Regression toward the mean – a detection method for unknown population mean based on Mee and Chua's algorithm
2008

Detecting Regression to the Mean in Observational Studies

publication Evidence: moderate

Author Information

Author(s): Ostermann Thomas, Willich Stefan N, Lüdtke Rainer

Primary Institution: University of Witten/Herdecke

Hypothesis

Can we develop a method to detect regression to the mean in situations where the population mean is unknown?

Conclusion

The proposed method can effectively differentiate between regression to the mean effects and actual treatment effects in uncontrolled studies.

Supporting Evidence

  • The method was applied to three real-world examples demonstrating its effectiveness.
  • It can clarify evidence from uncontrolled observational studies.
  • The approach helps in meta-analysis and health-technology reports.

Takeaway

This study helps us understand how to tell if changes in health outcomes are real or just due to random chance when measuring the same people over time.

Methodology

The study extends Mee and Chua's algorithm to situations where the population mean is unknown, using differential calculus to estimate treatment effects.

Potential Biases

The model may overestimate treatment effects if the correlation between measurements is incorrectly specified.

Limitations

The method requires an estimate of the correlation between baseline and follow-up values, which is often not reported in studies.

Statistical Information

P-Value

0.0504

Statistical Significance

p<0.05

Digital Object Identifier (DOI)

10.1186/1471-2288-8-52

Want to read the original?

Access the complete publication on the publisher's website

View Original Publication