Recursive partitioning

From Citizendium
Revision as of 23:25, 1 April 2008 by imported>Robert Badgett
Jump to navigation Jump to search
This article is a stub and thus not approved.
Main Article
Discussion
Related Articles  [?]
Bibliography  [?]
External Links  [?]
Citable Version  [?]
 
This editable Main Article is under development and subject to a disclaimer.

Recursive partitioning is a statistical method for the multivariable analysis of diagnostic tests.[1][2][3]. Recursive partitioning creates a decision tree that strives to correctly classify members of the population based on a dichotomous dependent variable. Compared to other multivariable methods, recursive partitioning:

  • Advantages are:
    • Generates clinically more intuitive models that do not require the user to perform calculations.[4]
    • Allows "adjustable misclassification penalties" or misclassification costs in order to create a decision rule that has more sensitivity or specificity. This has also been called the "dispersity index" which is the sum of the false-negatives and false-positives. Either the false-negatives and false-positives can be weighted in order to preferentially reduce their occurrence.[5]
  • Disadvantages are:
    • Does not work well for continuous variables[6]
    • May overfit data.

A variation is 'Cox linear recursive partitioning'.[5]

Examples are available of using recursive partitioning in research of diagnostic tests.[7][8][9][10][11] Goldman used recursive partitioning to prioritize sensitivity in the diagnosis of myocardial infarction among patients with chest pain in the emergency room.[11]

Studies comparing the predictive accuracy of recursive partitioning with logistic regression have reported both recursive partitioning[12] and logistic regression to be more accurate.[7]

Calculations

Unweighted analysis

An 'Initial misclassification rate' is calculated by assigning all cases to a single class. Using the example in the CART Manual, in trying to diagnose 59 patients who have a disease of interest among a sample of 189 patients, arbitrarily classifying all patients as normal yields:[13]
For the top node, also called node 1:

For the next nodes: The next best independent, classifier variable creates two nodes:

  • Left node: 159 patients; 118 have disease and 41 are false positives who do not have the disease
  • Right node: 30 patients; 12 have disease and false negatives

Weighted analysis

In this analysis, the sensitivity is preferenced by weighted it by 2.2, which is the ratio of patients without disease over patients with disease.
For the top node:

For the next nodes: The next best independent, classifier variable creates two nodes:

  • Left node: 159 patients; 118 have disease and 41 are false positives who do not have the disease
  • Right node: 30 patients; 12 have disease and false negatives

References

  1. Breiman, Leo (1984). Classification and Regression Trees. Boca Raton: Chapman & Hall/CRC. ISBN 0-412-04841-8. 
  2. Lewis RJ (2000). An introduction to classification and regression tree (CART) analysis. Accessed March 12, 2008
  3. Yohannes Y, Hoddinott J (1999). Classification and regression trees: an introduction. Accessed March 12, 2008
  4. James KE, White RF, Kraemer HC (2005). "Repeated split sample validation to assess logistic regression and recursive partitioning: an application to the prediction of cognitive impairment". Statistics in medicine 24 (19): 3019-35. DOI:10.1002/sim.2154. PMID 16149128. Research Blogging.
  5. 5.0 5.1 Cook EF, Goldman L (1984). "Empiric comparison of multivariate analytic techniques: advantages and disadvantages of recursive partitioning analysis". Journal of chronic diseases 37 (9-10): 721-31. PMID 6501544[e]
  6. Lee JW, Um SH, Lee JB, Mun J, Cho H (2006). "Scoring and staging systems using cox linear regression modeling and recursive partitioning". Methods of information in medicine 45 (1): 37-43. PMID 16482368[e]
  7. 7.0 7.1 Fonarow GC, Adams KF, Abraham WT, Yancy CW, Boscardin WJ (2005). "Risk stratification for in-hospital mortality in acutely decompensated heart failure: classification and regression tree analysis". JAMA 293 (5): 572-80. DOI:10.1001/jama.293.5.572. PMID 15687312. Research Blogging.
  8. Stiell IG, Wells GA, Vandemheen KL, et al (2001). "The Canadian C-spine rule for radiography in alert and stable trauma patients". JAMA 286 (15): 1841-8. PMID 11597285[e]
  9. Haydel MJ, Preston CA, Mills TJ, Luber S, Blaudeau E, DeBlieux PM (2000). "Indications for computed tomography in patients with minor head injury". N. Engl. J. Med. 343 (2): 100-5. PMID 10891517[e]
  10. Stiell IG, Greenberg GH, Wells GA, et al (1996). "Prospective validation of a decision rule for the use of radiography in acute knee injuries". JAMA 275 (8): 611-5. PMID 8594242[e]
  11. 11.0 11.1 Goldman L, Weinberg M, Weisberg M, et al (1982). "A computer-derived protocol to aid in the diagnosis of emergency room patients with acute chest pain". N. Engl. J. Med. 307 (10): 588-96. PMID 7110205[e]
  12. Kattan MW, Hess KR, Beck JR (1998). "Experiments to determine whether recursive partitioning (CART) or an artificial neural network overcomes theoretical limitations of Cox proportional hazards regression". Comput. Biomed. Res. 31 (5): 363-73. PMID 9790741[e]
  13. Steinberg, Dan and Phillip Colla. (1995). CART: Tree-Structured Non- Parametric Data Analysis. San Diego, CA: Salford Systems, 39. 

External links