Epi - Lecture 5 - Critical appraisal 2

evidence hierarchy

• system used to rank different types of research based on their reliability, validity, and strength

internal validity

• refers to the extent to which the results of a study accurately reflect the relationship between the exposure and the outcome

external validity

• refers to the extent to which the study's findings can be generalized to other populations, settings, or times

two forms of external validity

• population validity → the participant characteristics have to represent the larger population you want to generalise

• ecological validity → how similar the setting is to the real world and how the world changes over time

shortcomings evidence hierarchy

• does not take into account that the quality of research is also affected by how the research is performed,

• qualitative (anthropological) studies are not included → shortcoming because these studies are crucial to understand cultural factors (important for health interventions in certain communities).

different conflicts of interest

• financial → source of funding

• non-financial → strong personal, political or religious beliefs on a topic

why is a critical appraisal important

• helps to narrow down quantities of literature to relevance and high quality

• helps to assess validity and applicability

• helps to identify bias in studies

• is essential to a literature review

• is important for treatment and intervention planning

what are the 2 critical appraisal tools

• CAT (critical appraisal tool) → assesses whether a study is internally and externally valid.

• standard reporting guidelines → focuses on guiding researchers on what to report or what elements to include in an article.

CAT, critical appraisal tool

• to evaluate/assess whether a study is internally and externally valid

• focuses on evaluating methods and results of a study, but does not always thoroughly investigate the whole stud

standard reporting guidelines

• provides guidelines for researchers on what to report or what elements to include in an article

• focuses more on guiding researchers on what to report while drafting a research article

• more thorough → looks at all elements of the study, from intro to discussion

things to specifically check when assessing a case control study

• selection of the control

• recall and social desirability bias

• how well do controls fit with the cases

things to specifically check when assessing a cohort study

• participant characteristics (what does the cohort look like, is it representative)

• the follow up (relevant for the outcome)

• confounders (what has been controlled for)