Confidence interval approach for evaluating bias in laboratory methods
1988

Evaluating Bias in Laboratory Methods Using Confidence Intervals

Sample size: 21 publication Evidence: moderate

Author Information

Author(s): K. F. Yee

Primary Institution: Beecham Pharmaceuticals

Hypothesis

Can confidence intervals provide a better evaluation of bias between laboratory quantitation methods than traditional significance tests?

Conclusion

The study suggests that using confidence intervals can provide a more meaningful evaluation of bias between laboratory methods than conventional significance tests.

Supporting Evidence

  • The study found that the flame photometry method was biased, measuring 1.6% higher than the Astra method.
  • The confidence interval approach allows for a more practical interpretation of bias between methods.
  • Using a maximum acceptable difference, H, can help determine if two methods are equivalent.

Takeaway

This study shows that when comparing two lab methods, it's important to see if their results are close enough, not just if they are different.

Methodology

The study compares two laboratory quantitation methods by evaluating the confidence interval of their mean difference.

Potential Biases

There is a risk of misinterpreting small differences as significant bias if not evaluated in practical terms.

Limitations

The study primarily focuses on normally distributed data and may not apply to other distributions without adjustments.

Participant Demographics

21 patient serum specimens were analyzed.

Statistical Information

P-Value

0.025

Confidence Interval

(1.0, 2.2)

Statistical Significance

p<0.05

Want to read the original?

Access the complete publication on the publisher's website

View Original Publication