when using evidence from clinical studies to inform treatment decisions involves accepting the 'average' results obtained in a group of people and generalising these to individual patients. However clinical studies often employ strict subject inclusion and exclusion criteria - though this approach can help to remove some potential biases in the study
it may mean that the characteristics of subjects in the study do not reflect the characteristics of all patients who will receive the intervention
possible reasons why study subjects may not reflect patients seen in routine clinical practice include
different risk factors for disease
presence of co-morbidities
differing levels of care received in different care settings
these factors, and many potential others, can mean that responses or susceptibilities to the effects of interventions may differ between those observed in study subjects and those potentially seen in actual patients
Notes:
allocation by chance in a randomised controlled trial should mean that the groups being compared are similar in terms of both measured and unmeasured baseline factors - however this is not so in cohort studies, and therefore cohort studies are vulnerable to selection bias (2). Therefore consider possible selection bias and confouding factors when appraising a study
Annotations allow you to add information to this page that would be handy to have on hand during a consultation. E.g. a website or number. This information will always show when you visit this page.