Center for Interdisciplinary Research in Pediatric Critical Illness & Injury (CIRPCII)
Interdisciplinary research in pediatric critical care
 |
Heat map
|
Children with critical illness pose challenges at multiple levels in academic medicine. They comprise the most medically complex of patients, requiring substantial resources for their care. Management decisions must be made quickly, often with little or no published data to guide such management. In many areas of critical care therefore, clinical practice is empirical. There is insufficient understanding of molecular mechanisms of injury, basic pathophysiology, or indeed the mechanisms of repair. A major challenge to these formidable problems is the discipline-based approach to research and funding. One major obstacle to progress is the organization by discipline or disease.
The intent of the CIRPCII in supporting the Xenobase-Critical Care program is to attempt to overcome these barriers by enabling studies related to critical illness by incorporating data from all ICU admissions over a 14 year period and by enabling analyses which can address issues ranging from questions regarding access to care, socioeconomic status, physiology, costs of care, and outcomes. Xenobase is a bioinformatics software developed by Van Andel Institute, Grand Rapids, MI. The Xenobase database contains data from all PICU admissions since 1993. Included in these data are physiology, laboratory, administrative and demographic information for over 13,000 admissions. These data and the analytic tools available with Xenobase should be of use to investigators with diverse interests in pediatric critical care. Cohorts may be selected for analysis for example based on administrative data (procedures), demographics (census tract data), physiology or laboratory data. These analyses may be hypothesis driven or hypothesis-generating, using cluster analyses or multi-dimensional scaling. Examples of these approaches are shown with a cluster analysis of all intubated patients, and multidimensional scaling. In addition to using these tools of genetic analyses for large datasets, the investigator can also test hypotheses by generating and testing new prediction tools such as the ROC also shown on the right.
|
 Time course analysis
|
Multidimensional scaling 3D
|
ROC curve
|
Venn Diagram
|