Friday, April 21, 2017

Search strategy for chronic kidney disease

Chronic kidney disease?  Chronic kidney failure?   Chronic renal failure?

There are many ways to describe this (1).  After a few searches on the topic, I tried to come up with a Medline strategy.  I think it may include some conditions that Hsu and Chertow would not, but I have this:

1.            ((endstage or “end stage” or established or chronic or progressive) adj1 (renal or kidney) adj1 (failure or disease* or insufficienc*)).ti,ab
2.            (Chronic adj1 nephropath*).ti,ab
3.            (“Chronic uremia” or “chronic uraemia”).ti,ab
4.            (CKD or CKF or CKI or CRD or CRF or CRI).ti,ab
5.            (ESKD or ESRD or ESRF).ti,ab
6.            kidney diseases/ and chronic.ti,ab
7.            exp Renal insufficiency, chronic/ 
8.            Renal insufficiency/ AND chronic.ti,ab

9.            1 or 2 or 3 or 4 or 5 or 6 or 7 or 8

This is for Medline via HDAS.   

I started with terms I had used previously, and added terms used by colleagues at UHL, MeSH Used For, and terms used in strategies in First Consult, SIGN guideline 103 and NICE CG182.  I then asked people on lis-medical, and also had some further suggestions when I circulated the strategy there.

This, for the record, is the version dated 22nd June 2016 with tiny minor amendments made in April 2017 to remove some unnecessary full stops! 

(1) Hsu CY, Chertow GMChronic renal confusion: insufficiency, failure, dysfunction, or disease. Am J Kidney Dis. 2000 Aug;36(2):415-8.

Guidelines for comparing guidelines

This was asked at a writing workshop that  I led recently.   How do you go about comparing two guidelines?

Critical appraisal

Comparing is related to,  but not the same as, critically appraising a guideline, in which you are looking at guidelines individually.   Trisha Greenhalgh's very useful book How to read a paper has a chapter about appraising guidelines, with ten questions to ask about a clinical guideline.  There is also a systematic review of appraisal tools in this article in PLoS One by Ulrich Siering and colleagues.


The National Guidelines Clearinghouse has a "Compare" tool.    You choose the guidelines you want to compare, by ticking a box in the search results list.  Then you choose which sections you want to compare, and you can then see on screen or in a spreadsheet, those sections side by side.   It takes the sections as they are, which means that you may be comparing very differently arranged texts.

A PubMed search for comparison clinical guidelines finds some papers that compare guidelines for specific conditions.  It also finds many papers about comparing other things, not comparing actual guidelines.  The search worked much better once I had corrected comparision to comparison although the wrong spelling still found a handful of things!

Missing out the word clinical, so comparison guidelines, finds more (and more irrelevant things too, of course).

Searching more specifically, for example comparison guidelines valvular heart disease, would of course be a way forward.

A quick look at MeSH headings for something old enough to have them suggests:

Practice Guidelines as Topic/ 
Comparative Study/.  

Searching those headings (type in practice guidelines as topic comparative study) together finds a lot of relevant looking papers that I had not seen before, (as they did not use the word "comparison"), amongst other papers comparing decisions or results related to recommendations in guidelines.    Combine this search with a clinical area or subject.

Reporting guidelines

Comparing is related to reporting guidelines too.  The AGREE Instrument helps assess the methodology of a guideline.  This is in the Equator Network Library along with the RIGHT Statement for reporting practice guidelines and some other things.

Thursday, April 06, 2017


One of the things about being married to a minister is that I sometimes end up helping with resources.    And so it was that I was asked how many tears someone would cry in a lifetime, for use as an illustration in a service.

This turns out to be a difficult question.  If you just count tears of sadness, then of course it depends on how often you are sad, and how sad, as well as how long you live.   But I ended up with a sort of answer, and although it was on the internet, it wasn’t found through a regular internet search.

I started with a regular internet search, for something like “how many tears are cried?”.  That brought up a lot of news stories, which give widely different figures, with no sources.

The chain of research then went something like this:

I found something from the Huffington Post, , and something from How Stuff Works.

These alerted me to the three types of tear (possibly a medical textbook would have told me that too).  Equipped with alternative search terms, I headed for PubMed.  I found this, searching for basal tear flow.

I then went back to Google, but Google Scholar.   I searched for basal tears volume, and along with the Farris article cited above, I found this from the British Journal of Ophthalmology from 1953.  That cites this 1903 paper by Schirmer.  I had no access to the full text of this but looking for it in Google finds this related article by De Roetth from the AMA Archives of Ophthalmology (related because it also cites Schirmer).     

De Roetth cites Schirmer but also mentions that Schirmer is citing work by Ahlstrom (also a related article to Schirmer in Google Scholar - 1895. Ahlström, Über die antiseptische Wirkung der Tränen. Zentralbl. f. Augenheilk. S. 193).   

I had no access to Ahlstrom either but de Roetth gives Alhstrom's figure (as well as describing the limitation of Ahlstrom’s method, which might mean the figure is too high).   The figure was 4 grams of basal tears a day.   De Roetth was not available in full online, but the first page, with this information, was.

I then made an assumption that 1g of tears was 1ml in volume, which was, I think, quite close, and another assumption about how long our hypothetical person would live.   Taking that as 80 years gave 4 x 365 x 80 ml of tears, which is 116.8 litres of tears in a lifetime.  

So, an answer for illustrative purposes.
And not a new answer.   A figure from 1895, cited in 1903, with that paper cited in 1953.  Although it did come up in an internet search, it was not found straight  away. 
Having read this far, you deserve this tear related song from my youth by The Beat, and then this one by Tears for Fears, with a video set in a library.