JMIR Publications

ABSTRACT

Background:

The World Health Organization reported that cardiovascular diseases (CVDs) are the number one cause of death globally. CVDs are characteristically chronic, with complex progression patterns involving episodes of several comorbidities and multi-morbidities. When dealing with chronic diseases, physicians often adopt a “watchful waiting” strategy, and actions are postponed until information from an evolving clinical scenario is available. Given a patient’s current state, the ability to predict progression paths can enable effective monitoring leading to timely intervention decisions. Population-level transition probabilities and progression patterns can be revealed by applying time-variant stochastic modeling methods to longitudinal patient data from cohort studies. The resultant computational model can be embedded in interactive decision support tools for clinicians. However, to date, no study has attempted to accomplish this for the CVDs.

Objective:

This study aims to apply advanced stochastic modeling methods to uncover the transition probabilities and progression patterns from longitudinal episodic data of CVD patients and thereafter utilize the computational model to build an interactive clinical decision support artifact demonstrating actionability of such models in monitoring disease progression.

Methods:

Our data was sourced from nine epidemiological cohort studies by the National Heart Lung and Blood Institute (NHLBI) and comprises chronological records of 1274 patients associated with 4839 CVD episodes across 16 years. We then employed the Continuous-time Markov Chain (CTMC) method to develop our model, which offers a robust approach to time-variant transitions between disease states in chronic diseases.

Results:

Our study presents time-variant transition probabilities of CVD state changes, revealing distinct patterns of CVD progression against time. We find that the transition from myocardial infarction (MI) to stroke has the fastest transition rate (mean transition time 3 days), and MI to angina is the slowest (mean transition time 1457 days). Congestive Heart Failure (CHF) is the most probable first episode (44%), followed by stroke (26%). The resultant artifact is actionable as it can act as an eHealth decision support tool, helping the physicians gain critical insights into treatment and intervention strategies by predicting a patient's disease progression path and comparing it against the population pattern visualized by the model.

Conclusions:

Past research does not provide actionable decision support tools based on a comprehensive ten-state CTMC model to unveil complex CVD progression patterns from real-world patient data and support clinical decision-making. This paper aims to address this crucial limitation in extant health informatics literature. Our stochastic model-embedded artifact can help clinicians in efficient disease monitoring and timely treatment decisions guided by unbiased insights from real patient data. Furthermore, the proposed model can unveil progression patterns of any chronic disease of interest by inputting only three data elements: a synthetic patient identifier, episode name, and episode time in days from a baseline date.