Article type
Year
Abstract
Background: Regional activity-based HTA is used to evaluate mainly new technologies before introduction in our health care region. In 2009, we added the database EMBASE for our literature searches to complement the use of PubMed, the Cochrane Library, Cinahl, PsycInfo and others. An increased yield of published studies could potentially have an impact on the conclusions and the quality of evidence.
Objectives: To evaluate the impact of the addition of EMBASE in the literature search process on conclusions in the reports and the quality of the evidence according to GRADE.
Methods: In all literature searches from November 2009 to December 2011 we listed articles identified by EMBASE alone. Through a qualitative evaluation, the impact on conclusions and evidence levels for the relevant outcomes in each such article was independently assessed by five in-house HTA experts using, when appropriate, a sensitivity analysis.
Results: Twenty-two HTA-reports have been completed following the addition of EMBASE (Table 1). Six of these reports (27%) included publications that were only identified in EMBASE. Three of them assessed medical methods (3/13), two medical devices (2/5) and one pharmaceutical therapy (1/4). Exclusion of the EMBASE publications would have resulted in a minor modification of the conclusions regarding two outcome variables in one report. Furthermore, it would also have led to reconsideration of the quality of evidence for three outcome variables in two of the reports and a final change in evidence level (GRADE) for one of those outcomes.
Conclusions: In one fourth of our HTA-projects the utilization of EMBASE resulted in a yield of unique publications. It is impossible to predict by type of technology topic, whether EMBASE will provide additional publications. Exclusion of EMBASE is associated with small but significant rates of slightly altered conclusions and evidence levels.
Objectives: To evaluate the impact of the addition of EMBASE in the literature search process on conclusions in the reports and the quality of the evidence according to GRADE.
Methods: In all literature searches from November 2009 to December 2011 we listed articles identified by EMBASE alone. Through a qualitative evaluation, the impact on conclusions and evidence levels for the relevant outcomes in each such article was independently assessed by five in-house HTA experts using, when appropriate, a sensitivity analysis.
Results: Twenty-two HTA-reports have been completed following the addition of EMBASE (Table 1). Six of these reports (27%) included publications that were only identified in EMBASE. Three of them assessed medical methods (3/13), two medical devices (2/5) and one pharmaceutical therapy (1/4). Exclusion of the EMBASE publications would have resulted in a minor modification of the conclusions regarding two outcome variables in one report. Furthermore, it would also have led to reconsideration of the quality of evidence for three outcome variables in two of the reports and a final change in evidence level (GRADE) for one of those outcomes.
Conclusions: In one fourth of our HTA-projects the utilization of EMBASE resulted in a yield of unique publications. It is impossible to predict by type of technology topic, whether EMBASE will provide additional publications. Exclusion of EMBASE is associated with small but significant rates of slightly altered conclusions and evidence levels.
Images
