GRADE tables Flashcards
(22 cards)
What are the four categories of overall certainty in GRADE tables?
High, moderate, low, and very low
What is the starting certainty level for RCTs in GRADE tables?
High certainty
What is the starting certainty level for observational studies in GRADE tables?
Low certainty
How does serious concern in one domain affect overall certainty?
Decreases overall certainty by one category
How does very serious concern in a domain affect overall certainty?
Decreases overall certainty by two categories
What factors can lead to an increase in overall certainty?
Large effect, dose-response, or confounders that would only reduce the effect size
What is the risk of bias based on in GRADE tables?
Assessment of risk of bias from individual studies
What should be considered if a study has a high risk of bias?
Its weight in the meta-analysis for the outcome
What weight contributes to a ‘not serious’ rating in risk of bias?
<10%
What weight contributes to a ‘serious’ rating in risk of bias?
> 50%
What is used to judge inconsistency of results?
Eyeball test and Chi-square test with cutoff p-value 0.10
What does the eyeball test involve?
Checking if all confidence intervals (CIs) overlap the overall effect point estimate
What is the significance of I2 in assessing inconsistency?
> 50% is probably concerning; <30% is likely not concerning
Why is I2 more important for dichotomous outcomes than continuous outcomes?
Dichotomous outcomes are on relative scale; continuous outcomes are on absolute scales
What should be considered if all studies have a CI on one side of a threshold?
No concern for serious inconsistency regardless of p-value and I2
What must be done if there is heterogeneity in the results?
Search to explain it, such as through subgroup analysis
What is assessed in the indirectness of evidence?
Characteristics of included studies in relation to the PICO question
What is the focus of indirect comparisons?
Head-to-head comparison (e.g., drug A vs. drug B)
What is the main consideration in assessing imprecision of results?
Overall effect (bottom diamond)
Does the individual study size influence the assessment of imprecision?
No, it does not matter
What indicates imprecision in decision-making?
If different decisions would be made for low value and high value
What should be checked regarding the CI of the pooled effect?
Whether it includes a clinically important difference on both sides of the null
