Scientific publication is an important indicator of a researcher’s productivity. Scientists must publish the results of their work to inform their peers of their findings, support grant and funding applications, and enhance their potential for promotions. Therefore, the pressure to publish is high. Scientists usually look for prestigious journals, i.e., those with high impact factors and a fair review process, in which to publish their work.
Meanwhile, journal editors are engaged in competition to attract and select the best papers and strive to keep their journals interesting, while staying within a limited page allotment. The editors of Clinical Chemistry are interested in attracting articles that have high potential interest and impact in novel technologies/applications, biochemical markers, proteomics, molecular diagnostics, and clinical and epidemiologic studies. Papers are selected on the basis of current and potential future importance to our field, novelty, relevance to the audience, and clarity of presentation. Although many researchers are quite capable of designing excellent scientific studies, we have found as editors that not all are capable of clearly articulating their findings in a well-formulated scientific paper. Clinical Chemistry currently accepts approximately 15% of submitted original reports. Although common reasons for rejecting papers include lack of novelty or a poorly designed study, it is important to emphasize that papers that present their findings most clearly have a definite advantage in the peer review and decision-making process.
In our editorial experience, we find that about a third of submitted papers fail to fulfill at least one journal requirement as described in the Information for Authors (1) and the detailed checklist for manuscript preparation. Common mistakes include having too many references, incorrect reference formatting, incorrect units, and inclusion of supplemental files and figure legends in the main text of the paper. Even after peer review, roughly two-thirds of papers are returned to the authors by the Deputy Editors at the final stages of editing for such things as clarification of text, additional method information or clarification, additional results, revision or shortening of the abstract, unclear figures or tables, unjustified conclusions, and improper use of statistics.
In Clinical Chemistry, papers can be rejected at any stage of the review process. Just because the reviewers’ critiques have been satisfied does not indicate that a manuscript is of sufficient quality to warrant publication. Granted, reviewers make recommendations regarding the science in manuscripts, but, in addition, they assign 7 priority scores to each manuscript, among which are ratings for validity of conclusions, style and grammar, and quality of the manuscript. To intelligently select papers for publication, the editors must take these priority scores into account, looking at each manuscript in the context of other submissions and evaluating its relative importance. This overall selection process can be influenced by how well a paper reads and how clearly the data are presented; a paper that is difficult to read may fail to transmit the true scientific value of the work.
Although there is no single best method for writing a scientific manuscript, here we wish to help younger scientists and junior faculty improve their chances of publication in general, and specifically in Clinical Chemistry, by offering suggestions and points to consider when preparing manuscripts for submission.
Guidance in Manuscript Preparation
The issue of authorship must be addressed before any manuscript preparation. Decisions about whom should be included as authors and the order in which to list the authors should be made at this time. Only those who fulfill the 3 requirements of the journal for authorship, as laid out in the Information for Authors (1), should be listed as authors.
Some authors find it helpful to prepare an outline or list of bullet points to be covered in each section of the paper. Such lists help to avoid the omission of important points the author wishes to make, and provide an underlying organization to the manuscript. More experienced writers may not need such lists, but new or inexperienced writers will find them very useful.
order of manuscript preparation
The order in which information is presented in an article (e.g., introduction, methods, results, discussion) does not necessarily reflect the order in which these sections are written. Table 1⇓ presents an order for preparing a manuscript that works for many individuals. For authors who have difficulty getting started, preparation of the Materials and Methods section represents a momentum builder, since it requires the gathering and organization of simple procedural details but does not usually require a large amount of scientific reasoning. The most important step for an author is to have a clear understanding of the data and the best means of presentation to convey the intended message. Many authors spend considerable time on the preparation of possible tables and figures. The widespread availability of computer software greatly eases this process. Tables and figures tell the story and set direction of the manuscript. It is worth spending ample time in preparation of tables and figures to decide the best and clearest way to present the data. Below, we provide specific recommendations for preparation of each of these sections, using the order in which they appear in print.
Titles provide a self-explanatory brief summary and create interest in a paper. The title gives information about the content of the article. Thousands of readers look at the table of contents or scan through an issue to see what topics are being covered, and a captivating title may be the sole opportunity to entice someone to read the abstract or paper itself. Just as important is the fact that most scientists do not subscribe to many journals, and therefore, a majority of readers will become aware of papers through search engines (e.g., Google) or electronic databases (e.g., PubMed) that key on words found in the title.
