|
In 1999 Carter et al evaluated 5 post-lumpectomy treatment options for a 74yo women with early stage breast cancer, using 3 modelling methods: Markov Process, Analytic Hierarchy Process and Analytic Network Process. They returned the same 'best' option.  | KJ Carter, NP Ritchey, F Castro, LP Caccamo, E Kessler, BA Erickson Analysis of three decision making methods: a breast cancer patient as a model Medical Decision Making 19: 49-57 |
The AHP results are represents in the Annalisa snapshot below. - Observe
- Radiation
- Tamoxifen
- Radiation plus Tamoxifen
- Mastectomy
- Local recurrence
- Metastasis
- Pain
- Disfigurement
- Inconvenience
- Death
- Seroma
- Thrombosis
- Rib fracture (omitted in this Annalisa)
- Secondary Cancer
- Pulmonary thrombosis
From Carter et al. They were produced by the 'strength of preference' approach adopted in the AHP methodology and may bear no necessary relationship to direct  Ratings from the evidence. In any case they may now be out of date, so as usual the example is illustrative of the way Annalisa may be used. Note: We assume the published rating for Radiation/Local recurrence of .029 should be .249From Carter et al. using AHP 'strength of preference' method. From Carter et al. Radiation plus Tamoxifen emerged as top-ranked in all three methods and Mastectomy ranked last in all three. The other rankings varied between the methods. Carter et al. concluded that "the choice of one approach over another depends upon the urgency of the decision and the detail desired from the analysis" 
|
