How to optimally work with references for your journal manuscript

Tudor Toma


One of the problems with manuscripts submitted to the Pneumologia journal is the inconsistency in which references are cited and listed. In this article we discuss how to optimise the work with article references, including: how to select the references, how to cite in text and how to format the reference section of a manuscript, so that the process is as accurate and as easy as possible. In addition, the use of special reference management programs is also discussed. The information outlined herein has the potential of encouraging and helping others to establish their own efficient technique of working with references.


medical writing, references



Progression in the medical career continues to rely to a large extent on the scientific publications in the portfolio of each physician. Nevertheless, the technique of medical writing is generally ignored during medical education, which is why many continue to consider that there is, in fact, no technique involved, but rather an inborn talent for writing and, perhaps, professional opportunity. One of the problems seen in many of the manuscripts submitted for publication to the journal Pneumologia is the inconsistency with which references are cited and listed. References, however, are a crucial part of a scientific article. References must be selected, cited and listed correctly in order for a manuscript to be acceptable for publication. Although it may seem like an administrative and relatively simple activity, which can be left in the responsibility of a student or a junior doc-tor that will be mentioned as authors, unless it is per-formed accurately, the reference arrangement is a highly time-consuming activity, prone to numerous mistakes. These easily avoidable mistakes create problems for medical journal editors, but mainly for the authors – who do not only see their article rejected, but have to reformat their references again, in order to be able to submit the article to another journal.

Why are references so important? On the one hand, by citing the work of an author, the source of informa-tion is acknowledged. On the other hand, the use of a standardised format for references allows automatic counting of citations, enabling the evaluation of rele-vance (quantified by the number of citations) of an indi-vidual author, and of the journals that publish frequently cited articles.

The technical aspects for correctly working with ref-erences are discussed further below: how to select the references for your article, how to manage them, how to cite them in text and how to format the reference sec-tion. The information herein is only a guideline, and not the absolute. This article should primarily help and encourage potential authors to establish their own tech-nique of working with references in the process of writ-ing a manuscript.

How to find and select the right reference

The process of selecting references for an article is complex and time-consuming. It should be performed according to an algorithm, following a specific method. The process should be developed by each author, accord-ing to his/her preferences or to the advice and experi-ence of his/her academic mentor, and/or other information previously published in the literature. There are many articles where the author selects incor-rectly, incompletely and, sometimes, even randomly, or based on certain preferences, the articles entered as references. It should, however, be taken into account that for the peer-review journals, the reviewers also check carefully the references and, if there are mistakes, the article is returned to the author or rejected.

How should references be found and selected? There are at least two ways in which the text of an article can be generated:

  1. The article can be written alla prima, out of mem-ory, and, subsequently, the statements are linked to references.

  2. Or, more commonly, the references for a particular topic are first collected, indexed and reviewed, and, subsequently, the text of an article is written according to the statements in the references.

Which is the best method? Every author builds and adapts his/her own specific method. It is usually a com-bination of the two methods. Ideally, the author should write the text out of memory, without references, to produce a body of text as original as possible, and, sub-sequently, the general statements are linked to particu-lar references. This is, however, difficult to accomplish for the novice authors, as it requires a thorough knowl-edge of the topic covered by the manuscript.

Usually, the author collects and reads in extenso as many references as possible on the topic in question before starting to write. This is the period when the author collects a large number of articles, which will not all be cited in the final article, but which have to be ordered and indexed in a database for further use.

Ordering can be done manually, only on paper, using bibliography cards, or electronically, either using data-bases such as Excel or Word, or by saving all articles as PDF files. For optimising the management of a large number of such references, the use of specialised refer-encing software is recommended.

How to find and select references that matter?

First, several general keywords that are, neverthe-less, relevant for the subject of the article in progress are selected, and the publications available are retrieved and analysed, in particular the ones on PubMed or other databases available online.

The last good article published in close relation to the topic of the manuscript is selected as grounds for the article in progress, and the references in this article are thoroughly examined. Some references are essential, others are merely specific details. Essential references are further sought out and read in extenso. The respec-tive articles will include, in their turn, references, which should be searched for and perused. Thus, in an “oil stain” pattern of searches, the reference database expands, one reference leading to another. The process can be repeated for other keywords or topics associated with the main topic.

Not all articles are available free online, and not all journals that matter are easily available for the author in the library of the institution where the author works. There is, however, an unwritten solidarity between pub-lished and future authors, which should be used when one finds it impossible to access an article that is crucial for the study in progress. An e-mail asking the corre-sponding author to send their article as PDF file for research and citation purposes is, in most cases, fol-lowed by a positive reply. This occurs when the author has the right to disseminate the full-text form, given that some journals limit such rights. Authors do this, because they can thus increase the number of citations for their articles. A positive reply to such a request is not solely out of sympathy for younger authors, but also out of the motivation to increase the number of personal citations.