Titles need to be both succinct and descriptive. In most cases, titles turn out better if words are removed rather than added. The title should include key words or terms that authors want indexing services and readers to associate with their paper. The title is best left for last when writing a paper. Once a list of key words or terms has been created, or after the Abstract has been written, it is often easier to develop a title that reflects the points the authors wish to make.
The abstract provides an overview of the paper, highlighting the major findings and conclusions of the work. Abstracts are almost always the initial decision point that an editor uses to further consider submitted papers. Likewise, abstracts also provide the first impression of the value of the work for reviewers and can influence the outcome of the review process. Once in print, virtually all published abstracts are readily available via electronic literature indexing services (e.g., PubMed, ISI) and provide a first glimpse of the paper to potential readers who are deciding whether to read the full article. Poorly written abstracts with insufficient information or unsupported conclusions will likely diminish readers’ interest in the work. A well-written abstract will present a rationale for why the study was performed, a summary of the methods used, data that support the main message, and conclusions that do not extend beyond the findings. Abstracts should be written concisely in the past tense, without references, and should never exceed the allowed number of words.
The introduction to a paper should be brief (not more than 2 double-spaced typed pages), yet contain 3 salient aspects:
A brief background establishing the context of the subject matter reported in the paper.
A description of the rationale for the study (e.g., identifying a gap in the knowledge base, a missing application, drawbacks of prior work, or points of controversy).
A summary of the goals of the current study or the hypothesis being tested to address the problem at hand.
Details of materials and methods or findings of the study should not be presented in the introduction.
materials and methods
The Materials and Methods section should contain enough information to allow readers to reproduce the work, adequately evaluate the findings, and compare the study results with other published studies. If anything, the methods should be detailed to a fault. Common errors include leaving out information such as pH or concentration of a buffer; source and grade of solvents and chemicals; length, diameter, and particle size of a column; type and volume of specimen collected; anticoagulant used; collection or storage container type (polystyrene, polypropylene, glass); and storage conditions. As an example, the reagents in an LC-MS assay may be described as containing methanol or acetonitrile, without description of the vendor or purity. Yet it has been shown that both methanol and acetonitrile from different vendors, or of different purities, can affect LC-MS assays (2)(3). Although detail is important, do not provide exhaustive descriptions of already published methods. If the paper describes a newly developed method, however, extensive details are necessary.
If human subjects were part of the study, journals require statements that review and approval of the study was obtained from the appropriate ethics committee or review board and that informed consent was obtained when required. Similarly, if animals were used, declaration should be made that local, national, or international standards of treatment were followed.
Diagnostic studies and clinical trials.
Many submitted papers describe results of studies of the diagnostic utility of laboratory tests and biomarkers or effects of drugs or therapies on diseases. In these types of studies, it is imperative that, at a minimum, the information listed in Table 2⇓ be provided. Even in cases where the study may appear to involve primarily looking at the presence or absence of a biomarker in a disease state, much of this information is still important for readers to evaluate the study and the applicability of the findings.
Many sets of guidelines have been published to assist authors in conducting and publishing studies of various types. Authors submitting papers to this journal are encouraged to consult the reporting guidelines listed below as appropriate, all of which can all be accessed through the Equator Network (http://www.equator-network.org):
CONSORT: randomized controlled trials;
STARD: studies of diagnostic accuracy;
QUOROM: systematic reviews and meta-analysis;
STROBE: observational studies in epidemiology;
MOOSE: meta-analysis of observational studies in epidemiology.
A description of the statistical methods used in data analysis must be provided. If computer software packages were used for statistical analysis, this information should be provided. Measures used to summarize the location (spread) of data should be specified, such as mean (SD), median (range), or median [interquartile range (IQR)1 ]. The latter 2 approaches are best when the underlying data distribution is not normal or not known.
Tests used in significance testing should be described, including the underlying P value used to establish significance and what methods (e.g., Bonferroni, Scheffé, Tukey–Kramer, etc.) have been used to adjust for multiple statistical testing. Nonparametric statistical tests are generally preferred over parametric tests when the experimental data are not normally distributed or the underlying data distribution is not known. When parametric procedures, including ordinary regression analysis or analysis of variance (ANOVA), are employed, the data should comply with underlying assumptions of these procedures [e.g., uniformity of variance (homeoscedasticity) and normality]. Noncompliance with these assumptions may indicate the need for data transformation (e.g., logarithmic or square-root transformation) before such procedures are applied.
