RESULTS
Eleven systematic reviews authors were contacted. Five systematic review
authors answered. Among them, two systematic reviews met the inclusion
criteria and authors consented to participate.
Systematic review 1 aimed at exploring the barriers and enablers
perceived by healthcare workers toward deprescribing in long-term care
facilities8. Eight bibliographic databases were
searched by the authors. Systematic review 2 aimed at identifying
guidelines for deprescribing in older patients with
diabetes9. Three bibliographic databases and 18
websites were searched by the authors.
Results for MEDLINE search strategies are available in Table 1. Twelve
articles included in systematic review 1 were indexed in MEDLINE. The
original search strategy retrieved six of twelve included articles from
10 articles retrieved in MEDLINE (sensitivity: 50.0%, precision: 2.8%,
NNR: 35). During inclusion process, authors were required to read two
articles in full text to include one relevant articles. The implemented
search strategy using the maximized sensitivity filter retrieved seven
of the twelve included articles from 389 articles in MEDLINE
(sensitivity: 58.3%, precision:1.7% and NNR: 55). During inclusion
process, authors were required to read 2.1 articles in full text to
include one. The implemented search strategy using USDeN filter
retrieved five of twelve included articles from 9 articles in MEDLINE
(sensitivity: 42%, precision: 5.1%, NNR: 20). During inclusion
process, authors were required to read 1.6 articles in full text to
include one. Four articles included in systematic review 2 were indexed
in MEDLINE. The original search strategy retrieved one of four included
articles from 1,119 articles retrieved in MEDLINE (sensitivity: 25%,
precision: 0.09%, NNR: 1,119). During inclusion process, authors were
required to read fourteen articles in full text to include one. The
implemented search strategy using the maximized sensitivity filter
retrieved one of four included articles from 454 articles in MEDLINE.
(sensitivity: 25%, precision: 0.2%, NNR: 454). During inclusion
process, authors were required to read three articles in full text to
include one. The implemented search strategy using USDeN filter
retrieved one of four included articles from 55 articles in MEDLINE
(sensitivity: 25%, precision: 1.8%, NNR: 55). During inclusion
process, authors were required to read two articles in full text to
include one.
Results for Embase search strategies are available in Table 2.
Systematic review 1 included twelve articles indexed in Embase. The
original search strategy retrieved four of twelve included articles from
97 articles retrieved in Embase. (sensitivity: 33%, precision: 4.1%
and NNR: 24). During inclusion process, authors were required to read
1.7 articles in full text to include one. The implemented search
strategy using the maximized sensitivity filter retrieved 7 of 12
included articles from 339 articles in Embase (sensitivity: 58%,
precision: 2.1%, NNR: 48). During inclusion process, authors were
required to read 2 articles in full text to include one. Among
systematic review 2 included articles, none were indexed in Embase. The
original search strategy retrieved 997 articles in Embase. None of them
were finally included. The implemented search strategy with maximized
sensitivity filter retrieved 339 articles. None of them were finally
included.