II. Special guidelines for original research articles preparation
Introduction. In this section, a research background should be clearly formulated: a problem essence and its significance should be specified. Authors should introduce a studied problem to readers, to describe in brief what is known in this field, to mention works of other authors, to specify drawbacks of previous studies if there are any, in other words, to reason the necessity of their study. One shouldn’t list all published works on the topic, it is enough to mention the most significant works directly relating to the topic. It is recommended to make references not only to national, but also to foreign research in the studied topic.
At the end of this section, state a study aim or a hypothesis tested by authors. The aim should be formulated in such a way that readers have a complete grasp of what is planned to study, persons that are going to be studied and methods.
Methods. The section should include only those methods which were supposed to be used at a project planning stage according to an original study protocol. If in the process of a study implementation, a need in additional methods appears, they should be presented in the section “Discussion”. The section should be very detailed, so that readers could independently assess methodological benefits and implications of a study as well as reduplicate it if desired.
In this section, it is recommended to present clear descriptions of the following aspects (it is not necessary to separate them into subsections): a study type; methods of selection of study participants; methods of measurements; methods of data presentation and processing; ethical principles.
- Study design. In this subsection, a type of an implemented study is clearly specified (a literature review, observational, experimental etc.). In implementation of an observational study, it should be indicated, if the study is descriptive or analytical. In analytical studies, a kind of a study is determined: cross-sectional (one stage study), case-control, cohort, environmental (correlation) etc. It is recommended to specify a year and a month of a study implementation, particularly if symptoms with seasonal variation were studied. In literature reviews, criteria of their publication selection should be clearly indicated.
- Study objects selection. In this subsection, it is clearly specified how patients or laboratory animals were selected for observations or experiments. Criteria for inclusion of potential participants in a study and exclusion from it are specified. It is recommended to indicate the total formation for participant selection, what is also used for result extrapolation. If the parameter “race” or “ethnicity” was used in a study, it should be explained how the parameter was assessed and why it was important to use it. For observational studies, one should specify methods of sample creation (simple random, stratified, systematic, cluster, multistage etc.) and to ground inclusion of that exact number of participants in a study. For experimental studies, it is necessary to indicate presence or absence of study participants’ randomization. A randomization procedure description should be presented. Besides, one should indicate, if a masking procedure was carried out. Calculations of sampling minimum necessary size for check of statistical hypotheses or retrospective calculation of statistical power for main calculations are welcome.
- Measurement technique. All procedures of parameter measurements, data collection, medical or diagnostic interventions should be described in detail, so that it is possible to reduplicate a study according to a presented description. If necessary, it is possible to make a reference to a detailed description of used methods. If a researcher uses his/her own variant of an earlier described method or proposes a new method, it is necessary to present a brief description of the used variant or the proposed method, as well as any reasons against use of generally accepted methods. Titles of medicinal agents (both trade and international), chemical substances, doses and ways of administration of preparations applied in a study should be specified. Used devices, instruments, medicinal preparations etc. should be accompanied by references to manufacturers.
- Methods of data presentation and processing. It is necessary to describe used methods of data processing so thoroughly that readers having access to basic data can check results. In cases of doubt, the Editorial Body can request basic data from article authors in order to check presented results. In this subsection, definitions to all statistical terms, symbols and abbreviations used in an article should be given. For example, М — arithmetical mean, SD — standard deviation, m — standard error of arithmetical mean, Mе — median, Mo — mode, etc. If statistical hypotheses were checked in a study, one should specify the significance critical level (р) accepted by authors. Hypotheses must be stated clearly and described in popular language.
The Journal Editorial Board does not recommend to rely only on use of the reached level of significance during check of statistical hypotheses, as the value р does not indicate complete information. It is recommended to present results with corresponding errors and uncertainty parameters (confidence intervals). In description of statistical methods, references to manuals and guides should be given with obligatory page indications. Besides statistical procedures for check of hypotheses, it is recommended to calculate the effect value for most significant comparisons.
If several statistical criteria were used in a study, one should mention all of them and indicate which of criteria and in what situations were used. Application of one or other methods of data processing should be clearly reasoned. For example, if parametric criteria were used, one should describe what criteria were used for distribution check. It is necessary to specify how observation of conditions necessary for use of methods was checked. Each of used criteria should be specified so that to exclude different interpretations. For example, if sample means were compared with help of a Student's criterion, one should indicate which of the Student's criteria (for independent samples or for paired observations) was used in the study. It is not enough to say that a correlation analysis was used, it is necessary to indicate which of the correlation coefficients was calculated. In use of the multidimensional methods for data processing, it is specified, what was a method of selection of variables for inclusion into models and what categories were used as comparison categories. If a rare method of data processing was used, it is necessary to indicate, why exactly that method has been chosen, to make a reference to a literary source and to describe in brief the used method.
