Nobody disputes that bibliographic references are essential to any scientific article: they support what the author has to say, they enable the reader to find the bases for any data of interest to their own studies, they endorse appreciation of the contemporariness of the study and the breadth of the areas underpinning its publication.

However, searching for a bibliographical reference is not usually easy. We cannot always access all journals, some journals disappear, the authors are difficult to contact, and references cannot therefore be directly requested.

Lately this difficulty has increased due to the web page links provided: the speed with which the web alters suggests that these bibliographical references will be hard to locate in upcoming years.

To determine whether this supposition was real I reviewed the bibliographical references of the first 3 issues (one quarter) of 2018 and of the same issues of the year 2007. I noted the number of links cited and how many of them were still available.

In the first 3 issues of volume 3 in 2007 663 bibliographical references were made, of which only 4 (.6%) were links to web sites and only one is attainable.

Between the 2018 January–March–May issues (Vol. 14-2018) there were 957 references, 36 (3.8%) of which referred to web pages. Of these, 22 (61%) are attainable and 14 (39%) no longer are. These data offers us double confirmation: references to web pages are on the up and it is difficult to gain access to many of these web pages.

All of the above has led me to review the publication regulations which your publishing house establishes for its authors. I have been able to confirm that the journal itself defaults in its instructions to its authors. It explains how the format of a reference to a web page should be with an example using this link, which is no longer active: http://www.cancerresearchuk.org/aboutcancer/statistics/cancerstatsreport/.1

I propose 2 steps as a solution to this problem: (1) request that the proofreaders be stricter with regards to this important publishing issue, checking that each one of the references is attainable when the article is sent to the editorial board and (2) that the publishers request a PDF copy from the authors of the referenced web sites and that these be stored with the publication documentation, because the persistence of the link and the author's contact details cannot be guaranteed.