Thursday, July 23, 2015

Post Ebola syndrome, and choice of search terms

On the BBC news recently was an interview with a survivor of Ebola, which made reference to "post Ebola syndrome".  A search of the web finds a lot of news items mentioning this term, a short Wikipedia page and an article from Medecins Sans Frontieres.  According to that MSF article, joint pain, eye problems and depression, among other things, can follow Ebola.

Another item found by a web search for "post Ebola syndrome" is this blog post on the H5N1 blog. It refers to an editorial in Lancet Infectious Diseases (the link in the blog post goes to the Lancet but here is the PubMed record):

Bausch DG.
Lancet Infect Dis. 2015 Apr 21. pii: S1473-3099(15)70165-9. doi: 10.1016/S1473-3099(15)70165-9.
PMID: 25910638
A search of PubMed for post ebola syndrome finds only 6 items, which are perhaps not the most relevant.  A search for ebola sequelae, taking the hint from this editorial, finds 89.   Among them is this, which is mentioned in that blog post:

Clark DV, Kibuuka H, Millard M, Wakabi S, Lukwago L, Taylor A, Eller MA, Eller LA, Michael NL, Honko AN, Olinger GG Jr, Schoepp RJ, Hepburn MJ, Hensley LE, Robb ML.
Lancet Infect Dis. 2015 Apr 21. pii: S1473-3099(15)70152-0. doi: 10.1016/S1473-3099(15)70152-0.
PMID: 25910637
PubMed's Automatic Term Mapping maps post ebola syndrome to both relevant MeSH terms, virus and disease.  Searching for ebola sequelae does the same, but also to the subheading Complications.   This can be applied to the disease heading (Hemorrhagic Fever, Ebola), and effectively is a floating subheading with the virus heading (Ebolavirus).  Adding complications to the search (ebola (complications OR sequelae)) finds 89 as well.

However, using MeSH alone would find neither item from Lancet Infectious Diseases, as neither has been indexed yet. 
 
So, the literature searching lessons here - try alternative terms, try to identify the ones that are used in the medical literature and not just ones used in the general media, and try to find terms that make best use of PubMed's Automatic Term Mapping.  

No comments: