EBM Flashcards
selection bias
when subjects included in the study are not truly representative of the target population
definition bias
study subjects should be sharply defined
e.g. not just ‘tuberculosis’ but ‘sputum positive montoux positive TB’
prevents study designer from being subjective in interpretation
bias in concepts
lack of clarity about the concepts that are to be used in the proposed research
bias due to concurrent disease
selected patients may suffer from something that is seemingly unrelated, but their response may differ due to this, or any other medication they are on
information bias
occurs when key information is either measured, collected or interpreted inaccurately
interviewer/observer bias
when one is able to elicit a better response from one kind of patient than another kind
e.g. well educated vs. illiterate
recall bias (two types)
recency recall bias - events that happened a short time ago are more easily recalled
severity recall bias - easy to recall because they live longer in memory
response bias
relates to tendency of subjects to respond inaccurately or falsely to questions
e.g. when asked about sexual history, pt may not disclose having an STD due to stigma
attrition bias
rate of drop-out during clinical trials can effect sample sizes and data
e.g. if the most severe cases drop out during a trial
what is the hawthorn effect?
if a subject knows they are being measured or observed, their behaviours and responses may change
confounding bias
failure to account for cofounding variables that may effect the outcome of the trial
statistical bias
validity of statistical tests rely on the number of subjects in the study. Can produce false positives/false negatives if the sample size isn’t big enough
