The choice of keywords is frequently a compromise decision. Where contact with the study authors identifies unique terms, these can be added to the search, but we found no clear evidence that this leads to literature that has been missed by standard searching; on the contrary, such approaches may lead to an expansion of the search, while at the same time bringing in unique elements that may be lacking in the standard searching alone . Lonie et al.  report the addition of key terms identified from known sources of grey literature: ‘sorting directory’, ‘civil service organisation’, ‘embassy’, and ‘university’. Similarly, Malpass  made use of a search in the Social Care Online database which returned papers not previously identified through standard searching. However, we found no evidence that the addition of these terms results in studies being missed from the standard searching alone .
In the face of evidence that text mining often lacks precision when applied to qualitative data [42, 91], the authors question the benefits of text mining in the context of qualitative reviews as it is likely to exacerbate the risk of missing relevant items [41, 43]. The authors did not identify any examples of text mined qualitative reviews, which may be a reflection of early methodological development in this field. 7211a4ac4a