Background: The amount of medical publications is growing exponentially. This will increasingly lead to the retrieval of a large overhead of irrelevant citations in any kind of comprehensive search strategy. Objectives: In order to reduce workload and increase manageability of search results, we propose some adjustment in the deployment of search strategies and in the use of search filters. Methods: We stated the following hypotheses: Searching has to be modelled closely alongside subject matter, and specific subjects warrant different strategies altogether; searchers may start with highly sensitive and precise preset filters to define outer (high and low yield) borders and after that deploy an optimized filter most suitable to the topic of interest, if needed. We searched PubMed for topics of different conception and impact regarding epidemiology, research area, publicity, and sponsorship. We employed strategies that made use of all available MeSH terms as well as strategies that focused on word string searching with automatic term mapping. After development of a most efficient search strategy for each topic, we applied the corresponding sensitive and specific quality filters from the Hedges Group and The Cochrane Collaboration, and we adapted the quality filters individually for each topic. To estimate the quality of the resulting sets, we used the NCBI filter feature for clinical trials. Results: The use of explicit MeSH terms helped to achieve better precision in cases of highly researched subjects and small difference in hit count between text string search and MeSH-only search. In cases of less well-known concepts, with a large difference in hit counts between MeSH-only and text string, automatic mapping seems to be generally more efficient. Sequentially adding sensitive and specific validated quality filters for the corresponding purpose category guided the formulation of subjectoptimized quality filters. Conclusions: Adaptive strategies, taking into account topic-related qualitative features, lead to more effective search results. There are some general rules regarding the application of controlled vocabulary terms and much more individual rules, depending on the disease, intervention or area of interest. These rules include the use of validated filters as guidance for optimized topic-related quality filtering in a ‘bottom-ceiling-approach’.