Bibliographic research in scientific literature

[reading time: 7 minutes]

The ability to trace the information saved in the databases of the scientific literature constitutes a fundamental aspect not only for the researchers but also for the professionals and the students, so much so that there are countless handbooks, courses and articles on the subject (almost all university institutions offer advice, guides or manuals for the use of PubMed, (for example the University of Padua and the University of Milan).

It is evident that the life of a patient may depend on whether a doctor is up to date on the most recent studies of his specialty. Moreover, a rigorous bibliographic research not only is indispensable to the researchers who have to know the state of the art before designing a new study to avoid duplications and the waste of resources, but also it is an indispensable prerequisite for the conduct of the systematic reviews, which affect the development of the guidelines and therefore on the care erogated by the health systems.

The corpus of the articles and documents containing the scientific publications is defined literature. By convention it is defined as white literature, namely the totality of the texts printed in books, magazines and journals freely accessible by everyone. The white literature is opposed to the gray literature,¹ represented by the information produced by channels different from the commercial traditional ones, which includes, for example, reports, final dissertations, conference proceedings, seminars, round tables, bulletins, registered trials, patents.

The various types of indexed studies in literature are also characterized by planning and different designs, to which an increased level of quality is based on the pyramid of evidence, which provides a hierarchy of studies according to their methodological rigor. For example the observational studies are considered of lower quality compared to the randomized controlled trials (RCT).

Studies are also classified as primary literature, consisting of all unfiltered publications, that is, individual studies, while they are called secondary literature when they are short studies summarizing and critically evaluating the results of primary studies, such as systematic reviews. This kind of study is considered the most reliable and it is collocated at the top of the pyramid.
The articles published in journals are collected in various databases, which index them and allow bibliographical research within their websites by the mean of keywords. The most known is MEDLINE, a search engine for the biomedical domain, free for all internet users and updated weekly. It contains more than 20 millions bibliographical references and is made available by the American National Library of Medicine. Amongst the other databases, some of which with reserved access, there are Embase for pharmaceutical and biomedical literature; CINAHL addressed to nursing and other health professions, to students, teachers and researchers; the Cochrane Library that divides the studies in different databases (for example CDSR which collects the systematic reviews competed by the Cochrane and the protocols of those in course of realization; DARE, containing more than five thousands structured abstracts and references to systematic reviews not produced by the Cochrane; CCTR/CENRAL, collecting clinical controlled studies found in sources different from Medline, Embase and other platforms); Google scholar, which consents to carry out searches easily and immediately in any language. although it gives more ample results, sometimes irrelevant and non pertinent from disomogeneus sources.

The gray literature is harder to find: for example, it is possible to consult the website ClinicalTrials.gov and the International Clinical Trials Registry Platform (ICTRP) of the WHO for registered clinical studies and the website Networked Digital Library of Theses and Dissertations (NDLTD) for final dissertations.

All search engines have more or less refined filters, through which it is possible to set the visualization of articles published in a determined period of time, or the articles referring to a specific kind of study, or those available for no charge, or all the articles written by a single author, etc.

To launch a basic query one or more keywords are simply typed in the search bar to obtain general information that can then be filtered. Professionals, however, have to create strategies to formulate specific queries, also according to their pursued purpose (a surgeon might need to consult the most recent systematic review to decide whether amputate a limb of a patient; or an equip engaged in the research or writing of secondary studies like systematic reviews and meta-analysis has to ensure to gather, as far as possible, all the available material).

The studies extrapolated from the literature through search engines are not consistent in regard to quality and scientific rigor, therefore it is necessary to acquire the right competence to test them, by carrying out a critical evaluation of the search results.

In order to launch a search aimed at selecting studies that can answer a specific query, it may be useful to remember the acronym PICO², in which:

• P stands for patient, population or problem to be solved;
• I stands for intervention;
• C stands for comparison;
• O stands for outcome

The terms of the search can be transferred directly in the query, integrating them with the boolean operators, for example, in the query was: “Is paracetamol better than ibuprofen to alleviate fever in children?”, in the search string P would correspond to children with fever, I to the paracetamol, C to the comparison between paracetamol and ibuprofen and O to the reduction of the fever³.

