If the sample size is less than 10, what should be done with reference values?

If the sample size is less than 10, what should be done with reference values?

• A. Do NOT report reference values.
• B. Report reference values with caveats.
• C. Use literature RI without validation.
• D. Create a combined RI.

Answer: (A)

When you’re trying to define a reference interval, you need enough healthy samples to accurately map the range of normal values. With fewer than 10 individuals, the estimate becomes very unstable: a single outlier can swing the ends of the interval, and the confidence limits around those ends become so wide as to be meaningless. Reporting an RI from such a small dataset would mislead clinical decisions. Therefore, the best course is to not report reference values. Collect more samples to meet minimum guidelines, or rely on validated external reference intervals with appropriate local validation.

When you’re trying to define a reference interval, you need enough healthy samples to accurately map the range of normal values. With fewer than 10 individuals, the estimate becomes very unstable: a single outlier can swing the ends of the interval, and the confidence limits around those ends become so wide as to be meaningless. Reporting an RI from such a small dataset would mislead clinical decisions.

Therefore, the best course is to not report reference values. Collect more samples to meet minimum guidelines, or rely on validated external reference intervals with appropriate local validation.