Interpreting Study Results Flashcards
How are study results presented?
- Mean values (or mean changes from baseline)
- Proportion (percentages, risk ratio, and odds ratio)
- Time to an outcome (survival curve, hazard ratio)
What types of outcomes can be successfully represented by means?
Continuous outcomes (ex. BP)
What are proportions?
The fraction of the total that possesses the outcome
Proportions are often used to compare outcomes between groups
ex. Compare failure rates between groups of the Vitamin D study
What are some types of proportions used to compare outcomes of studies?
- Absolute Risk Reduction (ARR)
- Relative Risk (RR)
- Relative Risk Reduction (RRR)
- Odds Ratio (OR)
- NNT/NNH
What is absolute risk reduction (ARR)?
The absolute difference between the probability of the event in the control group and probability of the event in the intervention group
ex. ARR of stroke= 8.3% (placebo) - 5.7% (Drug X) = 2.6%
Drug X reduces stroke by 2.6% vs. placebo
What is relative risk (RR)?
This shows the “risk” of the outcome in the intervention group is compared to the “risk” in the control groups
ex. RR= Intervention risk/placebo risk = 5.7%/8.3% = 0.68
Can continuous outcomes be used to determine relative risk (RR)?
No, relative risk can only be calculated for dichotomous outcomes (yes or no?)
How are relative risk (RR) values analyzed?
RR = 1.0 (No difference in risk between groups)
RR < 1.0 (Less risk of outcome in intervention group
RR > 1.0 (Higher risk of outcome in intervention group)
What is relative risk reduction (RRR)?
The degree to which baseline risk is reduced (or increased) by the intervention
RRR(relative risk reduction) = 1-RR(relative risk) = 1- 0.68 = 0.32
Drug X reduces the risk of stroke by 32% vs. placebo
Can continuous outcomes be used to determine relative risk reduction (RRR)?
No, relative risk reduction (RRR) can only be calculated for dichotomous outcomes
What is odds ratio (OR)?
This value shows the odds of the outcome occurring in the intervention group compared to the control group
Odds = (risk of having event)/(risk of not having event)
Odds Ratio = odds (intervention)/odds (placebo)
How can odds ratio (OR) values be interpreted?
OR = 1.0 (No difference in odds between groups)
OR < 1.0 (Less odds of outcome in intervention group)
OR > 1.0 (Higher odds of outcome in intervention group)
What is Number needed to treat (NNT)?
Number of subjects who would have to be treated (receive the intervention) in order for one additional subject to “benefit” in comparison to the control
Duration of study is also important when interpreting results
ex. NNT = 100%/ARD (absolute risk difference) = 100/2.6 = 38.5= 39
We need to treat 39 patients with Drug X to prevent stroke in one additional person compared to placebo
What is Number needed to harm (NNH)?
Similar concept as NNT, except in NNH the outcome that we prevent is an adverse effect
ex. NNH = 100, Need to treat 100 people to see one additional person to not experience a defined side effect vs. the comparator group
Is a small NNT always an indication for the studied therapy to be highly reccomended?
No, we need to look at the results of the study.
A study may have a small NNT, but both interventions studied may have grave adverse effects
What are some ways study results can be compared to each other as time to outcome?
- Survival curve
- Hazard ratio
What is the Kaplan-Meier survival curve?
Compares how long it takes subjects to reach the outcome (ex. median time until stroke)
Often reported as median survival time (time for half of the subjects to reach outcome)
What is the survival (outcome survival) rate at the start of the study?
100%
What is the hazard ratio (HR)?
This helps compare survival curves between groups
HR = 1.0 (No difference in hazard betweeen groups)
HR less than 1.0: Less hazard of outcome in intervention group
HR more than 1.0: Higher hazard of outcome in intervention group
What are confidence intervals (CI)?
- Range where the true effect of the intervention (treatment) lies
- The narrower the CI, the more precise the results
ex. 0.32 (95% CI 0.22 to 0.40). If we run our study 100 times, in 95 trials our study will acheive results within the confidence interval
What are advantages of confidence intervals?
CI can show us the magnitude of the effect, while p-values can only give us a binary outcome (significant, and insignificant)
CI values also show us the direction of effect
CI values also help distinguish between clinical and statistical significance
How can confidence intervals be used to compare groups?
If the CI crosses the threshold for “no difference”, then the result is not statistically significant
For mean values or proportions: no difference = 0
For relative risk or odds ratio: no difference = 1
review slides 33 and 34
What are superiority studies?
Most studies are superiority studies
These studies want to show an intervention is better than control (active/placebo)
What are non-inferiority trials?
These studies try to determine whether new intervention is not substantialy worse than an established intervention
What is considered to be not substantially worse must be predetermined before conducting study
Why are intention-to-treat (ITT) and per-protocol analyses performed for non-inferiority trials?
IIT (include all participants, including droupouts. Therefore minimizing effects of drug in results)
per-protocol (only those that followed treatment fully, but will see greater drug effects)
What are equivalence studies?
Performed rarely
The new intervention is neither worse nor better than an established intervention
Review slide 42 for review of interpreting non-inferiority margin
What are propensity scores?
They are used in observational studies to determine the probability that a subject would be in a particular treatment group based on their observed baseline characteristics
What is adaptive clinical trial design?
These types of studies have the ability to have pre-determined modifications to the study design (this makes them these types of studies more efficient compared to traditional trials)
