Library

Searching the literature - finding the evidence

Search Filters

Search filters (often called 'hedges') are designed to improve retrieval of clinically relevant research from health science databases. Using search filters for a particular topic or methodology may improve your search sensitivity or decrease the number of irrelevant results.

Australian Resources
CareSearch: Subject-based search filters for use in the PubMed database. These filters provide the best available evidence on a variety of topics including palliative care and advanced care planning, bereavement, heart failure, etc.

Llt.search: This search tool provides easy access to Aboriginal and Torres Strait Islander health literature in the PubMed database. 27 topic searches are available.

Flinders Filters: Flinders Filters is a specialist information retrieval research team which provides standardised and systematic subject-based searches with known levels of performance. Topics include opthamology, primary health care, stroke, contraception, etc.

International Resources
McMaster University Health Information Research Unit (US): Methodological search filters coveringcovering therapy, diagnosis, prongnosis, etiology, economics, reviews, qualitative etc. The McMaster filters are embedded in the Ovid databases at Austin Health.

ISSG (InterTASC Information Specialists' Sub Group) (UK): Search filters are grouped by study design (eg. epidemiological studies) or focus (eg. quality of life).

SIGN (Scottish Intercollegiate Guidelines Network SIGN): develops evidence based clinical practice guidelines for the National Health Service in Scotland. SIGN has devised search filters for: systematic reviews; RCTs; observational, diagnostic & economic studies; and patient issues

PubMed Clinical Queries: designed to filter one search by three clinical research areas: Clinical Study Categories, Systematic Reviews, and Medical Genetics.

Using the 'Humans' limit

When the 'humans' limit is added to a MEDLINE or Embase search it can exclude records not yet indexed or relevant records which have not had the 'human' index term applied. To ensure a thorough search, the following is recommended by Cochrane:

Medline

  • Construct your search in Medline
  • On the next search line type the phrase: exp animals/ NOT humans.sh.
  • Combine your last two lines with NOT.

For example:

....

#8 #6 OR #7 etc.

#9 exp animals/ NOT humans.sh.

#10 #8 NOT #9

Embase

If you are using the Embase database, the syntax required is:

NOT ((exp animal/ or nonhuman/) NOT exp human/)