Date: Monday, October 22, 2018
Session Title: Measures and Measurement of Healthcare Quality Poster II
Session Type: ACR Poster Session B
Session Time: 9:00AM-11:00AM
Background/Purpose: Combination therapy with DMARDs for treating RA is considered as standard of care. However, certain rates of adverse events (AEs) are unavoidable. The stigma is which drug should be stopped first once AEs emerge and the following sequence if AEs persist for optimal risk reductions. The decisions made by clinicians are always empirically. The purpose of this study is to develop an algorithm for decision making on drug withdraw sequence in face of adverse events with combination therapy based on data mining and machine learning from the smart system of disease management (SSDM).
Methods: SSDM is an interactive mobile disease management tool, RA patients can input medical records (including medication and laboratory test results) and perform self-evaluation via applications (App). The data synchronizes to the mobiles of authorized rheumatologists through cloud and advices could be delivered.
In order to develop the master algorithm, abnormal white blood cell counts (WBC) and alanine aminotransferase (ALT) elevation were targeted. WBC, ALT and medication data was collected, extracted, validated, and then based on Bayesian networks, data mining, modeling, calculating, analyzing were performed. WBC under 4,000/ml is defined as leukocytopenia (LP), over 10,000/ml as infection predisposing (IP), and ALT > 40 U/L as ALT elevation.
Results: From June 2014 to June 2018, 32,130 RA patients from 587 centers registered in SSDM. 7,086 are male and 24,144 are female with mean age of 49.82 year. 129 different drugs and 479 types of combination therapies are identified. Lab test results showed LP happened in 311 and IP 217, ALT 316 in 554 mono or combinational treatment regiments. Among them we selected prednisone (Pred), leflunomide (LEF), MTX, HCQ as an example to develop a master algorithm based on Bayesian networks and learning model. Image 1 shows Bayesian network and data processing, in which, quartet are correlating with 15 different regiments. Drug withdraw sequence for LP is HCQ, then LEF and then Pre, and the risks of LP are reduced by 39%, 33% and 23%, respectively. For IP, withdraw sequence is Pred, then MTX and then HCQ, and the risks of IP are reduced by 47%, 51% and 15%, respectively, For ALT, withdraw sequence is MTX, then Pred and then HCQ, and the risks of ALT are reduced by 51%, 28% and 16%, respectively.
Conclusion: Big data system can be built using SSDM via empowering patient. Through data mining, networking, modeling, and Bayesian calculation, a master algorithm for drug withdraw strategy in reduction of adverse events with combination therapy is developed, which can be applied on the other AEs in SSDM and may replicated in other diseases. Following the continuing data inputs and machine leaning, an artificial intelligent system in assisting clinical decision making may be achieved.
Limitations: This study only focus on rate of AE without considering the efficacy, without stratifying dosing.
To cite this abstract in AMA style:Zhao Y, Yang J, Huang J, Wei H, Wang Y, Mu R, Zuo X, Wang H, Duan X, Xue J, Sun H, Wu B, Kang L, Wei F, Mi C, Zhao Y, Li Y, Chen H, Li Z, Meng Q, Jia Y, Xiao H, Xiao F. Develop a Master Algorithm for Drug Withdraw Strategy in Reduction of Adverse Events – a Machine Learning Model from the Smart System of Disease Management (SSDM) [abstract]. Arthritis Rheumatol. 2018; 70 (suppl 10). https://acrabstracts.org/abstract/develop-a-master-algorithm-for-drug-withdraw-strategy-in-reduction-of-adverse-events-a-machine-learning-model-from-the-smart-system-of-disease-management-ssdm/. Accessed January 17, 2020.
« Back to 2018 ACR/ARHP Annual Meeting
ACR Meeting Abstracts - https://acrabstracts.org/abstract/develop-a-master-algorithm-for-drug-withdraw-strategy-in-reduction-of-adverse-events-a-machine-learning-model-from-the-smart-system-of-disease-management-ssdm/