New research article preparation guidelines

Cephalalgia publishes papers from various fields of headache research, including randomized trials, clinical research studies, population-based studies, neurophysiologic studies, studies involving animal data and others. Although each of these fields has different methodologies, it is the journal’s role to ensure comparable manuscript structure and standard reporting to ensure that readers will get the most out of a paper.

The editors of Cephalalgia have recently approved new research article preparation guidelines that are posted on our website under point 2.1 of the Manuscript Submission Guidelines (http://www.uk.sagepub.com/msg/cep.htm#ResearchArticlePrepGuide). These guidelines have three main sections: general aspects, specific reporting aspects, and methodological and statistical guidelines. The new guidelines are summarized in Table 1.

Table 1.

Cephalalgia’s new research article preparation guidelines.

General aspects
Title: Use titles that describe the main aspect of your study, stimulate interest, are easy to read and concise, and state the design of the study (i.e. randomized controlled trial, case–control study, cohort study, etc.). Main findings or interpretation of the study should not be included in the title.
Abbreviations: Do not use abbreviations unless absolutely necessary. Often abbreviations make it difficult for readers to follow a paper, particularly if they are not experts in your field; consider abbreviating long names of chemical substances, genetic polymorphisms and terms for therapeutic combinations. It is appropriate to use abbreviations that are largely known, such as DNA. If abbreviations are used in tables or figures to save space, please explain all abbreviations in the footnote or legend.
Introduction: Briefly introduce the background and significance to concisely set up the context of the specific research question for readers. Remember that your audience and the readership are generally knowledgeable about the issues related to common headache disorders and these aspects do not need to be repeated. Economy of words is important and comments should be essential and specific to the subject matter of the manuscript and need for the study. End the introduction with a clear statement of the study’s objectives or hypotheses.
Methods: For studies involving humans, describe how participants were selected and enrolled, and the sites or setting from which they were recruited. Describe study procedures including any details of interventions (if applicable), measurement and classification of main exposure (if applicable) and outcomes, and other data collection techniques. Consider the use of a figure to show study processes. Report how many individuals were eligible, how many declined to participate and how many were lost to follow-up. For studies that have numerical data and use statistical inference, include a section that describes all details of the statistical analyses (how groups were compared, model building strategies, specific software(s) used, etc.). See also specific statistical reporting guidelines below.
Results: Fully describe the sample and setting of the study (if applicable) and provide characteristics of your study population. Present the finding of the primary outcome first followed by the result of secondary outcomes, exploratory outcomes and subgroup analyses. Consider presenting main results in tables or figures and avoid repeating the same numbers in text, tables and figures.
Discussion: We encourage authors to structure the discussion and cover the following aspects: Summary of the main findings (primary outcomes first followed by secondary outcomes), discussion on how the findings compare with previously published studies, a brief description on potential biological mechanisms (if applicable), clinical, scientific and/or public health implications, strengths and limitations, unanswered questions and suggestions for further targeted research (if applicable).
Funding: For all studies, include a statement describing all and any funding sources and the role of each funding source for the study. If the study had no external funding source or if the funding source had no role in the study, state so explicitly.
Ethics or institutional review board approval: Please clearly indicate that the study obtained appropriate approval (or a statement and explanation of why it was not required), including the name of the ethics committee(s) or institutional review board(s) and the number/ID of the approval(s). For human studies, please also add a statement that participants gave informed consent before taking part in the study.
Study protocol: If your study protocol is registered (ClinicalTrial.gov, etc.), please provide the registration number (required for intervention studies). We encourage the registration of observational study protocols.
Specific reporting aspects
The following checklists are required and must be uploaded with submission: CONSORT statement (for reporting of randomized controlled trials: please use the appropriate extension to the CONSORT statement, including the extension for writing abstracts). STARD (for reporting of diagnostic accuracy studies). STROBE (for reporting of observational studies in epidemiology). PRISMA (for reporting of systematic reviews and meta-analysis of randomized controlled trials). MOOSE (for reporting of meta-analyses of observational studies). STREGA (for reporting of gene–disease association studies) STrengthening the REporting of Genetic Association studies – an extension of the STROBE statement. ARRIVE (Animal Research: Reporting In Vivo Experiments). Links to the forms can be found on the webpage of Cephalalgia.
Methodological/statistical guidelines
Percentages: Report percentages to a maximal of one decimal place (i.e. XX.X%). In studies with <300 participants it is recommended to show full numbers (i.e. XX%).
p -values: In the Methods section, please indicate whether you have calculated two- or one-tailed p-values and which cut-off you have set for statistical significance. Please report all p-values. Showing ‘n.s.’ for not significant is not acceptable. For p-values between 0.001 and 0.10, please report the value with three decimal places. For p-values greater than 0.10, please report p-values with two decimal places. For p-values less than 0.001, report as ‘p<0.001’. Exceptions are genome-wide association studies. Do not only show p-values for group differences but show the appropriate effect measure (i.e. relative risk, absolute risk, difference of means etc.).
Relative risk estimates: Show all relative risk estimates with appropriate (i.e. 95%) confidence intervals. Do not show more than two decimal places. In smaller studies, there is often only power to show one decimal place. The term relative risk is often used as a generic term for odds ratios, hazard ratios or rate ratios. We encourage using the precise term depending on the model used to calculate the relative risk measure (i.e. odds ratio for logistic regression models, hazard ratio for Cox proportional hazard models). If you use the term ‘relative risk’ as a generic term, please indicate in the Methods section what relative risk stands for (e.g. we used a logistic regression model to calculate odds ratios as a measure of the relative risk).
Absolute event rate: Please indicate in the table(s) or text how many people had the outcome event(s) of interest according to the exposure or intervention status. In other words, do not show just the relative risk estimate or proportions without showing how many subjects went into the calculation.
Absolute risks: Please consider showing absolute risk (i.e. risk difference, attributable risk, etc.) in addition to showing relative risk estimates. Often relative risks are large when there is only a small absolute effect (i.e. in setting where either the exposure or the outcome are rare), which can lead to over-interpretation of findings. On the other hand, please keep in mind that absolute effects assume causality in a specific setting when interpreting absolute effect estimates. So caution should be used before making strong inference, in particular from observational research (i.e. XX% of the outcome events are explained by the exposure or XX% of the outcome can be avoided when the exposure is eliminated).
Trend: Use the word trend only when you have tested a trend across a specific variable (i.e. dose response) and report an appropriate p-value for trend.
Model building: We discourage the use of ‘stepwise’ or automated selection procedure methods (i.e. such as forward or backward selection procedures) to build multivariable models. Exceptions are studies aiming to build prediction or prognostic models or studies that are set up to generate hypotheses for subsequent research (i.e. hypothesis generating studies, data mining, etc.). Regardless of the approach, the authors should clearly state in the Methods section how a multivariable model was built.
Subgroup analyses: It is encouraged to limit the number of subgroup (or stratified) analyses. Subgroup analyses should be pre-specified and based on biological or clinical plausibility. p-values of appropriate test for interaction should be provided. The inclusion of any non-pre-specified subgroup or stratified analyses must be accompanied by a correction for multiple comparisons (e.g. Bonferroni).
Missing data: Please report the amount of missing data and how you dealt with this.

Although these are guidelines and not firm requirements, the editors of Cephalalgia will rely on them and the manuscript may be returned at any stage of the review process to authors to address specific aspects.

The guidelines will be regularly reviewed and, if necessary, updated. We also plan to add more specific general guidelines for animal studies.

We hope that the guidelines will help to further improve the standards and transparency of research reporting in Cephalalgia and to provide high-quality articles for our readership.