Information Technology Reference
In-Depth Information
models remain dominant in the methodology. Since it is clear that the threshold value is defined by the
level 5 mortality index for disease staging, it would be a relatively simple matter to shift patients into
level 5 to increase the prediction of mortality.
The disease staging indices appear to assume a lesser degree of severity in the patient population
compared to the previous Charlson and APRDRG indices. For this reason, a significant portion of the
patient population is in the highest level of severity. Moreover, for patients undergoing cardiovascular
surgery, none are classified in the lowest two index levels. The main problem here with a substantial
portion of the patients in the highest levels is that there is no ability to classify the sickest patients into
a higher category of severity, and the sickest patients are combined into the same class level as patients
with more moderate conditions. Therefore, predictions for length of stay and total charges will not reach
as high as they should given the gamma distributions of the outcome variables. These disease staging
models have difficulty with predicting true patient outliers.
references
Kotelchuck, M. (1994). An evaluation of the Kessner adequacy of prenatal care index and a proposed
adequacy of prenatal care utilization index. American Journal of Public Health , 84 (9), 1414-1420.
doi:10.2105/AJPH.84.9.1414
Mapel, D. W., Picchi, M. A., Hurley, J. S., Frost, F. J., Petersen, H. V., & Mapel, V. M. (2000). Utili-
zation in COPD: patient characteristics and diagnostic evaluation. Chest , 117 (5Suppl 2), 346s-353s.
doi:10.1378/chest.117.5_suppl_2.346S
Thomas, E. J., Goldman, L., Mangione, C. M., Marcantonio, E. R., Cook, E. F., & Ludwig, L. (1997).
Body mass index as a correlate of postoperative complicatons and resource utilization. The American
Journal of Medicine , 102 (3), 277-283. doi:10.1016/S0002-9343(96)00451-2
Tu, W., Morris, A. B., Li, J., Wu, J., Young, J., & Brater, D. C. (2005). Association between adherence
measurements of metoprolol and health care utilization in older patients with heart failure. Clinical
Pharmacology and Therapeutics , 77 , 189-201. doi:10.1016/j.clpt.2004.10.004
Ville, B. D. (2006). Decision Trees for Business Intelligence and Data Mining: Using SAS Enterprise
Miner . Cary, NC: SAS Press.
Search WWH ::




Custom Search