1,” followed by a single carefully constructed table. In reality, presenting research results is never this simple, but the use of tables and figures can help organize numbers in ways that cannot be accomplished in words. Also, the Results section should begin with some explanatory prose before sending the reader to the first table. Keep in mind that tables and figures are expensive for the journal to produce and are a leading source of error. On balance, however, most Results sections benefit from one or more tables or figures. The paper on the July spike in deaths, for example, has four figures, all bar graphs.

Tables and figures for all types of publication models are discussed in detail in Chap. 4. Here I will only emphasize the importance of creating a legend for each that explains the table so that it can be reasonably understood without the accompanying text. That is, a lecturer can incorporate your table with its legend into a PowerPoint presentation (with credit to you, of course) without adding explanatory prose.

Tables and figures should not duplicate data presented in the text. Select only one location to present the numbers.

Discussion

In Chap. 1, I stated that each article must face the “So what?” question. The Discussion section should answer that question by stating the relationships among facts discovered, relating them to prior studies (the ones you mentioned earlier in the introduction), and postulating what it may all mean—the conclusions. Discuss the results, but do not restate what has already been said under Results. A good way to begin this section is with the phrase: “Our study showed. . . ” [10]

The Discussion section is where you might describe your opinion of the novelty of your findings, or how they may affirm or contradict previous research in the field. For example, in the study of breast cancer-related lymphedema described above, the authors write: “Our results confirm and contrast with findings from 2 recent studies of BCRL.” Here is also where you describe any study limitations, or even any disagreement among co-authors regarding the interpretation of results.

The Holy Grail in all of this is generalizability, a neologism that is not in my Microsoft Word spell checker or Dorland’s Medical Dictionary, but that all researchers recognize. Does what you have found apply only to your group of subjects, a weakness of the small sample or the single-institution study? Or do the results found have implications for similar patients elsewhere, the obvious advantage of the large trial involving thousands of subjects in various sites, such as the aircraft noise vs. fatal myocardial infarction study mentioned at the beginning of the chapter.

State any weaknesses of the study design, or these will surely be described enthusiastically by reviewers or in letters to the editor. The Discussion section is also where you should tell about any factors that may have biased collection of the data, such as unexpected events, attrition of subjects, or ­mid-study changes in methods, such as terminating one of the study groups. The aircraft noise study, for instance, describes the possibility of bias in the coding of deaths.

In the last paragraph (where the grazing reader may go right after reading the abstract), present a summary of your conclusions and what your team thinks they mean. State the implications for others in your field, the generalizability of your work, and perhaps how they might translate to patient care in the office or hospital. Write this paragraph very carefully. It represents the outcome of months of effort.

References

Your references are where you have obtained background information and indicate your awareness of prior work in the area of your research. A focused list of citations is more valuable to your reader—and to you, as author—than a very long list of unselected papers.

References serve other purposes. Readers often use them as part of their own research on topics. For these individuals, your list is already a little out of date by the time it is published, but it can be useful at times. Your reference list also represents a sort of “merit badge” for the authors, indicating that you valued their papers enough to cite them as credible sources.

When using a reference citation to support a statement, be sure that you are conveying the actual meaning of the author. I have seen too many references used to support statements when the paper cited says something entirely different. Today, the ready availability of PubMed and other sites makes it easy to match author assertions and the actual words of authors.

The technical considerations of presenting references are similar for all publication models, and are presented in Chap. 4, along with the most familiar models (Table 4.2). Here I will list just a few additional suggestions and comments:

  • The ideal reference citation is the original research source.

  • Uniform Requirements recommend that you avoid citing abstracts, such as “grey literature” conference abstracts, as references [4].

  • If in doubt in listing the name of a journal, write it out, because, for example, “Psych” could mean psychiatry or psychology.

  • By custom, a journal with a single word title, such as Nature or Science, is written in full and is not abbreviated.

  • A paper accepted for publication, but not yet published, can be cited as “in press” or “forthcoming.” [4] If the paper is published before your article goes to press, the citation can be updated in page proofs to provide the details of publication.

  • Try to avoid using websites as references in a scientific report; these sites contain a lot of specious data and outright fiction.

  • If you must cite a website: Because the site of electronic citations can change or disappear altogether, the author citing a Web site should print out a copy of the online material, in case it is requested later.

  • Never cite a source you have not read and copied for your electronic or paper files.

Acknowledgments

Some papers have a final section listing those who assisted with the work. This includes “all contributors who do not meet the criteria for authorship, such as a person who provided purely technical help, writing assistance, or a department chair who provided only general support.” [4] For example: At the end of his article “How to Write a Research Paper,” [10] Alexandrov states: “The author is not a native English speaker. I am indebted to John Norris, MD, FRCP, for—among many things during fellowship training—his patience with my ‘a’s’ and ‘the’s’ and the first lessons in study design, analysis and presentation.” If financial or material support has not been disclosed elsewhere, it should be included here.