A database of references that matter for the topic of the article in progress is thus formed and should be organised. This is a time-consuming process that requires scheduling and planning.

Who should be cited?

The statements regarding “what is known”, which is usually part of the Introduction section of an article, should be linked to the articles that have demonstrated the issues in question. A citation should be made for the original paper that proved the statement, and not for an article addressing the same topic and which, in turn, cites a secondary article. The author should follow the citation chain until finding, for each statement, the original source paper. If the statement has evolved, the last paper to demonstrate the statement in question should be cited, and not necessarily the very first early version of the statement demonstration. There are, nev-ertheless, many articles that cite editorials, meta-anal-yses, secondary articles or book chapters – and this is most of the times inaccurate. In addition, articles cited should be available and accessible, and this is the reason why articles written in English are cited more frequent-ly. Self-citation must be avoided – this is when the author is citing himself/herself in order to boost the number of citations. Similarly, citations within the same journal, only to improve the impact factor of the journal in question, are not considered an ethical practice.

For the Methods section, the manuscript should refer-ence the methodological elements that were used in other articles. In addition, the references should be only articles that introduce and present in detail a particular method used in your paper, and not editorials or second-ary articles. A secondary article can be cited, however, if the respective article adapted or optimised a particu-lar method, and the author is currently using this opti-mised/adjusted method, and not the original method.

The Results section of a paper usually has no cita-tions, unless the expression of results follows a particu-lar model used by other articles.

In the Discussion section, all source articles contain-ing similar studies which are compared with the current study should be cited. The discussion should be able to answer with references to the question: “What have others done in the same field, and why were the results similar or different compared to other studies?”

How are references arranged and managed?

For small studies, with a limited number of refer-ences, the in-text arrangement, entry and listing of references is not a major problem. It can be done manu-ally. Most often, however, a software specialised in the collection, management and arrangement of references is used.

There are several software programs for reference management. In the past, the most widely used was Reference Manager; unfortunately, this program is no longer available, having been replaced by Citavi(1). For Windows users, Citavi is very good software, but it lacks a version for Mac users, although, with some software tricks, Windows can also be run on Mac systems. Mac users can use Endnote (see endnote.com), which features a free basic version (EndNote basic) as well – and this is recommended as “perfect for your first article”(2,3).

What does reference management software do?

Reference management software collects all the ref-erences selected by the authors in a large database that communicates directly with PubMed, or other online databases. Thus, the author does not need to copy the citations word for word and comma for comma, but just to click Save in a reference management database, and the references can thus accumulate easily. The program can also enter a code in the text of your manuscript where a certain reference is required, and, subsequently, it will arrange automatically both the in-text numbering and the formatting in the bibliography list according to the style required by the journal where the article is submitted(4). Thus, the job of writing by hand the refer-ence’s details and formatting it correctly is automati-cally done by the click of a mouse.

Different medical journals have different require-ments for reference style and formatting. These require-ments are usually clearly stated in the section Instructions for authors. Reference format uniformiza-tion is an extremely laborious and time-consuming activity if done manually. The programs for reference management will automatically format all references according to the individual requirements of each of the

existing journals. This is because such programs include within their databases the style requirements for all major journals and automatically generate references based on the desired journal. Also, minor journals, gen-erally, adopt a style that is similar to that of one of the major journals, and, therefore, even if the title of the journal where the article is to be submitted is missing from the program database, a similar title can be used for quick and correct formatting.

Moreover, if the editor returns the article with cer-tain modifications, and paragraphs with in-text cita-tions have to be inserted or removed, the software renumbers and rearranges automatically the in-text citations along with the reference in the list.

In addition, such a program immensely eases the workload for authors if they have to submit the article to another journal with different style requirements (when rejected).

A summary of the different reference management pro-grams designed by Wikipedia can be found here: (https:// en.wikipedia.org/wiki/ Comparison _of_reference_ management_software)!(5).


References should be found, read and managed per-haps according to the methods suggested above. This activity requires time, must be well-thought out, and minutely planned. The use of reference management software is highly recommended in order to easily meet the requirements set forth in the instructions for authors. Without these programs, articles with a large number of references are challenging to edit and, if refer-ences are not formatted as required by the journal, the article is returned to the author or simply rejected.


  1. Available at: www.citavi.com, accessed: 2018/1/4.
  2. Available at: www.endnote.com, accessed: 2018/1/4
  3. Oliveira MAB, dos Santos CA, Brandi AC et al. Braz J Cardiovasc Surg 2015, 30(2):246-53.
  4. Kessler J, Van Ullen MK. Citation generators: generating bibliographies for the next generation. J Acad Librarianship 2005; 31(4):310-316.
  5. Available at: https://en.wikipedia.org/wiki/Comparison_of_reference_ management_software, accessed: 2018/1/4.


© Copyright 1993-2017 Societatea Romana de Pneumologie. All Rights Reserved.