As part of Paediatric Training it is expected that we will become proficient at using Evidence Based Medicine to guide our decision-making and therefore improve clinical care. A useful exercise to develop these skills is the completion of the Clinical Question section on Kaizen.
Have a look at the boxes for an examples of how you may want to present your piece and for the key requirements.
Forming a question:
The topic examined could be anything in the realm of medicine, but is particularly useful if is based on an area of personal interest or is driven by a particular clinical encounter.
It can be helpful to use the PICO format. The structure is simple, but it will help you to really reduce down the clinical topic to an answerable question. The PICO structure identifies a Patient (or Population), and Intervention, something to Compare (or a Control), and an Outcome. The more specific your description of the Population, the more likely it is you will be able to apply the findings to your own target population. Your clinical question may not precisely fit into the PICO format; consider using another framework if needed, (you may want to read about the use of SPICE and SPIDER as other frameworks).
It is best demonstrated using an example written here by Marita Macken and Ian Wacogne:
Example Forming a Question: Antibiotics and CSF
One scenario often faced is that of a potentially septic or meningitic child being admitted and commenced on antibiotics. I have frequently been in the position where I have had to review lumbar puncture results from a specimen taken after the administration of antibiotics. Clearly we do not want to miss a significant diagnosis of meningitis so it is important to know how CSF specimens are affected by antibiotics. We can now form our PICO question:
P – Patient Population including infants, children and adolescents.
I – How does the Intervention of antibiotic pre-treatment…
C – Compare to no antibiotic pre-treatment, affect…
O – the Outcome of the diagnostic value of cerebrospinal fluid.
Write your PICO question as clearly as possible with a specific aim. Also write about why this question is of interest to you and why you have decided to perform the search.
Searching for the evidence:
Trying to sift through the massive amount of evidence to find the answers you want can be intimidating. Having identified your precise question, you will want to limit your search to relevant documents. Many hospitals have comprehensive libraries with very experienced medical librarians you can take you through the stages of a literature. Consider booking a session during with a librarian to help develop your skills.
Example Evidence Search: Antibiotics and CSF
A search of MEDLINE and EMBASE was performed using search terms (adolescent OR child OR infant OR paediatric OR pediatric) AND (meningitis) AND (antibiotic OR antibiotic agent OR antibiotic therapy or anti-bacterial agents) AND (empiric OR pre treat OR pretreat OR time interval OR time to treat) AND (cerebrospinal fluid OR CSF OR lumbar puncture OR spinal fluid).
This identified 253 articles in English. Four articles were reviewed in depth.
Describe how you searched for your evidence in a similar way to above and give full references for the evidence you have reviewed. Select a minimum of four supporting pieces of evidence.
NHS Evidence: www.evidence.nhs.uk
This is a simple and intuitive system to use that can be a good place to start. It has more bespoke features for those that want to use them and is free to use for those that have an NHS Athens account (which is also free to create!)
You can perform a simple search by simply typing the question into the box (marked green). This will provide you with a selection of suggested resources including up to date guidelines.
Alternatively you can perform a more precise search by clicking the Journals and Databases (marked in red).
Click “Begin a search” within the Healthcare databases advanced search (HDAS) box.
Click “Login to OpenAthens” in the top right corner
You can search the huge databases of Medline, Pubmed and EMBASE amongst others.
Boolean search terms can be used (such as AND, OR and NOT).
Click the “Help” button to see a comprehensive guide for how to use the system including some of the really useful features such as combining searches.
You should look for at least four references to complete your Clinical Question. This should not include an existing publication that has reviewed evidence, e.g. a Cochrane review, as you will be reviewing the individual paper yourself!
Reviewing the evidence:
A complete discussion on how to comprehensively review evidence is outside the scope of this piece, but there are some excellent resources available to use. The skills practiced at this stage are vital for being able to sift through the huge amount of evidence out there to best decide what is appropriate to base your decisions on. A paper review is also now a part of the START assessment required by the Royal College.
Example Evidence Review: Antibiotics and CSF
Paper 1, Nigrovic et al.
245 children with confirmed or probable bacterial meningitis were identified presenting over a 3.5 year period. This was a retrospective cohort study and was published in 2008. 35% received antibiotic pretreatment within 72 hours of an LP and found a lower rate of CSF culture positivity with pretreatment. There was no significant association with CSF WCC or neutrophils or Gram stain. The exclusion criteria were appropriate but 14 had missing information about the duration of antibiotic pretreatment. There was no information on organisms identified on bacterial cultures.
Summarise each of your selected evidence in a similar way highlighting the important findings, the strength of the evidence and conclusions you draw from the paper. Use a CASP system (or similar) to appraise the evidence, don’t simply copy the abstract!
The Centre for Evidence-Based Medicine is based at the University of Oxford and provides training and qualifications in a variety of areas of Evidence-Based Medicine and holds a number of excellent resources for open use.
The tools found here have worksheets to guide your assessment of Systematic Reviews, Diagnostics, Prognosis and Randomised Controlled Trials.
The Critical Appraisal Skills Programme (CASP) was developed to help develop skills in Evidence-Based Medicine. The website contains some excellent features including a guide on how to search for evidence, plus tools to aid critical appraisal.
Forming a conclusion:
Combining evidence from a number of resources is a specialist skill and is one that requires particular training, such as if performing a Cochrane Review.
Evidence-Based Medicine in action:
Many publications have sections specifically utilising Evidence-Based Medicine. These great resources are useful for personal development, staying up to date with the latest evidence and as examples of how to construct and answer a Clinical Question.
Archives of Disease in Childhood publishes an Education and Practice Edition six times per year and is specifically aimed at developing paediatricians at all levels of training. You will find the ‘Picket’ section containing concise reviews of sentinel articles and gives you some ideas on how to review and summarise papers. The ‘Archimedes’ section addresses clinical questions in a way similar to what is expected in the Clinical Question write up you will be doing. Make sure you have a look at these pieces for inspiration and guidance. Also, you may want to consider if any of the Clinical Questions you develop are appropriate for publication. If you have an idea, have a chat with your supervisor or contact Dr Ian Wacogne, Edition Editor, Education and Practice, Archives of Disease in Childhood. (firstname.lastname@example.org)
You should complete two Clinical Questions per training year
Read the blue boxes above for examples and the key requirements
Try to separate your piece into the key sections of Question, the Evidence Search, the Evidence Review, and the Conclusion.
- Do antibiotics affect cerebrospinal fluid results? Macken, M. Wacogne, I. Arch Dis Child 2017; 102:990-993.
- Methley A, Campbell S, Chew-Graham C, McNally R, Cheraghi-Sohi S. PICO, PICOS and SPIDER: A comparison study of specificity and sensitivity in three search tools for qualitative systematic reviews. BMC health services research. 2014 14:579. 10.1186/s12913-014-0579-0.
Example Conclusion: Antibiotics and CSF
Conclusion after review of four papers:
The usefulness of Gram stain and culture of CSF is affected by pretreatment after only 1-2 hours of antibiotic for some bacteria. If there is a delay, PCR may be helpful but since it does not provide antibiotic sensitivity it is not a routine substitute for prompt, clinically safe LP. It is worth doing a LP up to 96 hours after antibiotic treatment to exclude or confirm meningitis by assessment of the CSF WCC. A ‘bloody tap’ may make the CSF WCC uninterpretable and is best avoided if at all possible.
Aim to form a consensus opinion after carefully reading all the appropriate evidence you have gathered.