The amount of results provided by the search engine can be filtered further by adding other keywords to the specific question, depending on the answers seeked, like:

• therapy/treatment/intervention
• diagnosis
• prognosis
• etiology/risk

Over the years the format PICO has been refined further, originating specialized queries defined by the formats PICOS, PICOT, PICOC (respectively adding to the query the variable “S” – design of the study, “T” – arc of time, “C” –context), EPICOT+, PIPOH and PECODR.⁴

The PICO format is useful to find qualitative studies; however, those who were to collect and express critical judgments on articles of qualitative type could find more effective to resort to different formats, such as SPIDER or SPICE^5,6, to mention some examples. Both these formats try to identify the fundamental characteristics of the string necessary to formulate an adequate query to find articles in the qualitative research. The first considers the sample (S), the phenomena (Pi), the design (D), the evaluation (E) and the research (R), whereas the second considers the setting (S), population (P), the intervention (I) the comparison (C) and the evaluation (E).

Particular attention is paid to the bibliographical research techniques aimed at extracting the literature necessary to develop the systematic reviews from the databases, since this type of study can have considerable consequences on the entire sector of the Clinical Governance.

Only as an indication, reference is made, for example, to a 2018 article that identified the search process in literature as a crucial factor for the completion of systematic reviews, reporting the existence of nine guides for the conventional approach to this type of search, united by eight fundamental steps.⁷

The PRESS guidelines are also relevant; they contain a checklist addressed to the peer reviewers in charge of evaluating or validating the strings adopted by the authors of a systematic review in their search of articles, in order to detect any errors⁸ .

The PRISMA-S⁹ guidelines (an extension of the PRISMA guidelines, which dictate the criteria for the reporting of the systematic reviews and of the meta-analyses) have been developed especially to ensure maximum transparency and reproducibility of the undertaken searches to gather the studies on which to carry out the systematic review. The PRISMA-S contains a checklist of 16 elements, featuring explanations and examples – the element number 8 recommends that the researchers report the search strings used with precision in their article, copying and pasting them in the text exactly how they launched them.

With regard to qualitative systematic reviews, a team of researchers has examined the research methods used in forty-three different studies, noting considerable variations and hoping for the adoption of a uniform standard. To this end they proposed the STARLITE guidelines.¹⁰

Bibliography

1. Ferroni E. Parola chiave: Letteratura grigia, CARE 3, 2014: 32-33.
2. Richardson WS, Wilson MC, Nishikawa J, Hayward RS. The well-built clinical question: a key to evidence-based decisions. ACP J Club. 1995 Nov-Dec;123(3):A12-3.
3. Caldwell PH, Bennett T, Mellis C. Easy guide to searching for evidence for the busy clinician. J Paediatr Child Health. 2012 Dec;48(12):1095-100.
4. Lazzari, G., Salvini, L., Patella, S., Ausili, D., & Di Mauro, S. La strutturazione del quesito clinico per reperire le prove di efficacia: una revisione della letteratura. L’infermiere, 6 (2015).
5. Methley AM, Campbell S, Chew-Graham C, McNally R, Cheraghi-Sohi S. PICO, PICOS and SPIDER: a comparison study of specificity and sensitivity in three search tools for qualitative systematic reviews. BMC Health Serv Res. 2014 Nov 21;14:579.
6. Rehman, Y. (2021). Difference between Quantitative and Qualitative Research Question-PICO vs. SPIDER. American Scientific Research Journal for Engineering, Technology, and Sciences (ASRJETS), 77(1), 188-199.
7. Cooper C, Booth A, Varley-Campbell J, Britten N, Garside R. Defining the process to literature searching in systematic reviews: a literature review of guidance and supporting studies. BMC Med Res Methodol. 2018 Aug 14;18(1):85
8. McGowan J, Sampson M, Salzwedel DM, Cogo E, Foerster V, Lefebvre C. PRESS Peer Review of Electronic Search Strategies: 2015 Guideline Statement. J Clin Epidemiol. 2016 Jul;75:40-6.
9. Rethlefsen ML, Kirtley S, Waffenschmidt S, Ayala AP, Moher D, Page MJ, Koffel JB; PRISMA-S Group. PRISMA-S: an extension to the PRISMA Statement for Reporting Literature Searches in Systematic Reviews. Syst Rev. 2021 Jan 26;10(1):39.
   Booth A. “Brimful of STARLITE”: toward standards for reporting literature searches. J Med Libr Assoc. 2006 Oct;94(4):421-9, e205.