EBM Flashcards
3 main issues address when undergoing critical appraisal
Validity, the results, the clinical relevance
validity
how trustworthy the study is, does it measure what its supposed to. Minimising bias improves validity
Results
What does this show, is it statistically significant, size of the effect. (only worth thinking about implications of results if study design and methods are valid)
Clinical relevance
will the results help you and your patient. Differences between patient and participants identified to see effectiveness.
Whats involved in the critical appraisal checklist, what are the 6 questions
Was the study original?(does it add anything to current literature) Who was the study about? (how were Ps recruited/inclusion exclusion criteria) Was the study design sensible? Was bias avoided? (control group used, randomisation, other sources) Was assessment blind? Were preliminary statistical questions addressed? (size of sample, followup) different checklists are used for different study types.
Critical appraisal skills programme (CASP)
provides critical appraisal checklists that are designed to help you think about aspects of appraisal systematically.
CONSORT
focuses on RCT and aspects that should be included. Can be used to critically evaluate or design.
STROBE
helps design and appraise observational studies
Hierarchy of evidence
SR and MA. RCT. Cohort. CC. Cross sectional. Ecological. Case reports. Ideas and opinions.
GRADE
grading and recommendations assessment, development and evaluation. ranks quality of evidence. GRADE and SIGN take into account more dimensions.
PICO
patient, intervention, comparison, outcomes.
Reaching a decision about a paper
Reject it, interpret the results cautiously, accept it/trust.
Pre trial Bias
selection bias. not representative of the population. groups differ from each other.
Channeling bias - patient prognostic factors or degree of illness dictates the study cohort into which patients are placed
During trial bias
information bias. errors in measuring exposure, intervention or outcome.
interviewer. Chronology. Recall. Transfer. Performance
Post trial bias
confounding. statistical bias - misleading conclusions.
Citation. Internal/external validity.