Week 2: Variance Management Systems Issues

Bob Luttman, Robert Luttman & Associates

                                               
Home Page

Overview

Statistical Issues

Multi-collinearity

Cascade Effect

Sample Size

Halo Effect

Operational Issues: Documentation and Reporting

Too Much Data

Traditional Pathway / Variance Documentation

What's Wrong With This Picture?

Operational Issues: Summary

Legal Issues: Overview

Legal Issues

The Biggest Issue: So What?

Summary and Conclusion

Assignment

Feedback

Questions?

Statistical Issues

Variance data is, potentially, the mother lode of clinical process improvement data. Everyone wants the data, and visions of regression functions dance in analytical minds.

Then reality bites. Hard.

Statistical issues seriously limit what you can and cannot do with all that data you've collected. Variance data is - foremost - process data; and process data is inherently messy. These are not clinical trials, they are real world experiences.

The most serious statistical issues are :

  • Multi-collinearity: the lack of statistical independence amongst variances.
  • Cascade Effect: today's variances causes tomorrow's variances.
  • Sample Size: you can't have more variables than patients.
  • Halo Effect: everybody's LOS is declining.

 

Home Page | Overview | Statistical Issues | Multi-collinearity | Cascade Effect | Sample Size | Halo Effect | Operational Issues: Documentation and Reporting | Too Much Data | Traditional Pathway / Variance Documentation | What's Wrong With This Picture? | Operational Issues: Summary | Legal Issues: Overview | Legal Issues | The Biggest Issue: So What? | Summary and Conclusion | Assignment | Feedback | Questions?

rluttman@robertluttman.comt
Improving Healthcare Across the Continuum