Dynamic Transitions of Pediatric Sepsis: A Markov Chain Analysis

Abstract

Pediatric sepsis is a heterogeneous disease with varying physiological dynamics associated with recovery, disability, and mortality. Using risk scores generated from a sepsis prediction model to define illness states, we used Markov chain modeling to describe disease dynamics over time by describing how children transition among illness states. We analyzed 18,666 illness state transitions over 157 pediatric intensive care unit admissions in the 3 days following blood cultures for suspected sepsis. We used Shannon entropy to quantify the differences in transition matrices stratified by clinical characteristics. The population-based transition matrix based on the sepsis illness severity scores in the days following a sepsis diagnosis can describe a sepsis illness trajectory. Using the entropy based on Markov chain transition matrices, we found a different structure of dynamic transitions based on ventilator use but not age group. Stochastic modeling of transitions in sepsis illness severity scores can be useful in describing the variation in transitions made by patient and clinical characteristics.

Case report: evaluation of an open-source synthetic data platform for simulation studies

“…Microsimulation models have been used for real-world applications in forecasting policies 6 , 7 and assessing disease progression trajectories and intervention scenarios. 25 This demonstration shows that Synthea can be repurposed for simulation studies comparable to the reference study without advanced programming. Although workarounds were required, replication was feasible.…”

Beyond prediction: Off‐target uses of artificial intelligence‐based predictive analytics in a learning health system

“…We view this as a missed opportunity. In this case series, we demonstrate how visual AI‐based predictive analytics that were trained on events of clinical deterioration can be used in off‐target ways when the score represents underlying physiological stability 12,35 . It is not feasible to develop and train AI‐based models on every possible event or model outcome that clinicians may view as useful, but it is incredibly important to elicit feedback following implementation on how clinicians have incorporated the model output into their workflows 13 .…”

“…In this case series, we demonstrate how visual AI‐based predictive analytics that were trained on events of clinical deterioration can be used in off‐target ways when the score represents underlying physiological stability. 12 , 35 It is not feasible to develop and train AI‐based models on every possible event or model outcome that clinicians may view as useful, but it is incredibly important to elicit feedback following implementation on how clinicians have incorporated the model output into their workflows. 13 It is critical to collect long‐term data on uses, clinical actions, and quality outcomes within a learning health system cycle feedback to determine if further study or additional validations are needed due to data reasons (data drift, missingness).…”

Selection of the best operational amplifier based on the statistics of the complex moments

“…Markov chains: It allows in evaluation of an array of data in which the state of the next element depends only on the state of the previous ones [20][21][22]. It is widely used in probability theory in describing various information processes.…”