A Tree-Based Decision Model to Support Prediction of the Severity of Asthma Exacerbations in Children
- 229 Downloads
This paper describes the development of a tree-based decision model to predict the severity of pediatric asthma exacerbations in the emergency department (ED) at 2 h following triage. The model was constructed from retrospective patient data abstracted from the ED charts. The original data was preprocessed to eliminate questionable patient records and to normalize values of age-dependent clinical attributes. The model uses attributes routinely collected in the ED and provides predictions even for incomplete observations. Its performance was verified on independent validating data (split-sample validation) where it demonstrated AUC (area under ROC curve) of 0.83, sensitivity of 84%, specificity of 71% and the Brier score of 0.18. The model is intended to supplement an asthma clinical practice guideline, however, it can be also used as a stand-alone decision tool.
KeywordsDecision making Asthma Child Retrospective studies Decision trees
The authors would like to thank William Klement and Morvarid Sehatkar from the School of Information Technology and Engineering, University of Ottawa for help with data analysis and Chris Drummond and Peter Turney from the National Research Council of Canada for advice about contextual normalization. The authors also would like to thank anonymous reviewers for helpful comments and suggestions. Research reported here was supported by NSERC-CIHR grant from the CHRP program.
- 2.Barnett, P. J., and Oberklaid, F., Acute asthma in children: Evaluation of management in a hospital emergency department. Med. J. Aust. 154(11):729–733, 1991.Google Scholar
- 4.National Heart, Lung and Blood Institute, Guidelines for the diagnosis and management of asthma. Available from http://www.nhlbi.nih.gov/guidelines/asthma/index.htm. Last accessed on September 15, 2008.
- 5.Canadian Association of Emergency Physicians, Guidelines for emergency management of paediatric asthma. Available from http://www.caep.ca/template.asp?id-3272F9E47A064ED4820C829B15BB1BCD. Last accessed on September 15, 2008.
- 12.Gorelick, M. H., Stevens, M. W., Schultz, T. R., and Scribano, P. V., Performance of a novel clinical score, the Pediatric Asthma Severity Score (PASS), in the evaluation of acute asthma. Acad. Emerg. Med. 11(1):10–18, 2004.Google Scholar
- 13.Ducharme, F. M., Chalut, D., Plotnick, L., Savdie, C., Kudirka, D., Zhang, X., Meng, L., and McGillivray, D., The pediatric respiratory assessment measure: A valid clinical score for assessing acute asthma severity from toddlers to teenagers. J. Pediatr. 152(4):476–480, 2008. doi: 10.1016/j.jpeds.2007.08.034.CrossRefGoogle Scholar
- 19.Becker, A. B., Nelson, N. A., and Simons, F. E., The pulmonary index: Assessment of a clinical score for asthma. Am. J. Dis. Child. 138(6):574–576, 1984.Google Scholar
- 21.Schuh, S., Johnson, D., Stephens, D., Callahan, S., and Canny, G., Hospitalization patterns in severe acute asthma in children. Pediatr. Pulmonol. 23(3):184–192, 1997. doi: 10.1002/(SICI)1099-0496(199703)23:3<184::AID-PPUL3>3.0.CO;2-O.CrossRefGoogle Scholar
- 22.Smith, M., Iqbal, S. M. S. I., Rowe, B. H., and N’Diaye, T., Corticosteroids for hospitalised children with acute asthma. Cochrane Database Syst. Rev. 1:CD002886, 2003.Google Scholar
- 23.Pearson, S. D., Goulart-Fisher, D., and Lee, T. H., Critical pathways as a strategy for improving care: Problems and potential. Ann. Intern. Med. 123(12):941–948, 1995.Google Scholar
- 25.Abu-Hanna, A., and Lucas, P. J. F., Prognostic models in medicine. AI and statistical approaches. Methods Inf. Med. 40:1–5, 2001.Google Scholar
- 26.Quinlan, R., C4.5: programs for machine learning. Morgan Kaufmann, San Mateo, CA, p. 302, 1993.Google Scholar
- 29.Kohavi, R., and Quinlan, R., Decision-tree discovery. In: Klosgen, W., and Zytkow, J. M. (Eds.), Handbook of Data Mining and Knowledge Discovery. Oxford University Press, New York, pp. 267–276, 2002.Google Scholar
- 31.Beck, J. R., Garcia, M. E., Zhong, M., Georgiopoulos, M., and Anagnostopoulos, G. C., A backward adjusting strategy and optimization of the C4.5 parameters to improve C4.5’s performance. In: Wilson, D., and Lane, H. C., (Eds.), Proceedings of the 21st International Florida Artificial Intelligence Research Society (FLAIRS) Conference (FLAIRS 2008), Coconut Grove, Florida, USA, May 15–17, 2008, p. 35–40.Google Scholar
- 34.Grzymala-Busse, J.W., and Hu, M., A comparison of several approaches to missing attribute values in data mining. In: Ziarko, W., and Yao, Y. Y., (Eds.), Rough Sets and current Trends in Computing. Second International Conference, RSCTC 2000 Banff, Canada, October 16–19, 2000. Revised papers. Berlin: Springer,, p. 378–385, 2000.Google Scholar
- 35.O’Sullivan, D., Elazmeh, W., Wilk, S., Farion, K., Matwin, S., Michalowski, W., Sehatkar, M., Using secondary knowledge to support decision tree classification of retrospective clinical data. In: Ras, Z, Zighed, D, and Tsumoto, S, (Eds.), Mining Complex Data. ECML/PKDD 2007 Third International Workshop, MCD 2007, Warsaw, Poland, September 17–21, 2007. Revised selected papers. Berlin: Springer, p. 238–251, 2008.Google Scholar
- 36.Acuna, E., and Rodriguez, C., The treatment of missing values and its effect in the classifier accuracy. In: Banks, D., House, L., McMorris, F. R., Arabie, P., and Gaul, W., (Eds.), Classification, Clustering and Data Mining Applications. Proceedings of the Meeting of the International Federation of Classification Societies (IFCS), Illinois Institute of Technology, Chicago, 15–18 July 2004. Berlin: Springer, p. 639–648, 2004.Google Scholar
- 43.Li, D., German, D., Lulla, S., Thomas, R. G., and Wilson, S. R., Prospective study of hospitalization for asthma. A preliminary risk factor model. Am. J. Respir. Crit. Care Med. 151(3 Pt 1):647–655, 1995.Google Scholar