There is one important caveat: Be sure that all the people who you thank are pleased to be acknowledged and that they actually agree with the substance of the paper. Being mentioned allows readers to infer that those acknowledged support the data and conclusions, whether this is true or not. For this reason, you must have written permission from all persons listed in the acknowledgments. Some journals have specific online forms for this purpose; others will accept a signed note on a letterhead.

Common Problems in Reports of Clinical Studies

What are the common mistakes seen in reports of clinical studies? Despite the many hours of labor that go into scientific manuscripts, there are a few errors that occur even with the most experienced medical researchers and writers. Maybe some creep in as unhappy compromises during group wordsmithing. Others may be the result of midnight editing, when not all the mental light bulbs are on. Whatever the reason, we make mistakes in following the recipe for writing research reports. To help you avoid these missteps, I offer the following to act as a checklist to use when you think your manuscript is done.

  • Be sure to prepare a Title Page listing the article title, the names and affiliations of all authors, sources of support such as grants, the number of tables and figures, a word count, and anything else requested in your target journal’s Instructions for Authors.

  • Remember that the title page should specify the “corresponding author”—the author who will represent the research team in conversations with the editor—along with this individual’s full contact information.

  • Check once again to assure that your title accurately describes your study, and that it just might prompt the casual reader to learn more.

  • Ask yourself: Have I stated the problem clearly?

  • Perform a last-minute review of the literature to assure that you have not overlooked a recent key report.

  • Review the Results section of your research report to be sure that it does not contain background information (which should be in the Introduction).

  • Ask yourself: Have I tried to put too much in my tables and figures?

  • Also check to be sure that you have not repeated the same data in tables, figures and text.

  • Verify that interpretation of what you found is in the Discussion section and not in Results.

  • Remember that the Discussion section is not the place to introduce new information.

  • The ICMJE Uniform Requirements advises: “Avoid non-technical uses of technical term in statistics, such as ‘random’ (which implies a randomizing device), ‘normal,’ ‘significant,’ ‘correlations,’ and ‘sample.’” [4]

  • Eliminate overly-clever phrases and clichés.

  • Consider ruthless removal of anything that causes you to think, “I just want to get this fact in print.”

  • Reconsider acknowledgements: Remember that leaving someone out can lead to hard feelings.

  • Assure that you have clearly identified any potential conflict of interest.

  • Have a last meeting of the research/writing team to assure that everyone knows exactly what is being submitted and that all agree, once again, on the initial target journal.

Thoughts About Research and Research Reports

Quality Writing and Research Design

Medical composition is a laudable skill, one that we should all work to improve. Wager, in an article telling “What Medical Writing Means to Me,” [11] observes that medical writing “inhabits a strange boundary zone between science and art.” When it comes to writing a report of a clinical research study, however, the art of medical writing skill must take a back seat to research design. Have you ever read a research report and wondered whether the skillful prose—perhaps composed chiefly by an editorial assistant—masks questionable methods or unjustified conclusions? As Dirckx [12] has written, one should guard “against the temptation to cover his lack of information with a rhetorical snow job, to palm off muddy thinking under a veneer of smooth writing.” Medical writing, especially in the case of research reports, is chiefly about medical science, and here art cannot trump science.

Stating What You Really Think

Reports of clinical research studies are often written by committee; the members seek consensus on what will appear in print. Perhaps this is why the final version of the paper does not always include the heartfelt opinions of some researchers on the team and often does not reflect the diversity of author opinions. Richard Horton, editor of The Lancet, surveyed contributors to ten research articles published in The Lancet. Thirty-six of 54 contributors to the ten articles responded to questions in a qualitative analysis. The research question in the study was: “To determine whether the views expressed in a research paper are accurate representations of contributors’ opinions about the research being reported.” [13] The study found unreported concerns about study weaknesses, and disagreements among authors about findings and their significance. The study concludes that one remedy for the problem of suppressed opinions may be structured Discussion sections in research papers, as we now see in Abstracts.

Research Mentors

Research is best undertaken in teams, and members of the team bring different skills, one of which may just be mentoring. Research mentors can be especially important team members, who provide nurturing and guidance to the less experienced. They help keep young researchers on track, which can yield surprisingly good outcomes. Hoff [14] writes: “When I finished medical school, I did not intend to do research as part of my life in surgery. That all changed when I met a mentor who inspired me during my training days. I had some protected time, assembled space and equipment, developed a hypothesis, and went to it. I’ll never forget my first experiment and publication. Frankly, it was my best.”

Getting Your Research Report in Print

General Douglas MacArthur once said, “There is no substitute for victory.” In academic medicine, there is no substitute for publication. You can have a brilliant idea, perform groundbreaking research, and write the results with great proficiency, but if the paper is not published—so that it can be cited, criticized, or praised—then the effort has been largely wasted. The advancement of science depends on sharing of knowledge in print. Chapter 12 discusses how to achieve publication, for your research report or other publication models.