Predicting diabetes second-line therapy initiation in the Australian population via timespan-guided neural attention network

Abstract

IntroductionThe first line of treatment for people with diabetes is metformin. However, over the course of the disease metformin may fail to achieve appropriate glycemic control, and a second-line therapy becomes necessary. In this paper we introduce Tangle, a timespan-guided neural attention model that can accurately and timely predict the upcoming need for a second-line diabetes therapy from administrative data in the Australian adult population. The method could be used to design automatic therapy review recommendations for patients and their providers without the need to collect clinical measures. DataWe analyzed seven years of deidentified records (2008-2014) of the 10% publicly available linked sample of Medicare Benefits Schedule (MBS) and Pharmaceutical Benefits Scheme (PBS) electronic databases of Australia. MethodsBy design, Tangle can inherit the representational power of pre-trained word embedding, such as GloVe, to encode sequences of claims with the related MBS codes. The proposed attention mechanism can also natively exploit the information hidden in the timespan between two successive claims (measured in number of days). We compared the proposed method against state-of-the-art sequence classification methods. ResultsTangle outperforms state-of-the-art recurrent neural networks, including attention-based models. In particular, when the proposed timespan-guided attention strategy is coupled with pre-trained embedding methods, the model performance reaches an Area Under the ROC Curve of 90%, an improvement of almost 10 percentage points over an attentionless recurrent architecture. ImplementationTangle is implemented in Python using Keras and it is hosted on GitHub at https://github.com/samuelefiorini/tangle.