The Complexity of Survival
0 likes121 views
This image shows the complex clinical pathways and shared events of over 15,000 heart failure patients who survived their final hospital admission. Each journey starts as a single point on the graph's periphery and connects as patients experience similar events, merging into larger points based on diagnosis. The graph was generated from California healthcare data from 2006 to 2011 by mapping patient utilization sequences and merging similar events to show the intricacies of survival.
The Complexity of Survival
- 1. The Complexity of Survival This image shows all different journeys toward surviving a final admission for Heart Failure. Each journey starts at a tiny brilliant point in the periphery of the huge graph network and represents at least 11 people who are traveling together on a shared clinical pathway of similar events toward the center (showed as edges). Surfing their specific paths toward a common destiny of surviving final hospitalization, at points each group joins others and forms bigger points. Points are colored based on general diagnostic group of the primary admission diagnosis that all patients share at each point. We used multilevel map of CSS diagnostic coding for this purpose. We used California HCUP inpatient datasets of 2006 to 2011 of 24 million patients to extract and map information of 153064 heart failure patients who survived a heart failure admission between 2009 and 2011 and had at least one admission before that in our dataset. To generate this graph we started by creating sequences of healthcare utilization events of patients. Then anchoring the latest event in the center of the graph we mapped each sequence on the graph. We merged similar events of patients and created branches each time a sequence was different than existing ones. Finally we simplified the graph as a tree structure. Data is processed using a number of tools, most importantly F# and R programming languages, Quickgraph and igraph libraries and visualized using Cytoscape V3. This image is result of a 2 years project for patient journey analysis (patient trajectory analysis) conducted in Center for Outcomes Research and Evaluation in the university wide Big Data To Knowledge, BD2K, movement. This image shows the complexity of life and surviving and contrasts with the other image that we are applying and shows the decisive and succinct patterns of death March 2015 Abbas Shojaee M.D. Associate Research Scientist in Health Data Science Abbas Shojaee, Yale University