evidence-based medicine:salvation or curse: evidence-based medicine: salvation or curse stephen hall cabrini institute


Slide2: Evidence-based medicine is ‘the integration of best research evidence with clinical expertise and patient values’


Slide3: Evidence-based medicine is ‘the integration of best research evidence with clinical expertise and patient values’ In practice is a methodology of acquiring and managing data to determine treatment effectives and safety


need for evidence-based medicine: need for evidence-based medicine Archie Cochrane claimed that only a small minority (10 -20%) of medical treatments were supported by evidence an expert committee of the US Institute of Medicine, (stated) only about 4% of all services have strong strength of evidence 1992


need for evidence-based medicine: need for evidence-based medicine there was a need to organize the evidence of effectiveness of treatments the term “evidence-based medicine” McMaster University


EBM as a mantra: EBM as a mantra


Slide8: is there a demonstrable need for evidence-based guidelines to treatment is it reasonable to use RCT data for all patients do techniques like meta-analysis value add can EBM be “subverted”


need for EBM: need for EBM Ellis J et al Lancet 1995 109 consecutive medical patients Oxford 4/95 82% evidence-based 53% RCT 29% convincing non-exptl data


need for EBM: need for EBM Ellis J et al Lancet 1995 109 consecutive medical patients Oxford 4/95 82% evidence-based 53% RCT 29% convincing non-exptl data no data eg DC for VF


need for EBM: need for EBM haematology > 70% psychiatry > 65% RCT general medicine > 50% RCT dermatology 60% with most others being “symptomatic and supportive” surgery !!!!!!


need for EBM: need for EBM when people look , they find


need for EBM: need for EBM when people look , they find the basis of EBM is the randomised controlled trial


Slide14: consider inclusion criteria for RA study 8 swollen and 8 tender joints raised ESR >30 raised CRP >20 methotrexate failure if early RA must have erosions


Slide15: consider exclusion criteria for RA trials 18 – 75 i.e. excludes the elderly no abnormality of renal, liver function no psychiatric disease no history of cancer save for non-melanotic skin cancer


Slide16: rheumatoid arthritis – 100 consecutive RA patients seen in practice only 15% ever eligible for RCT conducted for RA


Applying randomized controlled trial data to clinical practicePatients inclusion in randomized controlled trials: Included Patients (%) In these two studies, less than 21% of patients in clinical practice were eligible for inclusion in randomized controlled trials (RCT) Applying randomized controlled trial data to clinical practice Patients inclusion in randomized controlled trials 1. Dowd R, Recker RR, Heaney RP. Osteoporos Int 2000;11:533-6. 2. Hlatky MA, et al. Stat Med 1984;3:375-84.


Slide19: reports of trial data incomplete biased inconsistent with protocol


Slide20: 102 trials 62% major discrepancies in outcomes reported cf protocol 92% at least 1 incompletely reported efficacy outcome 81% at least 1 harm outcome Chan et al JAMA 2004


Slide21: we extrapolate from the refined patient group to the total pool ? validity


is there a better way to look at the data: is there a better way to look at the data literature review meta-analysis systematic review


meta-analysis: meta-analysis LeLorier et al NEJM 1997 examined large RCT >1000, adequate power over 3 years in premier journals then found meta-analyses prior to RCT


meta-analysis: meta-analysis


Slide26: Cochrane reports better than average but often flawed / poor quality ICU Delaney et al Critic Care Med 2007 asthma Jaddad et al BMJ 2000 maternal medicine Sheikh BMC Med 2000


Slide27: impossible to tell which articles were considered why they were rejected analysis difficult to replicate


Slide28: deus ex machina phenomenon


EBM in technology assessment: EBM in technology assessment Wooldridge M. “Australia first in world to adopt evidence based medicine”. Media release MW 77/98. 6 Apr 1998. PET scanning ?useful in Cancer treatment MSAC determined need for only limited number of machines


EBM in technology assessment: EBM in technology assessment Ware et al 2004 MJA MSAC “open + transparent” but minutes were incomplete record policy committee changed the conclusion of the scientific committee questions re use of data from research


EBM in technology assessment: EBM in technology assessment MSAC over 8 years considered of 56 health technology assess 31could be assessed no conflict of interest ever declared only 50% had validity assessments of studies hard to replicate the conclusions Petherik et al MJA 2007


clinical practice guidelines: clinical practice guidelines designed to improve patient outcomes hypertension, asthma, cigarette smoking 91 trials, 13 met criteria for inclusion 5 / 13 (38%) showed stat. significant results Worrall et al CMJ 1997 pay for performance linked to CPG – highly questionable Boyd JAMA 2005


Slide35: Eminence based medicine Isaacs & Fitzgerald BMJ 1999


Slide36: Eminence based medicine Vehemence based medicine


Slide37: Eminence based medicine Vehemence based medicine Eloquence based medicine


Slide38: Eminence based medicine Vehemence based medicine Eloquence based medicine Providence based medicine


Slide39: Eminence based medicine Vehemence based medicine Eloquence based medicine Providence based medicine Diffidence based medicine


Slide40: Eminence based medicine Vehemence based medicine Eloquence based medicine Providence based medicine Diffidence based medicine Nervousness based medicine


Slide41: Eminence based medicine Vehemence based medicine Eloquence based medicine Providence based medicine Diffidence based medicine Nervousness based medicine Confidence based medicine