Article Text

Download PDFPDF

Towards evidence-based medicine for paediatricians
  1. Bob Phillips
  1. Centre for Reviews and Dissemination, University of York Alcuin College, York, UK
  1. Correspondence to Dr Bob Phillips, Centre for Reviews and Dissemination, University of York Alcuin College, York, UK; bob.phillips{at}

Statistics from

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.

What could it be?

It is often the case we would like to know what the chances of different types of diagnosis are, or the chance of a scan finding something or some things with a presentation. These types of questions can be answered in an ‘evidence-based’ way, but they rarely are even thought about let alone addressed.

Their appraisal should follow the same model as others; does the paper produce valid findings? Are they meaningful and important? How far is the paper from the population or patient I am thinking of—too far to be helpful? Validity is best gained from an appropriate population, evaluated with fair diagnostic standards and if subgroups proposed, done without data dredging or microblading. Importance is partly about the certainty any ranking or proportions can get—these relate largely to numbers—and how if there were ‘wobbles’ in the ordering this might make a difference. The particularisation (not ‘generalisability’) of your patient or population to the paper will ask the usual things about time, presentation, elements of age, ethnicity and comorbidities that you would instinctively think of.

This can be valuable for thinking through the value of particular ‘diagnostic’ (but not ‘gold’ standard) tests, setting expectations or firing warning shots for families, and considering if there needs to be any further actions at all. It is also just quite interesting to set your own expectations, which are intrinsic biases, against what the data from a more objective collection can get you. Plus it might open your eyes a bit further to the breadth of things which could be evidence based?


  • X @drbobphillips

  • Competing interests None declared.

  • Provenance and peer review Commissioned; internally peer reviewed.

Linked Articles