DeepBlip: Estimating Conditional Average Treatment Effects Over Time
This provides a more efficient and interpretable method for estimating treatment effects over time in clinical settings, though it is incremental as it adapts existing SNMMs with neural networks.
The paper tackled the lack of neural frameworks for structural nested mean models (SNMMs) by proposing DeepBlip, which uses a double optimization trick to enable end-to-end training and achieves state-of-the-art performance on clinical datasets.
Structural nested mean models (SNMMs) are a principled approach to estimate the treatment effects over time. A particular strength of SNMMs is to break the joint effect of treatment sequences over time into localized, time-specific ``blip effects''. This decomposition promotes interpretability through the incremental effects and enables the efficient offline evaluation of optimal treatment policies without re-computation. However, neural frameworks for SNMMs are lacking, as their inherently sequential g-estimation scheme prevents end-to-end, gradient-based training. Here, we propose DeepBlip, the first neural framework for SNMMs, which overcomes this limitation with a novel double optimization trick to enable simultaneous learning of all blip functions. Our DeepBlip seamlessly integrates sequential neural networks like LSTMs or transformers to capture complex temporal dependencies. By design, our method correctly adjusts for time-varying confounding to produce unbiased estimates, and its Neyman-orthogonal loss function ensures robustness to nuisance model misspecification. Finally, we evaluate our DeepBlip across various clinical datasets, where it achieves state-of-the-art performance.