Specialty › Precision & CI width

Bland-Altman LoA

Precision-based sizing finds the n needed to estimate a Bland-Altman limit of agreement to a target half-width, using SE(LoA) ≈ 1.71·σ_d / √n (Bland 2015).

What is Bland-Altman LoA?

The LoA SE has a 1.71 multiplier on top of σ_d / √n because the LoA = bias ± 1.96·σ_d combines uncertainty in the bias and in σ_d. The exact formula is in Bland 2015; we use the simplified approximation here.

When should I use Bland-Altman LoA?

  • Planning the n for a method-comparison study where you want precise LoAs.

What data does it need?

Expected SD of differences σ_d + target half-width.

What does it report?

Required number of pairs.

Formula

n ≈ (1.71 · σ_d / E)²

How do I interpret the result?

The 1.71 multiplier reflects combined uncertainty in bias and σ_d — LoA precision needs more n than a single-mean CI to the same half-width.

See also

References

  • Bland (2015). An Introduction to Medical Statistics, 4th ed.