Presenting Evidence on Clinical Topics

How to do a presentation on EBD and Endodontics

A former student contacted me yesterday to ask if I could help provide some guidance on how to go about doing a presentation on EBD and endodontics. Where should he start, he asked and how should he go about it?

So I thought I'd blog a response in case anyone else has to do an EBD presentation on some aspect of dental care.

A little bit like doing endo itself, preparation is key to a successful outcome. So it is with a presentation on EBD. We may only have 10 or 20 minutes to present what we have found but probably several hours will be needed to deliver something that is informative and, importantly, backed by the most up to date research.

My first question back to my new colleague is how broad he would want to be in addressing the topic. Simply being asked to "present on EBD and Endo" unsurprisingly caused him to panic a little. The field is huge so what would he present on?

Questions

So, as with a research topic or any literature search, developing a clear idea of a clinical question to present evidence on would be my first step. Using the PICO structure (see a blog explaining this here and a prezi here) I would think about whether I was interested in a question about:

• diagnosis (e.g. how sensitive are tests for non-vitality?) 
• prevention (e.g. how effective is partial caries removal compared to full caries removal in preventing irreversible pulpal damage?)
• treatment or interventions (e.g. is one stage endo better than two stage?)
• prognosis (e.g. what is the success of re-treatment over 10 years?)
• patient or practitioner experience (e.g. how did clinicians get on with using a particular technique for obturation?)

Sometimes it takes a while to decide how broad / narrow you want to be and sometimes only after you have begun to 'scope' the literature do you get a sense of how much research there is likely to be to help answer your question.

At this stage it is also worth thinking about the best type of study or studies to answer the question. For non-complex interventions a systematic review of randomised controlled trials or the trials themselves may be most appropriate. For a prognosis question a cohort study that follows patients with a particular condition over a period of time could be suitable (or indeed one arm of a controlled trial). For a question about experience and values a qualitative study design could be best. The point is that we shouldn't concentrate only on RCTs when looking for evidence as they are not always the only or best way to answer certain questions. You can find some guidance on the best types of studies to answer questions here.

Search

The next stage, then, is to look for the research evidence. There is a growing recognition that we need to get better at recognising and being critical of non-research evidence too - particularly our own experience and the views of our patients - and of combining these in an optimal way (we're still working on it...). 

But we need to search the various medical databases and search engines to find the research evidence first. An efficient way to do this is to look first of all for summaries of evidence, such as guidelines, and systematic reviews. If we can find one that is up to date and relevant to the question we asked then we need not look for primary studies. So I would recommend beginning with the Cochrane Library for reviews or the National Guideline Clearinghouse for guidelines. You might search the EBD Journal website too to see if there are any commentaries on research there or the ADA Center for EBD.

Given the time - often a couple of years - to complete a review or guideline, anything that is more than a couple of years old is probably out of date since the most up to date research they include may by then be 4 years old. So if the review seems old, irrelevant or there simply isn't one, then we need to look for primary studies. PubMed is an open access medical database that allows this. There are a couple of helpful introduction videos by my colleagues at Oxford here and here.

As you become more familiar with PubMed you can limit the number of results you use by using filters for systematic reviews or randomised controlled trials. A video explains about this here. The advantage is that we can cut the number of articles we have to look through from hundreds or thousands to maybe dozens or less.

Of course, you could also ask the excellent library staff at the British Dental Association to do a search for you. As with your own search it's best to have a  clear question to give them or they may end up searching for things you're not interested in. This service is free to members of the association. The BDA also houses collections of papers on over 500 topics at their London site and these can be posted out to members at no charge.

Accessing the research you find

One of the biggest problems we face is that much of the research out there sits behind a pay wall and few of us are willing to fork out $25 to read a paper that may be irrelevant or of poor quality. I have blogged here about this problem. Again, the BDA can help out but at a cost of £2.50 per article. Unfortunately, my experience with Athens is that this provides minimal access to relevant journals. This is why up to date Cochrane Reviews are so valuable to us as they're free to anyone in the UK and several other countries (see here if you're not sure).

Get critical

Not all research is equal in terms of its validity. If you manage to find a systematic review there should be an indication of the quality of the primary studies included. There are various schemes for this and Cochrane now use GRADE criteria that assesses the research to be of very low to high quality. But if you're reading the primary studies yourself a checklist such as those produced by the CASP organisation are helpful to quickly get a sense of the methodological quality of a study and its usefulness to you.

My personal view is that any CPD presentation ought to indicate the quality of the evidence being presented. Normally on a course there is a mix of personal and research experience and I think that we are entitled to know which is being used. Likewise, if we are to do a presentation of the evidence-base for different topics around endo then the audience should be given a summary of how strong the evidence is. After all, why go through the cost and time of changing one's practice if the only research suggesting you should is of a very low quality? The quality of the evidence should determine whether or not we consider implementing it.

