Considering that all authors (whether main authors or collaborating) are affiliated to particular institutions in various countries, we can leverage this to extract papers with Kenya in the author affiliation.
$ esearch -db pubmed -query "Kenya[Affiliation]" | \
efetch -format abstract > kenpaps.txt
grep "PMID:" kenpaps.txt | cut -f2 -d' ' > kenpmids.txt
#| xargs | sed 's/ /,/g' > Kenpmids.txt
You can uncomment the line | xargs | sed 's/ /,/g' > Kenpmids.txt
and pipe the previous
line to it in order to convert the list of IDs into an efetch compartible format
(e.g. 26895239,26894324,26893967,26892879)
If we are only interested in the abstract of all papers with Kenyan authors, then the above code that returns the abstract should be enough and no need to fetch the PMIDs.
And from the line count in the kenpmids.txt file, we can see there are 13473 PMIDs relating to the same amount of papers with Kenyan authors.
Also attached is a file of all the abstracts for the 13473 papers.
The search strategy that returned the over 25,000 papers was based on searching Kenya as a search term.
That is
esearch -db pubmed -query "Kenya"
This strategy searches Kenya in all fields and medical subject headings of the PubMed
records and returns the combined results.
That is
"kenya"[MeSH Terms] OR "kenya"[All Fields]
Hence, we would not have accomplished our goal to fetch only papers that Kenyans are invovled in as authors.
NB:
The search strategy [Open science] AND Kenya
translates to
(Open[All Fields] AND ("science"[MeSH Terms] OR
"science"[All Fields])) AND ("kenya"[MeSH Terms]
OR "kenya"[All Fields])
And it returns only 140 records.