I am writing this using a free downloadable office suite due to the practice copy of Windows Office 2000 finally showing its age when run on a new computer.
Oh, the days before online licensing. It looks remarkably similar to Mr Gate’s version but has the added benefits of (a) working and (b) being free – which to me are two major selling points over its nearest rivals. If this appears unintelligible in print then I have either been proved wrong in the above or just drunk too much whilst writing it.
The way we acquire information (and tools such as this one I am using now) has changed markedly over the last decade and the volume of what is accessible has increased hugely as well. We constantly need information during the average working day as practising vets and have lots of options as to where to look for it.
We are also encouraged to practise “evidence-based medicine”. This is a phrase that has always intrigued me as I cannot really imagine what alternative we have. It is a bit like saying we have to practise “legs-based walking” or play tennis on “laws of physics-based principles”.
How else would anyone practise medicine? Who practises on “medicine with no evidence” or bases treatments on what they’ve just made up? OK, I guess to some extent we all do and all never do. So where do we get the evidence from then?
Evidence inserted into us…
Well, initially we all come pre-packed with five years worth of evidence inserted into us during vet school. And the stuff we are taught there we have to just believe (I’m simplifying here, stay with me…). So new grads have it easy for once.
What about the rest of us, where do we look for our evidence to base our practice on after the evidence we graduated with has faded in memory and become out of date? We all have our own means, and we also need to consider conflicting priorities and characteristics of where we look. Sometimes speed is the most important factor, so you ask a colleague, “I’ve got this cat in the consult room….” – no time to do major research, but then you are relying on the opinion of one person.
With a bit more time then, quality of information can be prioritised over speed of acquisition. You may, however, also want something up to date and new. Peer-reviewed papers are up to date and of high quality, but one paper does not equal the right answer: its conclusions and methods need to be tested by peers in further experiments or trials and results repeated for its conclusions to be accepted.
When dealing with clinical findings and treatments, this is especially pertinent. Academic papers also tend to be quite esoteric to us in practice, as only something new will get published in a peer-reviewed journal.
One drug company’s trial results, published in a sanitised glossy format, do not necessarily tell the whole story. Also, when looking at a paper, the experimental methods and statistical methods applied to the results need to be scrutinised. Was it a double-blinded placebo-controlled study (best), a cohort study, a case series? Were surrogate outcomes used?
Surrogate outcomes
I have recently learned about surrogate outcomes and feel the need to pass it on. Maybe most of you can spot them a mile off, but for the rest of us, surrogate outcomes are tricks used by drug companies. A surrogate outcome is, for example, when they say that drug A is meant to cause clinical outcome x.
Instead of showing you a trial where drug A causes x, they go on to say: clinical outcome x has been associated with a rise in chemical z in the body. We can show that our drug A causes a rise in chemical z, therefore it must also cause outcome x!
The rise in z is the surrogate outcome used to “prove” outcome x. They usually do this if the trials showed that drug A did not, in fact, cause clinical outcome x at all, but they did find it caused a rise in z, and as they can kind of find a link between z and x, drug A must work!
A step up from individual papers is to look for reviews of academic papers and meta-analysis. These pull together different papers, scrutinise the methods and statistical interpretation and discuss the results of them all, possibly finding some significant trends and overall results.
A Mecca for meta analysis is found in the Cochrane collaboration library where large numbers of studies in human health are brought together and put through their statistical and methodological paces. This is an amazing resource for doctors and public health professionals (although a GP friend admitted he has heard of it but never has the time to use it).
Finding new conclusions
As well as testing large numbers of large studies, it can also pull together small studies and find new conclusions that the original studies just weren’t large enough to prove. (To find it log on to pub med and there’s a link).
Textbooks are kind of reviews as well, and are usually authored by many experts and well referenced to peer-reviewed papers. The downside? They are usually a few years out of date when they go to press. This may not be altogether a bad thing – see later.
So we all probably use some or all of the above resources depending on the situation – a quiet day reading up on an obscure case or a quick exit from a consult room to check a textbook or ask a colleague. Down at the bottom of the scale of trustworthiness comes “The Internet”.
I did a quick trial on information gathered from the internet. I chose a question I knew the answer to (a good way of testing any resource, e.g. when browsing a new textbook) and asked the internet for the answer – my only criterion being that the answer had to be written by vets for vets.
Putting the question…
The question: “How do you treat uveitis in dogs?”
Firstly, straight “Google”. The answer: “The aim of treatment is to treat the uveitis, to treat the underlying condition and to reduce pain. Medications include systemic and/or topical steroids, NSAIDs, topical mydriatic and antibiotics if indicated…
“Complications in uveitis include posterior synechia between lens and iris, iris bombé, secondary glaucoma corneal oedema, cataract, retinal detachment, iris atrophy, retinal atrophy and phthisis bulbi…”, and so on.
