Data Visualization and Data Summary - Text Mining Techniques for Healthcare Provider Quality Determination - page 72

dIscussIon
Kernel density estimation to investigate the entire population distribution is important in a healthcare
setting where the distribution is usually not normally distributed. It can be used to examine segments of
the population rather than to focus on averages as most traditional statistical methods do currently. We
need to examine the costs and demands that are related to the most extreme patients.
Kernel density estimation also allows use to compare subgroups within the population because we
can overlay the curves on the same set of axis. As such, it is superior to the use of bar graphs, which de-
pend upon side-by-side comparisons. It is extremely flexible as a technique and can be used to examine
a variety of outcome measures, including length of stay, cost, and patient compliance.
references
Ahrens, T. (1999). Outlier management. Influencing the highest resource-consuming areas in acute and
critical care. Critical Care Nursing Clinics of North America , 11 (1), 107-116.
Anonymous-kde. (2008). Kernel Density Estimators [Electronic Version]. Retrieved from http://homep-
ages.inf.ed.ac.uk/rbf/CVonline/LOCAL_COPIES/AV0405/MISHRA/kde.html.
Anonymous-simpson. (2008). Simpson's Rule [Electronic Version]. Retrieved 2008 from http://math-
world.wolfram.com/SimpsonsRule.html.
Battioui, C. (2007). Cost Models with Prominent Outliers. University of Louisville, Louisville, USA.
Fu, Q., Borneman, J., Ye, J., & Chrobak, M. (2005). Improved probe selection for DNA arrays using
nonparametric kernel density estimation. Paper presented at the Annual International Conference of the
IEEE Engineering in Medicine & Biology Society, Shanghai, China.
Harley, M., Mohammed, M. A., Hussain, S., Yates, J., & Almasri, A. (2005). Was Rodney Ledward a
statistical outlier? Retrospective analysis using routine hospital data to identify gynaecologists' perfor-
mance. BMJ (Clinical Research Ed.) , 330 (7497), 929.
Hong, X., Chen, S., & Harris, C. J. (2008). A forward-constrained regression algorithm for sparse kernel
density estimation. IEEE Transactions on Neural Networks , 19 (1), 193-198.
Novis, D., Walsh, M. K., Dale, J. C., & Howanitz, P. J. (2004). Continuous monitoring of stat and routine
outlier turnaround times: two College of American Pathologists Q-Tracks monitors in 291 hospitals.
Archives of Pathology & Laboratory Medicine , 128 (6), 621-626.
Pfeiffer, K. (1985). Stepwise variable selection and maximum likelihood estimation of smoothing factors
of kernel functions for nonparametric discriminant functions evaluated by different criteria. Computers
and Biomedical Research, an International Journal , 18 (1), 46-61.
Silverman, B. W. (1986). Density Estimation for Statistics and Data Analysis (Monographs on Statistics
and Applied Probability . Boca Raton, FL: Chapman & Hall/CRC.
Weissman, C. (1997). Analyzing intensive care unit length of stay data: problems and possible solutions.
Critical Care Medicine , 25 (9), 1594-1600.