A search strategy for retrieving articles from the LILACS database in a systematic review of diagnostic accuracy studies: an experimental study

ID: 

288

Session: 

Poster session 2

Date: 

Monday 17 September 2018 - 12:30 to 14:00

All authors in correct order:

Pereira R1, Tersi P1, Puga M1, Atallah Á1, Ribeiro T2, Macedo C1
1 Paulista School of Medicine, Federal University of São Paulo (EPM-UNIFESP), São Paulo, Brazil
2 Science and Technology Institute, São Paulo State University (ICT-UNESP), São José dos Campos, Brazil
Presenting author and contact person

Presenting author:

Arnaldo Silva

Contact person:

Arnaldo Silva
Cristiane Macedo
Abstract text
Background:
LILACS (Latin American Caribbean Health Sciences Literature) is described in the Cochrane Handbook for Systematic Reviews as one of the resources for systematic reviews of diagnostic test accuracy. Although many studies have already recognized the importance of LILACS in systematic reviews of clinical trials, guidelines on the search strategies for diagnostic studies are lacking. The use of incorrect or incomplete search strategies can impair the retrieval of important studies that are indexed only in LILACS, therefore it is important to develop search strategies that are specific to this database and the type of study.

Objectives:
To propose a new search strategy model to be used in the LILACS (Latin American Caribbean Health Sciences Literature) database for systematic reviews of diagnostic accuracy studies.

Methods:
In this experimental study, we created a search strategy to be used in the LILACS database and then tested the strategy for effectiveness. We then tested the search strategy created for effectiveness in LILACS, in August 2017, using as the gold standard the systematic reviews of diagnostic accuracy studies already published in the Cochrane Library. We registered the number of studies retrieved.

Results:
Of 70 Cochrane Reviews of diagnostic accuracy studies, only 29 had searched the LILACS database and only three had used adapted strategies. These used limited boolean operators only, without modifications in tags or languages. The adapted strategy proposed here was able to retrieve a higher number of studies than the original strategies, especially when using Portuguese and Spanish terms.

Conclusions:
The changes made to the original strategies built for MEDLINE, adapted for LILACS, resulted in a more complex strategy that allowed the retrieval of more studies. These additional studies would have been missed in strategies that were not tailored for LILACS.

Consumer involvement:
There was involvement of researchers testing this new search strategy for the LILACS database to be used in a systematic review of diagnostic accuracy studies.

Relevance to patients and consumers: 

This study created a new search strategie for the authors wich produces of a kind of research called systematic review of diagnostic accuracy studies. This kind of research investigate and compare different diagnostic tests. This new tool proposed here can positively impact and improve the methodological quality of this important category of scientif study.