S3 + S4 Critical Appraisal

Question 1

What is critical appraisal?

Answer 1

The process of carefully and systematically examining research to judge its trustworthiness and its value and relevance in a particular context

Question 2

What is PICO?

Answer 2

Patient/population
Intervention/exposure
Comparison
Outcome

Question 3

What are the initial keys questions to consider when doing a critical appraisal?

Answer 3

1. Is this article relevant to my interests?
2. Is there a clear research question?
3. When was it published?
4. Has it been peer reviewed?

Question 4

What is confirmation bias?

Answer 4

Evaluating evidence that supports one’s preconceptions differently from evidence that challenges these convictions

Question 5

What is the summary of the process of doing a critical appraisal?

Answer 5

1. FORMULATE PICO qs
2. IDENTIFY keywords
3. PLAN the search
4. EXECUTE the search
5. REFINE the results
6. REVIEW the literature
7. ASSESS the evidence

Question 6

What should you think about a study’s method when you’re doing a critical appraisal?

Answer 6

• is there selection bias - is there a systematic error difference between the people that have and haven’t been selected?
• data collection - how were data collected and how might that impact on the study findings?
• confounding factors - what are they? Have they been discussed? Have they been adjusted for?

Question 7

What should you think about a study’s results when you’re doing a critical appraisal?

Answer 7

• what is the sample - how is the sample described? Is any justification given regarding the size of the sample?
• what stats are presented - what is the outcome measure presented? How are figures and table presented? Are the results statistically significant?
• are the results generalizable?

Question 8

What should you think about a study’s discussion of findings when you’re doing a critical appraisal?

Answer 8

• is the main finding of the study clearly given and does is reflect the aim of the study?
• how does the study fit with what we already know?
• are the strengths and limitations discussed - has anything been missed?
• is a clear final conclusion given as well as recommendations for next steps?

Question 9

How can statistics be misleading?

Answer 9

• ignore any withdrawals or non-responders in the analysis
• if see no interesting results, did any subgroups behave differently
• report any relationships are statistically significant
• if confidence intervals overlap zero difference between the groups and leave them out
• if analysing your data the way you planned doesn’t show anything, try a different tests

Question 10

What are we most concerned about with a case-control study?

Answer 10

• selection bias
• recall bias
• misclassification bias
• interviewer/observer bias
• confounding

Question 11

What are the key questions we should be asking when we read a case-control study?

Answer 11

• is the aim clear?
• is it well defined?
• do the cases represent the population of interest?
• how is exposure and outcome defined and measured?
• do we know about the response/drop out rate?
• is confounding discussed?
• is the number of cases and controls justified?
• is the key finding precise?
• are the results applicable to the population?
• does it fit with other research?

Question 12

What are we most concerned about with a cohort study?

Answer 12

• selection bias
• response bias
• measurement bias
• healthy worker effect
• loss to follow up
• confounding

Question 13

What are the key questions we should be asking when we read a cohort study?

Answer 13

• is the aim fo the study clear?
• how was the cohort recruited?
• how was exposure and outcome measured?
• can we be sure participants were outcome free at the start?
• was the follow-up period long enough?
• is confounding discussed?
• is the key finding precise?
• is the finding applicable to population?
• how does it fit with other research?

Question 14

What are we most concerned about with a cross-sectional study?

Answer 14

• selection bias
• response bias
• measurement bias
• confounding

Question 15

What are the key questions we should be asking when we read a cross-sectional study?

Answer 15

• is the aim clear?
• how representative is the sample?
• what do we know about non-respondents?
• how has data been collected?
• is confounding discussed?
• is the key finding precise?
• how applicable are the results to the population?
• how does it fit with other research?

Question 16

What are issues with ecological studies?

Answer 16

• is the population clearly defined?
• is the source of information well described?
• confounding

Question 17

When is a cross-sectional study used?

Answer 17

To determine burden of disease in a population (prevalence)