What is evidence based medicine?
Evidence based medicine (EBM) grew out of the work by Archie Cochrane, a doctor who noticed prior to and during the second world war (as a prisoner of war medical officer) that most medical treatments were not particularly effective nor beneficial and that many wounded soldiers recovered despite being given ineffective treatments. As an epidemiologist he felt obliged to do something about this and help the British National Health Service focus on effective treatments so as to get better value for money.
He referred to the first randomised controlled trial (RCT) published in 1952 as the beginning of a new era in measuring efficacy.
From these beginnings evidence based medicine developed slowly. It was later describe by Iain Chalmers as evaluating the latest evidence, preferably from randomised controlled clinical trials, and seeing if and how it can best be applied to the patient’s particular needs and values.
In 1991 at a conference in Manchester (UK) it was estimated that only about 15% of medical interventions had been proven to be beneficial based on good evidence, such as from a RCT. A major potential boost to EBM occurred following the Manchester conference with the establishment of the Cochrane Collaboration that was an attempt by medical researchers throughout the world to increase the proportion of medical interventions that were both safe and effective.
According to Iain Chalmers, one of the pioneers of evidence based medicine following in Archie Cochrane’s footsteps, practising EBM involves five distinct steps:
- Converting the need for information into an answerable question. The question can be a “background” question involving general knowledge about a particular disease or disorder such as “what host resistance factors protect asbestos-exposed workers from contracting mesothelioma?”; and a “foreground” question involving information about managing patients with this specific disorder, such as “in older patients with pulmonary mesothelioma from crocidolite exposure does adding radiotherapy yield enough reduction in morbidity to be worth its adverse effects?”;
- Tracking down the best evidence with which to answer that question;
- Critically appraising the evidence for its validity, size of effect and usefulness in clinical practice;
- Integrating this critical appraisal with clinical expertise and the patient’s unique biology, values and circumstances;
- Evaluating one’s effectiveness and efficiency in carrying out steps 1-4 and seeking ways to improve them both for next time.
In other words EBM involves evaluating the latest evidence, preferably from randomised controlled clinical trials, and seeing if and how it can best be applied to the patient’s particular needs and values.
The estimate quoted in 1991 at the Manchester conference of 15% of medical interventions proven to be beneficial based on good evidence has turned out to be a bit optimistic. Unfortunately little progress has been made since then mainly because of resistance by medical authorities to attempts to change the way they practise medicine. The British Medical Journal’s Clinical Evidence Group found that in 2008 an evaluation of more than 2,500 common medical interventions about 13% were effective with another 23% likely to be beneficial – Total 36%. In 2017 the figure for 3,000 common interventions was about 11% with another 23% likely to be beneficial – Total 34%. In other words, no change in 9 years. These categories are updated every 6 months.
CISS has estimated the comparable figures for cancer to be much less than this. For example there are no RCTs comparing cancer surgery, which accounts for many interventions, with no treatment, with the possible exception of comparing radical prostatectomy with ‘watchful waiting’ that included some treatment. This found no significant survival benefits. Similarly radiotherapy is rarely compared with no treatment in an RCT. It is sometimes added to surgery and compared to surgery only. Again there are few cases with significant benefits. Most only measure the ability to shrink tumours. Chemotherapy has been found to increase the percentage five year survival from about 65% to 67%.
By adding together the increased survival observed for different types of cancer using the main treatments of surgery, radiotherapy and chemotherapy CISS has estimated that about 3% of interventions have been proven to be beneficial. These are mainly through randomised controlled trials in terms of a slight increased survival (mainly of chemotherapy). There is another 7-10% with possible temporary benefits (such as removing or shrinking a tumour that is immediately life-threatening, such as pressing on the brain or obstructing the bowel. This figure also includes a larger increased survival following chemotherapy for acute childhood leukemias and some other rare cancers.
Cochrane AL. Effectiveness and Efficiency: Random Reflections on Health Services., British Medical Journal, The Nuffield Provincial Hospitals Trust, 1972, The Memoir Club 1989
Chalmers Iain (ed), Pregnancy and Childbirth, Oxford University Press, 1989