Robust Outcome Prediction for Intensive-Care Patients

Abstract

Missing data are a major plague of medical databases in general, and of Intensive Care Unit databases in particular. The time pressure of work in an Intensive Care Unit pushes the physicians to omit randomly or selectively record data. These different omission strategies give rise to different patterns of missing data and the recommended approach of completing the database using median imputation and fitting a logistic regression model can lead to significant biases. This paper applies a new classification method, called robust Bayes classifier, which does not rely on any particular assumption about the pattern of missing data and compares it to the median imputation approach using a database of 324 Intensive Care Unit patients.

Keywords

Emergency Medicine - Incomplete Data - Costs Analysis

• References

• 1 Champion HR, Sacco WJ. Measurement of patient illness severity. Crit Care Med 1982; 10: 552-3.
• 2 Rubin DB. Multiple Imputation for Non-response in Survey. New York: Wiley; 1987
• 3 Ramoni M, Sebastiani P. Robust learning with missing data. Machine Learning, to appear also as Technical Report KMiTR-28. Knowledge Media Institute: The Open University; 1997. Available at http://kmi.open.ac.uk/techreports/kmi-tr-28-abstract.html.
• 4 Ramoni M, Sebastiani P. Robust Bayes classifiers. Technical Report KMi-TR-82. Knowledge Media Institute. The Open University; 1999. Available at http://kmi.open.ac.uk/techreports/kmi-tr-82-abstract.html.
• 5 van der Gaag L. Computing probability intervals under independency constraints. In: Proceedings of the Sixth Annual Conference on Uncertainty in Artificial Intelligence (UAI-90). New York: Elsevier; 1991: 491-7.
• 6 Ramoni M. Ignorant influence diagrams. In: Proceedings of the Fourteenth International Joint Conference on Artificial Intelligence (IJCAI-95). S. Mateo: Morgan Kaufman; 1995: 1808-14.
• 7 Knaus WA, Draper EA, Wagner DP, Zimmerman JE. APACHE II: A severity of disease classification system. Crit Care Med 1985; 13: 818-29.
• 8 Chang RWS, Jacobs S, Lee B. Use of APACHE II severity of disease classification to identify intensive-care-unit patients who would not benefit from total parenteral nutrition. Lancet 1986; 1986i: 1483-7.
• 9 Lemeshow S, Teres D, Pastides H, Avrunin JS, Steingrub JS. A method for predicting survival and mortality of ICU patients using objectively derived weights. Crit Care Med 1985; 13: 519-25.
• 10 McCullagh P, Nelder JA. Generalized Linear Models, 2nd ed. London: Chapman and Hall; 1989
• 11 Castella X, Artigas A, Bion J, Kari A. and The European/North American Severity Study Group.. A comparison of severity of illness scoring systems for intensive care unit patients: Results of a multicenter, multinational study. Crit Care Med 1995; 23: 1327-32.
• 12 Langley P, Iba W, Thompson K. An analysis of Bayesian classifiers. In: Proceedings of the Tenth National Conference on Artificial Intelligence (AAAI-92). San Mateo: Morgan Kaufman; 1992: 223-8.
• 13 Pauker SG, Gorry GA, Kassirer JP, Shwartz WB. Toward the simulation of clinical cognition: Taking a present illness by computer. Am J Med 1976; 60: 981-5.
• 14 Little RJA, Rubin DB. Statistical Analysis with Missing Data. New York: Wiley; 1987
• 15 Rubin DB. Inference and missing data. Biometrika 1976; 63: 581-92.
• 16 Sebastiani P, Ramoni M. Model folding for data subject to nonresponse. In: Proceedings of the Seventh International Workshop on Artificial Intelligence and Statistics (Uncertainty 99). San Mateo: Morgan Kaufman; 1999: 287-92.
• 17 Kyburg HE. Rational belief. Behav Brain Sci 1983; 6: 231-73.
• 18 Hand DJ. Construction and Assessment of Classification Rules. New York: Wiley; 1997
• 19 McQuatt A, Andrews PJD, Sleeman DH, Corruble V, Jones PA. The analyses of head injury data using decision tree techniques. In: Proceedings of the Joint European Conference on Artificial Intelligence in Medicine and Medical Decision Making (AIMDM-99). Heidelberg: Springer; 1999: 336-45.