From Real‐World Patient Data to Individualized Treatment Effects Using Machine Learning: Current and Future Methods to Address Underlying Challenges

Abstract

Clinical decision making needs to be supported by evidence that treatments are beneficial to individual patients. Although randomized control trials (RCTs) are the gold standard for testing and introducing new drugs, due to the focus on specific questions with respect to establishing efficacy and safety vs. standard treatment, they do not provide a full characterization of the heterogeneity in the final intended treatment population. Conversely, real‐world observational data, such as electronic health records (EHRs), contain large amounts of clinical information about heterogeneous patients and their response to treatments. In this paper, we introduce the main opportunities and challenges in using observational data for training machine learning methods to estimate individualized treatment effects and make treatment recommendations. We describe the modeling choices of the state‐of‐the‐art machine learning methods for causal inference, developed for estimating treatment effects both in the cross‐section and longitudinal settings. Additionally, we highlight future research directions that could lead to achieving the full potential of leveraging EHRs and machine learning for making individualized treatment recommendations. We also discuss how experimental data from RCTs and Pharmacometric and Quantitative Systems Pharmacology approaches can be used to not only improve machine learning methods, but also provide ways for validating them. These future research directions will require us to collaborate across the scientific disciplines to incorporate models based on RCTs and known disease processes, physiology, and pharmacology into these machine learning models based on EHRs to fully optimize the opportunity these data present.

Precision Dosing: The Clinical Pharmacology of Goldilocks

“…As with other complex multi-variate challenges, machine learning approaches are being developed to tackle this problem, and they bring the possibility of determining the optimal doses from real-world data after drug approval. 10…”

“…This will take several years to have a clinical effect but, for drugs already on the market, there is no need to wait. There is already an opportunity for model‐informed precision dosing leveraging real‐world data 9 perhaps combined with novel machine learning techniques 10 and a comprehensive framework for enabling precision dosing for approved drugs and those newly in development is presented in this issue 22 . The current COVID pandemic represents an opportunity for precision dosing, 23 and it is to be hoped the urgency of the situation will bring the key stakeholders together to find ways to make precision dosing a broader reality that will subsequently also benefit many other diseases and drugs.…”

Clinical Pharmacology and Therapeutics: 2020 in Review

“…2 This was also the first example of an issue fully dedicated to a particular theme, including all original research articles. It also introduced the first CPT Tutorials, 3,4 a new article type for the journal providing practical educational material on tools, methodologies, and approaches in clinical pharmacology [5][6][7] ( Table 1).…”

A Novel Intervention Recurrent autoencoder for real time forecasting and non-pharmaceutical intervention selection to curb the spread of Covid-19 in the world

“…The RIN consists of an encoder and a decoder. Both of the encoder and decoder use the RNN as their basic architectures (Bica et al 2020).…”