Many published resources are available to assist in the description and reporting of statistical analyses of scientific data, several of which have been listed in References (4)(5)(6)(7)(8)(9). In cases where complex statistical analysis is required, authors are advised to consult with professional statisticians. In addition, authors are advised to carefully follow guidance given regarding statistical methods in the journal’s Information for Authors (1).
The Results section provides a clear description of the experimental findings without interpretation and is central to the message of any scientific paper. The information in Results should be presented in the order that is most logical to convey the message, not necessarily the order in which the experiments were performed. The information provided is usually structured around a series of informative figures and tables (discussed below).
Authors must be selective in what data are reported, since not all data may be relevant to the message of the paper. Results from highly technical mathematical or experimental procedures are often communicated best by providing a summary in Results and additional details in Supplemental Data files.
Experimental results should be presented in the most concise format possible. The use of a parallel reporting style often can be used to advantage in reporting several sets of results for different groups, e.g., “The mean (SD) admission troponin concentrations were 0.5 (0.3) μg/L, 4.3 (0.9) μg/L, and 13.3 (1.3) μg/L in the congestive heart failure, percutaneous intervention, and coronary bypass graft patient groups, respectively, whereas the mean (SD) admission B-type natriuretic peptide (BNP) concentrations in these 3 groups were 959 (305) ng/L, 89 (50) ng/L, and 136 (64) ng/L, respectively.” When data are reported like this it is not necessary to repeat the data in a table or figure.
Reported findings should be supported with appropriate statistical analysis including actual P values given to at most 2 significant digits, referring to the statistical approaches presented in the Materials and Methods section. Avoid terminology that draws attention to nonsignificant findings. Examples include “… tended to discriminate … slightly better” when the difference was not significant or “We found a pronounced, but not statistically significant, reduction in 2-month mortality risk. …”
This is the section in which the findings are interpreted and placed into context of earlier work and their significance and impact are explored. The length of the Discussion should be proportional to the amount of data presented. Reiteration of results should be avoided. It is often recommended to start with a paragraph stating the strength of the study and its major findings. Results of the primary analyses should be discussed first, followed by those of the secondary analyses. The order of discussion of the results should parallel their appearance in the Results section. In discussing the significance of the work, it is important to address the issue of statistical vs clinical significance. Speculations about the possible reasons for the findings and the potential benefits of the work are acceptable as long as they are kept within reason. In addition, the limitations of the study and the applicability of the findings, particularly in the case of clinical studies, should be articulated. Finally, the manuscript should conclude with a paragraph about the significance of the study and its potential impact on the practice in the field. Authors should avoid making statements that are not supported by the presented data, however.
In the Information for Authors (1), the allowed number of references for each article type and the required format are clearly indicated. Failure to comply with either requirement results in the returning of the submitted manuscript to authors, thus prolonging unnecessarily the review process time. Many individuals find an electronic reference manager a useful tool. Authors should avoid excessive self-citations to the exclusion of others.
Figures provide visual representations of data that help readers to see, understand, and remember information. Each figure should contribute to the text by the effective visual communication of information. Many resources exist to help with the preparation of effective figures. A representative sampling is included in the reference list (10)(11)(12)(13)(14). We offer some suggestions regarding the preparation of figures:
Always stay within a journal’s requirements for the number and size of figures (and tables). Submitted manuscripts that contain excessive numbers of figures, or figures that are too complex, have an immediate strike against them.
Avoid figures that are too complex. Usually, no more than 4 panels should be integrated into a given figure, and the information displayed in each of those panels should be highly related. Use adequately large font sizes and readable sans serif symbols. Reduce the figure size to 2 by 2 inches (5 by 5 cm) and ensure readability of the text in a figure if it is published.
Use shades that reproduce well. Avoid the use of gray colors and gray shading, because these often do not reproduce well when published. Avoid the use of color if the figure is to be published in black and white. Often much of the distinction between colors is lost when a figure is converted from color to black and white. If preparing color figures for publication, avoid the use of pastel shades. In Clinical Chemistry, we will soon require the use of a specific color palette by authors when color figures or graphs are desired.
Submit figures with adequately high resolution. If submitting bitmap graphics such as tagged image file format (TIFF), use resolution of at least 600 dots per inch (dpi). Ideally, figures should be prepared and submitted using vector graphics formats such as encapsulated postscript (EPS) or Windows metagraphic (WMG). Vector graphics formats can be enlarged or reduced in size with no change in the sharpness of fonts.
Fig. 1⇓ provides an example of a poor figure (top) and a good figure (bottom).