Implementation

One of the areas in EBD that is most complex is how to go about implementing change based on high quality evidence. It is recognised that most of us transform research findings rather than implementing them as reported in the research. I think that in a presentation it would be helpful to discuss what the barriers are to changing practice and how we might go about reducing these. Perhaps we need to compromise on some element of the protocols suggested by the research to make it practical and cost-effective in our practice. Perhaps we need to think about forming a group to keep each other motivated as we seek to change practice as most of us are very poor at changing what we do on our own. 

I would include these in any presentation as EBD is useless unless this important step is achieved.

I won't talk about presentation skills here - there are many much more gifted in those than me but I take inspiration from Steve Jobs who rarely used script and stuck to simple messages with plenty of graphics to enthuse the Apple-lovers out there.

Happy EBD presenting :)

A PICO Venn Diagram

PICO as a Venn Diagram

I created this Venn diagram to demonstrate how the different components of PICO overlap. As we look for studies we aim to find those that have as many of the components we are interested in as a possible. 

In this Prezi you can control where you go by clicking on different overlapping parts.

A PICO lecture - helping a customer buy the "right" phone

I have uploaded a lecture I recorded on using PICOs to help turn problems into structured questions.

It is aimed at 1st year undergraduates and so I have used instead of a clinical problem one that involves someone buying a phone. I hope it works...

Please take a look at it here.

Feedback always welcome

TRIP and its PICO

TRIP now includes PICO search

TRIP Database has just been redesigned and now incorporates a PICO search box alongside the google-style one. I've blogged about PICOs before and it's great to see a major search engine innovating with them. There's a screencast here http://www.screenr.com/c9y8

One suggestion, though - I would avoid trying to fill each of the 4 boxes. It's likely you'll end up with very few papers if you do. When I begin a search I usually start with search terms around just two concepts - say the problem / population and the intervention or intervention and comparison. If I find then that there are loads of hits with irrelevant outcomes, say, I may then add terms for the outcome of interest to me. But this is rare.

DEBTs in the EBD journal

You may be aware of papers in the EBD journal called DEBTs - or Dental Evidence-based Topics.

The idea is that you as a clinician or student will have a clinical question. You might wonder if you should be putting adhesive under fissure sealants or not, how best to treat caries in deciduous teeth or how long a posterior resin-bonded bridge is likely to last.

In order to arrive at a (hopefully) more informed position you then search for research evidence that could help using the PICO structure to guide you as discussed elsewhere in this blog.

It's important to think about what research would best inform your decision - it may not always be randomised controlled trials.

Having found what evidence there is, you then need to critically-appraise it and summarise whether the evidence you have found is valid internally and externally, and how it informs you in trying to answer your clinical question.

If well conducted the journal will publish these DEBTs because they bring the latest level of clinical evidence to the notice of other clinicians, who may well be wondering about the same clinical problem.

Any students who would like to have a go and get a publication to their name, feel free to drop me a note.

Dental Clinical Education Prescription - Presentation

This blog relates to my introduction to Dental Clinical Educational Prescription

PICO questions

Since I referenced PICOs in my last blog I thought I'd mention them here as they offer a great structure to help us find evidence for intervention studies. It's all about asking a clear and concise question so that you can - hopefully - find a clear answer.

The P bit stands for the patient or problem you are concerned about.

Let's imagine a 50 year old lady who had a load of amalgams placed mainly when she was young. She's read that dentists crown patients' teeth when they've got big fillings in because it stops them from breaking and losing their teeth.


What's the P here? The P is a problem in a patient - in this case a patient with heavily restored teeth.

The I bit relates to the intervention. In this case the I is a crown (or an onlay). You want to know, as this lady's dentist, whether or not you should crown some of her teeth and - if you're going to - which ones. So you want to compare placing a crown against not placing a crown to know if the crown is better.

So C is your comparison or control. In this case that is a filling - either amalgam or composite. If all we looked at was the survival time of a crown, without comparing it to the filling, we wouldn't know whether the filling might last just as long - or even longer.

Finally O is the outcome or outcomes you're interested in. What matters to the patient in this case? She is probably worried about losing her teeth. So the bottom line will be - does placing a crown mean that she'll keep that tooth longer than if she had just left it with a filling?

Another outcome might also be which restoration has to be replaced most frequently, or what the long term cost is of each treatment. And an outcome that is often forgotten is whether there are any adverse effects: does the tooth die, does it fracture catastrophically so it has to be extracted, does one cause mouth cancer?

We usually decide on one main outcome that really interests us. In this case I'd go for whether or not the tooth was retained longer as without this information it's very difficult to decide whether one is more cost-effective, for example.

So the PICO question here would be:

For an adult patient with an MOD or larger restoration in a molar tooth (P) does a crown or onlay (I) rather than a composite or amalgam (C) result in the tooth being retained longer (O)?

So why do this? Well, if you're going to search for evidence one way or the other - even if it means asking an expert (even though they lie way down the bottom on the hierarchies of evidence) - it is important to have a clear question. From this you can create a search strategy to use with Medline or the Cochrane Library and in a short time you can identify any good papers that might help you answer your question.

For more on this have a look at the Centre for Evidence Based Dentistry website