On 2015 Jan 24, Wichor Bramer commented:

I think this article is a rathor open door article, in that its conclusions can be drawn beforehand, wihout executing the experiments. If you consider PICOS as PICO + S, surely then of course PICO will retrieve more relevant articles and PICOS sarches will be more specific. The searches used for testing PICOS were the exact searches for PICO, with an added element of S, which in this case was very basic (two or three tersm), compared to the exhaustive translation of the other elements (more than 10 terms). PICO, PICOS and SPIDER should not be used in the creation of search strategies. Not every element in those methods is necessary in a search query. Including specific outcomes and controls can introduce bias in the search results. For a good thorough systematic review, only P and I are used, and when necessary combined with a sensitive filter on study design, that consists of more than three terms and is verified (when possible). PICO, PICOS ad SPIDER can be useful in the process of evaluating the retrieved full text references for inclusion, but should not direct the search strategy.

On 2015 Jan 24, Wichor Bramer commented:

I think this article is a rathor open door article, in that its conclusions can be drawn beforehand, wihout executing the experiments. If you consider PICOS as PICO + S, surely then of course PICO will retrieve more relevant articles and PICOS sarches will be more specific. The searches used for testing PICOS were the exact searches for PICO, with an added element of S, which in this case was very basic (two or three tersm), compared to the exhaustive translation of the other elements (more than 10 terms). PICO, PICOS and SPIDER should not be used in the creation of search strategies. Not every element in those methods is necessary in a search query. Including specific outcomes and controls can introduce bias in the search results. For a good thorough systematic review, only P and I are used, and when necessary combined with a sensitive filter on study design, that consists of more than three terms and is verified (when possible). PICO, PICOS ad SPIDER can be useful in the process of evaluating the retrieved full text references for inclusion, but should not direct the search strategy.