Tables provide a visual summary of a set of data or experimental parameters. Tables can be used to describe the characteristics of a study population or to list the conditions of an analytical method. When a table is carefully constructed, the information presented can be understood much more easily than if the same information were presented in the text. It is important to properly organize the information in a table. Generally the left header and column contain the variable experimental parameters, and subsequent columns, the corresponding results. For example, Supplemental Tables 1 and 2, which accompany the online version of this article at http://www.clinchem.org/content/vol55/issue7, show the relationship of temperature to signal intensity of an enzyme reaction. Is the presentation of data clearer or more visibly appealing in one of the tables?
There are also times when data do not need to be presented in a table. For example, assume that, in online Supplemental Table 2, only the data for Enzyme 1were being reported. Since the enzyme showed no activity below 40 °C or above 60 °C, is there a reason to use page space for a table when the same results could be described in 1 or 2 sentences in the text? The text could state, “The signal intensities for Enzyme 1 were 31 at 40 °C, 78 at 50 °C, and 26 at 60 °C. There was no measurable signal below 40 °C or above 60 °C.”
Responding to Reviewers’ Comments
The Editor-in-Chief and/or the Associate Editors evaluate each manuscript upon receipt to determine its suitability for the journal and whether to send it for review. Manuscripts that survive the initial screening process are sent to 2 or more reviewers. Authors should read carefully the comments provided by the reviewers and make every effort to respond to them. This may require performing additional experiments or further statistical analyses or modifying the text. If the authors do not agree with a reviewer’s suggestion, however, they should diplomatically articulate their disagreement and state their rationale. Authors are advised to respond point-by-point to the reviewers’ comments and highlight the additions and changes in the text to make it easier for the editor to evaluate the revision and render a decision. It can be very useful to ask a neutral colleague to read and evaluate the revised manuscript and the proposed responses to the reviewers’ comments before resubmission. It is important to remember that the Associate Editors and reviewers are volunteers who spend considerable amounts of time evaluating the submitted work with the goal of improving it or enhancing its presentation. Therefore, although it can be frustrating to read criticism of the work or adapt to suggested changes, it does no harm to sincerely thank the reviewers and the editors for their time and advice that invariably improved the manuscript.
advice for non–english-speaking authors
Writing a manuscript can be a daunting task for anyone, particularly for a non–native English speaker who may have difficulty clearly expressing thoughts in English. Such authors would be advised to have their manuscripts examined by a native English-speaking colleague or a language editing service before submission. The authors should reexamine their manuscript after language editing and before submission to ensure that their intended message was not altered by the editing service. For the benefit of authors, Clinical Chemistry provides a list of several language editing services in the Information for Authors (1), with the understanding that the Journal has no particular experience or arrangement with any of these agencies.
A clear articulation of the objectives and the findings of a study, adequate description of the methodologies used, and appropriate presentation of data enable the editors, reviewers, and readers to better assess the overall quality and the value of any work submitted for publication. By following the suggestions addressed in this editorial, we believe that authors will improve the presentation of their work and enhance the chances of getting their manuscript published.
Author Contributions: All authors confirmed they have contributed to the intellectual content of this paper and have met the following 3 requirements: (a) significant contributions to the conception and design, acquisition of data, or analysis and interpretation of data; (b) drafting or revising the article for intellectual content; and (c) final approval of the published article.
Authors’ Disclosures of Potential Conflicts of Interest: Upon manuscript submission, all authors completed the Disclosures of Potential Conflict of Interest form. Potential conflicts of interest:
Employment or Leadership: N. Rifai, Editor-in-Chief, Clinical Chemistry; T.M. Annesley, Deputy Editor, Clinical Chemistry; J.C. Boyd, Deputy Editor, Clinical Chemistry.
Consultant or Advisory Role: None declared.
Stock Ownership: None declared.
Honoraria: None declared.
Research Funding: None declared.
Expert Testimony: None declared.
Role of Sponsor: The funding organizations played no role in the design of study, choice of enrolled patients, review and interpretation of data, or preparation or approval of manuscript.
Acknowledgments: We thank Drs. Mitchell Scott and Amy Saenger for their thoughtful comments on the manuscript.
↵1 Nonstandard abbreviations: IQR, interquartile range; BNP, B-type natriuretic peptide; TIFF, tagged image file format; dpi, dots per inch; EPS, encapsulated postscript; WMG, Windows metagraphic.
- © 2009 The American Association for Clinical Chemistry