Examkin

Question

Difficulty: MediumStudy Design

An investigator initiates a randomized controlled trial comparing a high-intensity interval training (HIIT) program to standard moderate-intensity continuous training for improving glycemic control in patients with newly diagnosed type 2 diabetes. Five hundred patients are randomized equally between the two groups. Hemoglobin A1c is measured at baseline and at 12 months. During the follow-up period, 85 patients in the HIIT group abandon the study due to musculoskeletal injuries and excessive fatigue. In contrast, only 15 patients in the standard training group drop out. The investigator conducts a per-protocol analysis, excluding all dropouts from both groups, and reports a statistically significant reduction in Hemoglobin A1c in the HIIT group compared to the standard group. Which of the following phenomena most likely compromises the validity of this conclusion?

  1. A
    Confounding by indication
  2. B
    Effect modification
  3. Selection biasAnswer
  4. D
    Recall bias
  5. E
    Observer bias

Answer

Selection bias
The correct answer identifies selection bias, specifically attrition bias. When a significant and disproportionate number of participants drop out of one arm of a randomized controlled trial, the remaining participants are no longer representative of the original randomized group. By performing a per-protocol analysis (excluding dropouts) rather than an intention-to-treat analysis (analyzing patients in the groups to which they were originally randomized regardless of adherence), the study loses the benefits of randomization. The patients who tolerated the rigorous HIIT program and stayed in the study are likely systematically healthier, more resilient, or more highly motivated than those who dropped out due to injuries or fatigue. Comparing this self-selected subset to the standard group artificially inflates the apparent efficacy of the intervention. This differential loss to follow-up introduces a classic form of selection bias.

Step-by-Step Solution

1
Identify the study design and the method of data analysis used by the investigator.
The study is a randomized controlled trial that utilized a per-protocol analysis.
Understanding the design and analysis method is necessary to identify potential methodological flaws.
2
Analyze the loss to follow-up data between the two study arms.
There is a massive discrepancy: 85 dropouts in the HIIT group versus 15 in the standard group.
Differential attrition is a major red flag in clinical trials that can distort the true effect of an intervention.
3
Evaluate the impact of combining differential attrition with a per-protocol analysis.
Excluding the dropouts means the remaining HIIT participants are likely a self-selected, healthier, or more resilient cohort compared to the baseline randomized group.
Randomization only ensures groups are comparable at baseline. Differential dropout disrupts this balance, making the remaining groups fundamentally different.
4
Classify this specific methodological flaw into the correct epidemiological category.
Differential loss to follow-up introduces attrition bias, which is a specific subtype of selection bias.
Matching the mechanism of the error to formal epidemiological terminology is required to answer the question.

Key Concept

Attrition bias (a subtype of selection bias) occurs when there is differential loss to follow-up between study groups, which can be mitigated by using an intention-to-treat analysis rather than a per-protocol analysis.
Rate this question

Topics

Study Design

More questions from this topic