What is the minimal detectable change (MDC)?

What is the minimal detectable change (MDC)?

• A. The smallest amount of change beyond measurement error that reflects a true difference.
• B. The minimal clinically important difference.
• C. The p-value threshold for significance.
• D. The difference in baseline characteristics.

Answer: (A)

The main idea is that the minimal detectable change is the smallest amount of change that rises above the instrument’s measurement error to reflect a real difference. It depends on how precise the measurement is (the standard error of measurement) and the level of confidence you want. For a 95% MDC, you use the formula MDC = z × SEM × √2, with z = 1.96 for 95% confidence. This accounts for error in both the initial and follow-up measurements. If the observed change is smaller than the MDC, it could just be noise from measurement error; if it’s larger, you can be reasonably confident a true change has occurred. This is different from the minimal clinically important difference, which asks whether the change matters to the patient, and from other options like the p-value threshold or differences in baseline characteristics. Example: if SEM is 3 units, MDC95 ≈ 1.96 × 3 × 1.414 ≈ 8.3 units, so changes smaller than about 8.3 units are not considered real at 95% confidence.

The main idea is that the minimal detectable change is the smallest amount of change that rises above the instrument’s measurement error to reflect a real difference. It depends on how precise the measurement is (the standard error of measurement) and the level of confidence you want. For a 95% MDC, you use the formula MDC = z × SEM × √2, with z = 1.96 for 95% confidence. This accounts for error in both the initial and follow-up measurements. If the observed change is smaller than the MDC, it could just be noise from measurement error; if it’s larger, you can be reasonably confident a true change has occurred. This is different from the minimal clinically important difference, which asks whether the change matters to the patient, and from other options like the p-value threshold or differences in baseline characteristics. Example: if SEM is 3 units, MDC95 ≈ 1.96 × 3 × 1.414 ≈ 8.3 units, so changes smaller than about 8.3 units are not considered real at 95% confidence.