Using the following three-part quesion as an example, we explored with what search terms this paper would and would not have been found.
‘Does an antifibrinolytic agent reduce mortality in trauma patients with significant haemorrhage?’
Domain: trauma patients with significant haemorrhage
Determinant: antifibrinolytic agent
In this post, I want to focus on the use of thesaurus terms when doing a literature search. Let’s have a look which thesaurus terms are attached to this paper, taking MeSH (in PubMed) as an example.
Domain: wounds and injuries, hemorrhage
- wounds and injuries is a very broad MeSH-term, encompassing, among a host of other types of injuries, various forms of trauma. There is no single MeSH-term for trauma.
- Note that while it will limit the amount of references retrieved, it is not strictly necessary to include subheadings (e.g. wounds and injuries/complications) in a MeSH-search. In fact, it can cause you to miss relevant papers that have not been assigned that particular subheading (either on purpose or by mistake).
Determinant: tranexamic acid, antifibrinolytic agents
- When the action of a drug is being discussed, both the specific drugMeSH term (here: tranexamic acid) and the appropriate pharmacological action MeSH-term (here: antifibrinolytic agents) will be assigned to the article.
- If the pharmacological action MeSH-term is searched as [Pharmacological Action] or [pa] instead of [Mesh], all substances with that particular pharmacological action will automatically be included in the search
- Mortality is used here as a subheading. To search for a subheading irrespective of the MeSH-term to which is is attached, use mortality[subheading] or mortality[sh]
- Using the MeSH-terms mortality, death or fatal outcome would not have yielded this article. In the MeSH Database, the use of each MeSH-term is explained.
A word on outcomes:
In this case, I included one the primary outcome of the study in my three-part question. Indeed, this outcome (mortality) was mentioned in both title and abstract of the article, and was ‘awarded’ a MeSH-subheading. Often though, this is not the case with all outcome parameters, both primary and secondary.
For instance, in title and abstract of this paper, the following outcomes are decribed:
- mortality/death with the following categories: bleeding, vascular occlusion (myocardial infarction, stroke and pulmonary embolism), multiorgan failure, head injury and other
- vascular occlusive events (i.e. not necessarily fatal), blood transfusion
Some, but not all of the outcomes mentioned in title/abstract have been awarded a MeSH-term or subheading. In addition, additional outcomes (e.g. dependency) were studied that were not mentioned at all in title/abstract or MeSH-headings (Note the ‘and other’ in the abstract!). This often happens when either the outcome in question is not the main focus of the study, or when there is no significant effect of the treatment on that particular outcome…
The take-home message is that, for therapeutic questions, it can be a good idea to limit your search to domain and determinant only, and not to include the outcome(s). This will ensure that no papers are missed which contain (negative) data on an outcome, but in which that outcome is neither indexed with thesaurus terms nor mentioned in title/abstract.