Handsearching respiratory conference abstracts: a comparison with abstracts identified by an EMBASE search

Stovold E1, Hansen S2
1Cochrane Airways Group, St George’s, University of London, UK, 2Cochrane Airways Group, New Zealand

Background: Conference abstracts make an important contribution to the Airways Group Register (AGR). The abstracts of key national and international respiratory conferences are manually handsearched for reports of trials. Identifying these abstracts manually is time consuming. The EMBASE database now includes conference abstracts from relevant respiratory meetings and these are being identified by the weekly EMBASE searches run for the AGR.

Objectives: To compare abstracts found using the Airways Group EMBASE search strategy with the abstracts found by handsearching and determine if an EMBASE search would be sufficient in the future. The Chest 2010 meeting abstracts were used for the comparison.

Methods: The Chest 2010 conference abstracts were handsearched via the Chest website. Each individual abstract was read and assessed for study design and the citation was downloaded for submission to the CENTRAL database. Those abstracts relating to the scope of the Airways Group were tagged and added to the Airways Group Register of Trials. When the weekly search alerts from EMBASE were downloaded, the Chest 2010 abstracts that had been identified by the Airways Group search strategy were extracted. They were manually checked to see if all the abstracts relating to the scope of the Airways Group already identified through handsearching were present.

Results: The EMBASE search identified 151 abstracts from Chest 2010. Handsearching identified 47 RCT/CCTs, 34 of which were suitable for the AGR. A comparison of the two sets found 7/34 missing from the EMBASE set. Five were trials in pulmonary hypertension, and 2 of the missing abstracts were trials in asthma.

Conclusions: The addition of conference abstracts to EMBASE is a welcome development, but handsearching the major respiratory conference abstracts is still worthwhile in view of the missing references.