A reasonable answer and useful, but I would be unable to trust it as I don’t know the author or the publication. If I could Google up a publication I trusted, that would be better. More of that later.
Next “Wikipedia” and “Wiki vet”. I could not find anything at all. This is probably better than misinformation though!
Then pubmed: this is an extract from the first regarding treatment, and there were many more. “To compare effects of orally administered tepoxalin, carprofen, and meloxicam for controlling aqueocentesis-induced anterior uveitis in dogs, as determined by measurement of aqueous prostaglandin E(2) (PGE(2)) concentrations. ANIMALS: 38 mixedbreed dogs. PROCEDURES: Dogs were allotted to a control group and 3 treatment groups. Dogs in the control group received no medication. Dogs in each….”
This would score highly on detail, is peer-reviewed and is a placebo-controlled trial. However, I would need to do further literature searches to validate its findings and it is obviously of no help whatsoever if I am in a hurry and want to know what to do with the dog sat in the consult room with the sore eye.
A better bet might be pubmed reviews: “Uveitis is a commonly misdiagnosed ophthalmic condition with a wide aetiological base and often the cause cannot be established. Uveitis can be associated with systemic diseases like feline infectious peritonitis, feline immunodeficiency virus, feline leukaemia virus, lymphoma, toxoplasmosis and canine ehrlichiosis. The classification and general clinical signs….”
This is looking like it may be of more interest to a practitioner and would definitely be one to read up on one evening or quiet afternoon (as long as you spotted that the review was based on South African cases and so all the infectious diseases will be different).
Veterinary Ophthalmology magazine gave me lots of articles with catchy titles like:
- Surgical management and histologic and immunohistochemical features of a cataract and retrolental plaque secondary to persistent hyperplastic tunica vasculosa lentis/persistent hyperplastic primary vitreous (PHTVL/PHPV) in a Bloodhound puppy; and the relatively sparsely entitled
- Immunohistochemical characterization of intraocular metastasis of a canine transmissible venereal tumor; and many more.
So useful for researching a case for a casebook, or for a really in-depth trawl of the literature but not the best for everyone on a day-to-day basis.
No veterinary section yet…
The Cochrane library has no veterinary section yet so here is an example of an entry for human uveitis: “This is the protocol for a review and there is no abstract. The objectives are as follows: This study will systematically review the evidence from all randomised or quasi-randomised controlled trials examining the effects of antibiotic and other interventions for the treatment of people….”
You get the idea: they’re not messing about.
And lastly, the land where textbook meets internet, Chinese-data-entrysweatshop meets the great publishing houses. A world where copyright law seems to be laughed at, a huge online library where we are free to browse and drink beer whilst doing so: “Google Books”.
A search here took me to the correct page on “Slatter’s Fundamentals…” textbook (a newer version than mine) and also “Gelat”, one of the standard texts, the “Ettinger” of the eye world. It did prohibit access to another popular textbook but allowed viewing of one recent text written by UK ophthalmologists including some of Dr Mould’s photographs. It even highlighted the search words.
The whole book was there in my kitchen to be browsed and interrogated. This has all the benefits and drawbacks of real textbooks but the advantages of being free, a huge range of books is available, and you get to feel a bit naughty whilst doing your homework as it can’t be right that you can just read them all for free?
So acquiring information is a minefield, and sometimes the “best”, e.g. an academic journal you subscribe to, is less helpful in some situations than a general search engine. The middle ground I think has to be textbooks, whether pilfered online or in real form on the shelf.
Personally, I prefer the real ones, but online access is useful for unusual subject areas. Most vets have textbooks and subscribe to journals that interest them, and online access may be useful for the occasional case you see outside your normal comfort zone.
A well-referenced textbook fulfils most of the right criteria in terms of quality of information, except for being right up to date.
However, the novel is not always the best, and as can be seen from the excerpts above, hot-off-the-press journals aren’t actually that useful for us in practice, or intended to be. We are lucky in the veterinary world to have many peer-reviewed clinical journals that can fill the time gap between textbook and the current trends.
Despite having written over 2,000 words [sorry ed.] on the subject of getting the right information, I am by no means an expert. And that is the point.
At the sharp end
I am however, like most of you reading this, the one in the consult room, or in the lab. The one at the sharp end who may have been well trained in clinical matters and very experienced, but when it comes to knowing your cohort from your case controlled, what a p value really is, etc., we are not so good.
If I read a good academic paper on a successful novel treatment, am I the one to decide that it should be used in practice? In some cases I may be, but generally it might be better to let it filter through to the clinical journals, or even a textbook.
We need a way of seeing the wood for the trees, and we need the skills to be able to filter through the vast sea of information out there. Until someone sets up a veterinary Cochrane, we will have to carry on making our own judgements.
