1 Assessment of Individual Risk of Death Using Self-report Data: an Artificial Neural Network Compared to a Frailty Index Xiaowei Song, PhD, 1 Arnold Mitnitski, PhD, 2,3 Chris MacKnight, MD, MSc, 1,4 and Kenneth Rockwood, MD 1,4 1 Geriatric Medicine Research Unit, Queen Elizabeth II Health Sciences Centre, Halifax, NS, B3H 2E1, Canada, 2 Faculty of Computer Science, Dalhousie University, Halifax, NS, B3H 2Y9, Canada, 3 Department of Medicine, Dalhousie University, Halifax, NS, B3H 1V7, Canada, 4 Division of Geriatric Medicine, Department of Medicine, Dalhousie University, Halifax, NS B3H 2E1, Canada ABSTRACT OBJECTIVES: To evaluate the potential of an Artificial Neural Network (ANN) in predicting survival in elderly Canadians from self-report data. DESIGN: Cohort study with up to 72 months follow-up. SETTING: Forty self-reported characteristics were obtained from the community sample of the Canadian Study of Health and Aging. An individual frailty index score was calculated as the proportion of deficits experienced. For the ANN, randomly selected participants formed the training sample to derive relationships between the variables and survival and the validation sample to control over-fitting. An ANN output was generated for each subject. A separate testing sample was used to evaluate the accuracy of prediction. PARTICIPANTS: 8,547 Canadians aged between 65 and 99 years old, of whom 1,865 died during 72 months of follow up. MEASUREMENTS: The output of an ANN model was compared with an un-weighted frailty index in predicting survival patterns using Receiver Operating Characteristic (ROC) curves. RESULTS: The area under the ROC curve was 86% for the ANN, and 62% for the frailty index. At the optimal ROC value, the accuracy of the frailty index was 70.0%. The ANN accuracy rate over 10 simulations in predicting the probability of individual survival was 79.2 ± 0.8% CONCLUSION: An ANN improved accuracy of survival classification compared with an un-weighted frailty index. The data suggest that the concept of biological redundancy might be operationalized from health survey data. KEY WORDS: Aging; artificial neural network; frailty index; survival prediction.