About
 Introduction
 Search Concepts
 PubMed
 Ovid
spacer.gif (43 bytes) spacer.gif (43 bytes)

backspacer.gif (43 bytes)next

Backspacer.gif (43 bytes)Next

spacer.gif (43 bytes)
backspacer.gif (43 bytes)next
Backspacer.gif (43 bytes)Next

 spacer.gif (43 bytes)

spacer.gif (43 bytes)spubmed.gif (5318 bytes) backspacer.gif (43 bytes)next

PubMed Advanced Interface - V. Running the Systematic Search

To run the systematic search you will need to copy one from the tables below to Clipboard and paste it into the PubMed search box. Return to the Preview/Index page and clear the top search box. If you want to use the Haynes Therapy (specific) filter you would then copy and paste (double [WORD] AND blind* [WORD]) OR placebo [WORD] into the top box and press Preview.

Table 1. Systematic Searches
Category Search String (from Haynes et al.) Results
Therapy sensitivity "randomized controlled trial" [PTYP] OR "drug therapy" [SH] OR "therapeutic use" [SH:NOEXP] OR "random*" [WORD] 1196311
Therapy specificity (double [WORD] AND blind* [WORD]) OR placebo [WORD] 98649
Diagnosis sensitivity "sensitivity and specificity" [MESH] OR "sensitivity" [WORD] OR "diagnosis" [SH] OR "diagnostic use" [SH] OR "specificity" [WORD] 2640389
Diagnosis specificity "sensitivity and specificity" [MESH] OR ( "predictive" [WORD] AND "value*" [WORD]) 102645
Aetiology sensitivity "cohort studies" [MESH] OR "risk" [MESH] OR ("odds" [WORD] AND "ratio*" [WORD]) OR ("relative" [WORD] AND "risk" [WORD]) OR ("case" [WORD] AND "control*" [WORD]) 628689
Aetiology specificity "case-control studies" [MH:NOEXP] OR "cohort studies" [MH:NOEXP] 56309
Prognosis sensitivity "incidence" [MESH] OR "mortality" [MESH] OR "follow-up studies" [MESH] OR "mortality" [SH] OR prognos* [WORD] OR predict* [WORD] OR course [WORD] 1004204
Prognosis specificity prognosis [MH:NOEXP] OR "survival analysis" [MH:NOEXP] 178159

 

Table 2. Other Useful Searches
Category Search String Comment
REVIEWS ("systematic" [WORD] AND "review*" [WORD]) OR ("data" [WORD] AND "synthesis" [WORD]) OR ("published" [WORD] AND "studies" [WORD]) OR ("data" [WORD] AND "extraction" [WORD]) OR "meta-analysis" [MH:NOEXP] OR "meta-analysis" [TI] Mid-range sensitivity/specificity search as recommended by NHS CRD. Note that PubMed has no ADJ operator. See also article in ACP for other search strategies for reviews.
IRRELEVANT ("comment" [PTYP] OR "letter" [PTYP] OR "editorial" [PTYP]) OR ("animal" [MH:NOEXP] NOT ("animal" [MH:NOEXP] and "human" [MH:NOEXP])) Exclude irrelevant hits such as animal studies or letters by combining with other search using the NOT operator. e.g. REVIEWS NOT IRRELEVANT
GUIDELINES guidelines [MESH] OR guideline [PTYP] OR "practice guideline" [PTYP] OR "health planning guidelines" [MESH] OR guideline* [WORD] OR recommend* [WORD] OR consensus [WORD] This search is based on recommendation from the Canadian Centres for Health Evidence

 

Combining the Subject and Systematic Searches

Combine your subject search final line with your systematic search using syntax such as #1 AND #2 in the search box. Note that, for PubMed, the Boolean operator must be in capitals (Ovid doesn't mind upper or lower case) and the search numbers must be preceded by the hash character #. Press the "Go" button to display the search results.

Tidying Up

At this stage you can tidy up your search. I often run my 'Irrelevant' search and combine it using 'NOT' to exclude letters, comments and editorials, and animal studies. In addition, you can set the date range, and specify English language if you wish, using the Limits screen, which is fairly intuitive.

Summary

Going through the steps above has taken some time, but once you are familiar with the screens you can usually complete a search in a couple of minutes. The preliminary search strategy is to: -

  1. Type in search string, which PubMed automatically maps to a MeSH (exploded) and textword search, and Index it. This is the most sensitive subject search.
  2. Execute the relevant systematic search, and Index it.
  3. Combine the subject search and the systematic search on the Preview/Index screen, using Boolean AND.
  4. If you wish, add limits such as language and date.

If you find too many results you can increase the specificity of the subject search by using MAJR (focus), and by selecting only some of the floating sub-headings for your subject. You can also use a more specific systematic search.