Preferences or inclinations, perhaps stemming from prejudice impact on clinical medical practice daily, and unhappily if part of the culture of the medical practitioner themselves are unlikely to be perceived as such. The science of clinical statistics has demonstrated time and again the lack of objectivity in evaluating risk that applies to much of healthcare practice. This demonstration of statistical bias systematically favouring some outcomes over others will remain a problem, as human cultures and the human organism have evolved to be biased.
Common causes of bias
Common examples of bias in clinical medical practice are due to social and cultural attitudes to:
- The Individual
- The social group
- Caste/Social class
- Social values
- Power (culturally)
- Acts of the individual
- Acts of society
- English language sources on GANFYD
- Use of abstracts for references
- Biased towards secondary outcomes
- New Pharmaceuticals
- Better evidence base due to regulatory demand
- Sponsorship of a Study
- Meta-analysis of cost effectiveness studies shows those sponsored by company marketting a product of interest more likely to claim cost effectiveness than independent organisation studying same class of drugs
- Personal interest of researcher
- Financial, Pharmaceutical company entertainment, Power in profession
- Publication bias
- More "interesting" findings are more likely to be published. Positive results - e.g. those showing that a treatment is effective - are more likely to be published than negative results. Results which challenge received wisdom might also be more newsworthy, and therefore more likely to be published, than "boring" results that show what "we knew anyway" (although the published evidence may not be good).
- Smaller not fully reported studies tend to show less treatment effect than published larger ones which is important in ascertainment of all relevant trials in meta-analysis
- There is selective publication of drug development trial data. This effect has been analyzed in detail for antidepressant trials where an effect size of 32% from selective publication of clinical efficacy occurred . The amount of bias varied with drug.
- ↑ [http://www.cochrane-net.org/openlearning/HTML/mod15-2.htm "What is publication bias" Cochrane foundation. Last accessed 24/7/06
- ↑ Publication bias from numberwatch, last accessed 24/7/06
- ↑ [http://skepdic.com/posoutbias.html "Positive outcome (publication) bias" Skeptic's Dictionary. Last accessed 24/7/06
- ↑ Gunther Eysenbach. Tackling Publication Bias and Selective Reporting in Health Informatics Research: Register your eHealth Trials in the International eHealth Studies Registry (editorial). Journal of Medical Internet Research 2004;6(3):e35) doi:10.2196/jmir.6.3.e35
- ↑ Dickersin K. The existence of publication bias and risk factors for its occurrence. JAMA : the journal of the American Medical Association. 1990;263:1385-9.
- ↑ Hopewell S, McDonald S, Clarke M, Egger M. Grey literature in meta-analyses of randomized trials of health care interventions. Cochrane database of systematic reviews (Online). 2007; (2):MR000010.(Epub) (Link to article – subscription may be required.)
- ↑ Turner EH, Matthews AM, Linardatos E, Tell RA, Rosenthal R. Selective publication of antidepressant trials and its influence on apparent efficacy. The New England journal of medicine. 2008 Jan 17; 358(3):252-60.(Link to article – subscription may be required.)