Friday 22 December 2017

Finding case reports

I searched PubMed for a rare topic and found 361 references.

Filtering to “case reports" left 197.

Of the 164 not tagged in this way, some do appear to be case reports, with titles involving phrases like “two cases”, “a rare case”.  (See below for a discussion about whether reports of more than one case are a “case report”!).

Why are these 164 not tagged?   A glance at the titles suggests some are case reports.  Some of these will be PubMed Central and some very new, so therefore not indexed.   What about the rest?  The publication type "Case Report" was first used in 2008, but then applied retrospectively back to 1966, although I wonder how consistently it was applied to older material.  

Of the 361, 73 have the word "case" in the title and 20 have “cases”.   Some of those are general articles (“associated malformations… in cases of…”) but some are definitely reporting actual cases.   76 have “cases” in the title or abstract, and some of those are definitely reporting actual cases, without mentioning that in the title.

43 have “case” or “cases” in the title, but are not indexed as Case Reports.  

Some of the items mentioning “case(s)” in the title or abstract could be case control or case series, rather than a case report.   A case series is “A group or series of case reports involving patients who were given similar treatment. Reports of case series usually contain detailed information about the individual patients. This includes demographic information (for example, age, gender, ethnic origin) and information on diagnosis, treatment, response to treatment, and follow-up after treatment.”  (NCI Dictionary of Cancer Terms).   There is a discussion of how this differs from a cohort study in this Annals of Internal Medicine article (subscription needed).   A case control study involves a control group.

I couldn’t find a search filter for case reports (if you know of one, please let me know!).   For HDAS,  
(case or cases).ti,ab may be enough, although that will also find case series and case control studies, as well as things which are nothing to do with specific cases.      

To be more specific, you could try my filter suggested below, although that will miss titles like “two cases”, “a new case”, “a case of”.

Does a case report have to be a single case?   BMJ Case Reports require prospective authors to contact the office if there are more than three cases. Oxford Medical Case Reports and European Heart Journal Case Reports don’t specify, but the latter uses the CARE reporting guideline, which seems to imply a “case report” is one single case.   And certainly cases presented to our Cardiology Education Meeting are single cases.

MeSH defines “Case Reports” as “Clinical presentations that may be followed by evaluative studies that eventually lead to a diagnosis”, which could mean one case or more.

If you don’t want to use just the word case(s), here is my possible filter, for HDAS Medline, with my search topic:

1          ("ectopia cordis").ti,ab            (340)
2          "ECTOPIA CORDIS"/    (108)
3          (1 OR 2)           (355)
4          ((case OR cases) ADJ2 (report OR review)).ti,ab       (533,996)
5          ("Case Reports").pt     (1,857,467)
6          (4 OR 5)           (2,020,257)
 7         (3 AND 6)        (229)
 8         (3 AND 5)        (197)


Changing set 4 to (case or cases).ti,ab makes set 4 3,015,101 but set 7 275.

Why that topic?   Read more about Vanellope Hope Wilkins in the Leicester Mercury (she and her parents have featured in many other media outlets too).    

The HDAS search was run on 22nd December 2017 and all the links worked on that date.

No comments:

Post a Comment

Note: only a member of this blog may post a comment.