A comparison of Cox Regression and neural networks for risk stratificationin cases of acute lymphoblastic leukaemia in children

For most diseases there is considerable interest in the problem of classifi cation, both in relation to medical diagnosis and for prognosis. Multivaria te statistical methods are conventionally used as an aid to clinical decisi on making. Neural Networks (NNs) offer an alternative approach to this type of classification problem. Exploiting 1271 cases from the United Kingdom M edical Research Council UKALL X trial for childhood Acute Lymphoblastic Leu kaemia (ALL), cases were stratified as 'high risk' or 'standard risk' using both the survival analysis technique of Cox Regression and trained neural networks. Based on 10 random trials with a further 300 cases, and predictin g overall Jive year survival from age, sex and white cell count only, there was no significant difference between the two approaches in terms of mean Receiver Operating Characteristic area, though the regression model was sli ghtly superior to a single neural network at high sensitivity (Wilcoxon sig ned rank test; p = 0.033). A composite of two networks, one of which includ ed additional prognostic factors, restored the position of no significant d ifference. It was concluded that in the UKALL X dataset, factors predictive of outcome are fully described by a Cox regression analysis, and that a ne ural network-based analysis identified no additional prognostic features. T he value of the network analysis lay in suggesting that the maximum amount of prognostic information has been extracted from the database.