If a package of statistical programs was used for data processing, one should indicate its name and version. It is not necessary to inform, what computer was used for data processing since such information does not have practical value.
- Ethical principles. If an article contains a description of manned experiments, it is necessary to indicate if the procedure corresponded to the standards of the Ethical Committee responsible for this aspect of the work or to the Helsinki Declaration of 1975 and its revision of 1983. One shouldn’t call surnames and initials of patients, numbers of case histories, particularly if an article is accompanied by illustrations or photos. If laboratory animals were used in a study, it is necessary to specify animals’ species and number, used methods of anesthesia and sacrifice according to the “Rules of Work with Experimental Animals” approved by the USSR Ministry of Health or rules adopted in authors’ institutions, the recommendations of the National Research Council or the applicable law.
Results. The section is destined only for presentation of main research results. Results received during a study are not compared with results of analogous studies of other authors and are not discussed.
Results should be presented in a text, tables and figures in natural sequence on the basis of priority of study tasks and goals. It is not recommended to duplicate in a text results presented in tables or in figures and vice versa. Only significant observations relating to study tasks are described, separated and summarized. General characteristics of samples or studied groups should be presented in a table with indication of main studied symptoms. It is necessary to specify not only average values, but also measures of dispersion or confidence intervals for means and fractions.
Mean values should not be more exact than by one decimal sign in comparison with basic data. In descriptions of fractions, it is necessary to specify absolute numbers of observations, particularly in small samples. Percent is stated with two decimal signs only if a fraction is less than 1 %. If a fraction is 1-10 %, one decimal sign is enough. Use of 95 % confidence intervals is welcome both for mean values and fractions.
The reached level of significance for each of used statistical criteria should be given within the accuracy of three decimal signs. The Editorial Body highly recommends to avoid formulas like р < 0.05 (exception is in cases when p <0.001). Instead of the formulas p >0.05 or “differences are insignificant” one should specify the absolute value р within the accuracy of thousandths (e.g., р=0.032). Besides the reached level of significance, it is recommended to give actual criteria values and a number of degrees of freedom. For example, the Pearson’s chi-square criterion represents the following in presence of two numbers of freedom: χ2=29.2, d.f.=2, р <0.001. It is not correct to indicate that “the differences between the average values are reliable (unreliable)”, it is more correct to say about statistically significant (or insignificant) differences.
It should be always noted that detection of statistically significant differences does not mean that reliable or clinically significant differences as well as cause-effect relations are present.
Discussion. In articles describing original studies, this section begins with a brief (not more than 2–3 sentences) presentation of a study main results. Results corresponding to study goals and tasks are considered as main results. There is no point in drawing attention to side results, only because statistically significant differences have been detected during statistical hypotheses checking. Materials described in sections “Introduction” and “Methods” are not to be repeated in this section. It is necessary to specify new and important aspects of a study and, just as important, to explain reasons for obtaining these particular results. One should describe available shortcomings of a study, especially if they can influence significantly obtained results or their interpretation. Besides, one should note study strengths or its advantages over other studies in the field. Discussion of virtues and shortcomings of a study is an important part of the section and is qualified to help readers to interpret obtained results. It is always necessary to remember, that it is better to draw attention of readers to available shortcomings (there is no such thing as a perfect study), than to get a situation when these shortcomings are noted by reviewers or readers. In the section, it should be described, how obtained results correlate with results of similar studies conducted by other authors. Instead of bare mention of previous studies, one should try to explain why obtained results differ or do not differ from results of other authors. Possibilities of application of obtained results should be discussed as well as limitations of their application if any. It needs to determine directions of further studies resulting logically from study results. It is possible to formulate new hypotheses, only if it is justified, and one should clearly specify that these are only hypotheses. Sometimes in this section, operational recommendations for use of study results in practice can be presented.
Conclusions should proceed from study goals avoiding unfounded statements and conclusions which are not based on presented observations or calculations. For example, there is no point in making conclusions about economic expediency of application of a new method of treatment of patients with a disease Х, if in an article, comparative economic effectiveness is not analyzed. It is not worth to state priorities referring to incomplete work.