Using an optimized generative model to infer the progression of complications in type 2 diabetes patients

Abstract

Background People live a long time in pre-diabetes/early diabetes without a formal diagnosis or management. Heterogeneity of progression coupled with deficiencies in electronic health records related to incomplete data, discrete events, and irregular event intervals make identification of pre-diabetes and critical points of diabetes progression challenging. Methods We utilized longitudinal electronic health records of 9298 patients with type 2 diabetes or prediabetes from 2005 to 2016 from a large regional healthcare delivery network in China. We optimized a generative Markov-Bayesian-based model to generate 5000 synthetic illness trajectories. The synthetic data were manually reviewed by endocrinologists. Results We build an optimized generative progression model for type 2 diabetes using anchor information to reduce the number of parameters learning in the third layer of the model from $$O\left(N\times W\right)$$ O N × W to $$O\left((N-C)\times W\right)$$ O ( N - C ) × W , where $$N$$ N is the number of clinical findings, $$W$$ W is the number of complications, $$C$$ C is the number of anchors. Based on this model, we infer the relationships between progression stages, the onset of complication categories, and the associated diagnoses during the whole progression of type 2 diabetes using electronic health records. Discussion Our findings indicate that 55.3% of single complications and 31.8% of complication patterns could be predicted early and managed appropriately to potentially delay (as it is a progressive disease) or prevented (by lifestyle modifications that keep patient from developing/triggering diabetes in the first place). Conclusions The full type 2 diabetes patient trajectories generated by the chronic disease progression model can counter a lack of real-world evidence of desired longitudinal timeframe while facilitating population health management.

Translational analysis of data science and causal learning in real-world clinical evaluation of traditional Chinese medicine

“…Causal learning can help us deeply understand the real causal relationship in data, support clinical decision making and intervention measures, and even predict the expected responses of diseases and syndromes when changing interventions or making clinical decisions. Causal learning methods include categories of regression analysis methods, [61] causal inference methods [79][80][81][82] (e.g., propensity score under potential outcome framework, instrumental variables, and causal mediation analysis), causal relationship learning methods [83,84] (e.g., causal discovery algorithm, causal probabilistic graphical model, and causal graph learning), and causal representation learning methods [31,85,86] (e.g., causal analysis network, causal sparse coding, causal generative models or causal adversarial networks, and causal feature variable selection). Among these methods, the causal mediation analysis under the potential outcome framework is used to study the definition, recognizability, and estimation of ATE, allowing multiple mediators [80,87] to evaluate the pathways through which complex TCM prescriptions achieve therapeutic effects.…”

“…[30] Third, TCM big data analysis methods can quickly conduct data description, visualization, and modeling, as well as mine the explicit knowledge and implicit knowledge of TCM, making TCM studies understandable, acceptable, quantifiable, and empirical. [31] Fourth, the introduction of intelligent, rigorous, and systematic scientific methods in network science, information science, and complex systems science [32] will facilitate in-depth research on the clinical effects and mechanisms of TCM.…”

Global research trends of diabetes remission: a bibliometric study

“…In recent years, the incidence of diabetes has escalated precipitously ( 1 , 2 ), and its detriment to humans surpasses that of cardiovascular diseases, ascertaining its position first among the world’s top ten chronic diseases ( 3 ). Diabetes inflicts considerable physiological ( 4 – 7 ), psychological ( 8 – 10 ), and economic ( 3 ) adversities to patients. The cognizance rate of diabetes remains low ( 11 ), and an increasing number of young individuals are succumbing to it ( 12 ).…”

Remission effect of Canagliflozin in patients with newly diagnosed type 2 diabetes mellitus: a protocol for a multicenter, parallel-group, randomized, controlled, open-label trial

“…In recent years, the prevalence of diabetes has risen sharp [ 20 , 21 ], which causes great physiological [ 22 , 23 ], psychological [ 24 , 25 ] and economic [ 21 ] harm to patients. After achieving the remission of diabetes through relevant measures, the life quality of patients can be improved, the incidence of complications and concomitant diseases can be reduced [ 26 ], the physical and mental health of patients is benefited, and the economic burden of families as well as society is reduced.…”