This is the conscientious and judicious use of current bestevidence from clinical care research in the management of individualpatients taking into account their values.
A partial solution
50 journals are scanned not by experts in neonatal nephrology orthe left nostril, but by searchers trained to spot papers which have a directmessage for practice, and meet predefined criteria of rigour (below). Summariesare then published in Evidence-based Medicine.
Questions used to evaluate papers
- Are the results valid? (Randomized? Blinded? Were all patientsaccounted for who entered the trial? Was follow-up complete? Were the groupssimilar at the start? Were the groups treated equally, apart from theexperimental intervention?)
- What are the results? (How large was the treatment effect? Howprecise was the treatment effect?)
- Will the results help my patients (cost-benefitsum).
Problems with the solution
The concept of scientific rigour isopaque. What do we want? The science, the rigour, the truth, or what willbe most useful to our patients? These may overlap, but they are not thesame.
- Will the best be the enemy of the good? Are useful papers rejecteddue to some blemish? Answer: all evidence needappraising (often impossible!).
- By reformulating patients in terms of answerable questions, EBMrisks missing the point of the patient's consultation. He might simply want toexpress his fears, rather than be used as a substrate for an intellectualexercise.
- Is the standard the same for the evidence for all changes to ourpractice? For example, we might want to avoid prescribing drug X forconstipation if there is the slimmest chance that it might cause colon cancer.There are many other drugs to choose from. We might require far more robustevidence than a remote chance to persuade us to do something rathercounter-intuitive, such as giving heparin in DIC. How robust does the data needto be? There is no science to tell us the answer to this: we decide off the topof our head (albeit a wise head, we hope).
- EBM is a lucky dip if gathering all the evidence on a topic provesimpossible.
- What about letters columns? It may be years before fatal flaws areaired.
- There is a danger that by always asking รข€˜What is the evidence we divert resources from hard-to-prove activities (which may be veryvaluable eg physiotherapy for cerebral palsy) to easy-to-prove services. Theunique personal attributes of therapists are as important as the objectiveregimen. It is all too easy to transfer resources to some easy-to-quantifyactivity, eg neonatal screening for cystic fibrosis.
- Evidence-based medicine is rarely 100% up to date. Reworkingmeta-analyses in the light of new trials takes time if it is ever done atall.
- EBM contributes to the problem of data-overload by churning outendless guidelines that don't quite apply to the patient sitting in front ofyou.
Advantages of evidence-based medicine
- It improves our reading habits.
- It leads us to ask questions, and then to be sceptical of theanswers.
- As taxpayers, we should like it (wasteful practices can beabandoned).
- Evidence-based medicine presupposes that we keep up to date, andmakes it worthwhile to take trips around the perimeter of our knowledge.
- Evidence-based medicine opens decision-making processes topatients.
Conclusion
There is little doubt that, where available, EBM is better thanwhat it is superseding. It may not have much impact, as gaining evidence istime-consuming and expensive. Despite these caveats, evidence-based medicine isone of the most rational recent medical developments. Let's all join in bysubscribing to ideals and its journal.
Advertisements
