Examkin

Question

Difficulty: MediumMeasures of Association and Impact

An international cardiovascular research consortium examines the efficacy of a novel synthetic GLP-1 receptor agonist in preventing major adverse cardiovascular events (MACE). The study follows adults with type 2 diabetes and established atherosclerotic disease over a 4-year period. Event frequencies are documented as follows:

- GLP-1 receptor agonist group: 80 events out of 1,000 patients
- Placebo group: 120 events out of 1,000 patients

The researchers report a two-sided pp-value of 0.0040.004 for the overall difference in MACE incidence. During a secondary analysis, the investigators stratify the data by baseline statin use. In the subgroup of patients taking statins, the relative risk (RR) for MACE with the GLP-1 receptor agonist is 0.430.43 (95% CI:0.28−0.6595\% \text{ CI}: 0.28 - 0.65). In the subgroup of patients not taking statins, the RR is 1.01.0 (95% CI:0.85−1.1895\% \text{ CI}: 0.85 - 1.18).

Based on the overall cohort data, what is the number needed to treat (NNT) to prevent one MACE, and which epidemiological concept best explains the findings of the statin-stratified analysis?

  1. NNT = 25; Effect modificationAnswer
  2. B
    NNT = 25; Confounding
  3. C
    NNT = 3; Effect modification
  4. D
    NNT = 3; Confounding
  5. E
    NNT = 13; Confounding

Answer

NNT = 25; Effect modification
The correct answer combines accurate NNT calculation with the proper interpretation of stratified data. First, the Absolute Risk Reduction (ARR) is the difference in event rates between the placebo and treatment groups: 0.12−0.08=0.040.12 - 0.08 = 0.04. The Number Needed to Treat (NNT) is the inverse of the ARR: 10.04=25\frac{1}{0.04} = 25. This means 25 patients must be treated for 4 years to prevent one MACE. Second, the study evaluates the role of statins. The crude (overall) Relative Risk (RR) is 0.080.12=0.67\frac{0.08}{0.12} = 0.67. When stratified by statin use, the RRs are profoundly different from each other (0.430.43 vs. 1.01.0). Because the effect of the GLP-1 agonist varies based on the presence of a statin, statin use is acting as an effect modifier, not a confounder.

Step-by-Step Solution

1
Calculate the incidence of MACE in both groups.
Control Incidence (IcI_c) = 120/1000=0.12120 / 1000 = 0.12. Intervention Incidence (IeI_e) = 80/1000=0.0880 / 1000 = 0.08.
Baseline risks are required to determine the absolute risk reduction.
2
Calculate the Absolute Risk Reduction (ARR).
ARR = Ic−Ie=0.12−0.08=0.04I_c - I_e = 0.12 - 0.08 = 0.04 (or 4%4\%).
ARR represents the actual difference in risk attributable to the intervention.
3
Calculate the Number Needed to Treat (NNT).
NNT = 1ARR=10.04=25\frac{1}{\text{ARR}} = \frac{1}{0.04} = 25.
NNT is the reciprocal of the absolute risk reduction.
4
Interpret the stratified analysis to distinguish between confounding and effect modification.
The RR for statin users (0.430.43) is significantly different from the RR for non-users (1.01.0). This indicates effect modification.
When a third variable fundamentally changes the magnitude or direction of the association between an exposure and an outcome, it is an effect modifier. If it were a confounder, the stratified RRs would be similar to each other.

Key Concept

Calculating NNT using Absolute Risk Reduction and distinguishing effect modification (differing stratified relative risks) from confounding.
Rate this question

Topics

Measures of Association and Impact

More questions from this topic