gms | German Medical Science

Background: Horizon scanning systems use a variety of information sources to identify emerging health technologies (EHT) that will reach the markets in the coming years. Among these sources, PubMed stands out for its potential in detecting technologies, it is a free biomedical database maintained by the US National Library of Medicine in which a large number of healthcare specialties and their main medical journals are represented. Attending to this, Avalia-t validated an exhaustive search strategy in 2012, to identify EHT through the literature as soon as it is published. However, publications have experienced an exponential growth that has driven us to refine the strategy to reduce the number of references.

Methods: The strategy was designed in order to match adjectives and nouns commonly used for authors in titles e.g. “advance technology”, “novel device”, “innovative procedure”, “new intervention” among others. There were discarded from the strategy, using NOT, terms about food, environment, animals, or drugs. The search is launched each semester, filtering by language (only English and Spanish) and by record with summary. Editorials, letters, and opinion articles were discarded. The curation process of the information includes the screening of the retrieved results by HTA methodologists and the final filtration by clinical experts. The results analysed comprised a three year period (2021 to 2023).

Results: The strategy retrieved 3,531 references (0.20% of the amount of references in PubMed) in 2021, 2,233 (0.13%) in 2022 and 3,563 (0.21%) in 2023. Meanwhile, the previous search would account for 26,775 (1.5%), 27,230 (1.53%), and 26,247 (1.55%) respectively.

After a first screening made by HTA methodologists the records selected were 263 in 2021 (7.44%), 136 in 2022 (6.09%), and 69 (1.94%) in 2023. Then, the clinicians screened the records resulting in 55 technologies in 2021 (1.56%), 27 in 2021 (1.21%) and 22 in 2023 (0.62%).

The analysis of results from final prioritization for reports showed 3 technologies (0.08%) selected to develop an EHT report and 1 to HTA report (0.03) in 2021. In 2022, 2 (0.09%) technologies were prioritized for EHT report and 3 (0.13) for HTA report. 2023 data cannot be analysed yet.

Conclusion: The results show that search refinement was effective in reducing the number of references screened. The use of information sources other than PubMed within the HSS complements the identified technologies and avoids the loss of relevant data.