Search for contacts, projects,
courses and publications

The relational micro-structure of critical care transfer networks



Lomi A.



When multiple hospitals transfer patients to multiple other hospitals they create a patient transfer network (PTN). These relational arrangements between health care organizations have been shown to play a major role in the organization and delivery of critical care services - a context where inter-hospital patient transfer is common, but not well understood. The quality of critical care that a health system can effectively deliver depends delicately on relational coordination between heath care organizations. The objective of the research that is being proposed is to improve our understanding of the relation between the structure of PTN and patient-level outcomes in the context of critical medicine. Ultimately we want to be able to ascertain what network structures and processes are more likely to be beneficial for the patient. Three main questions orient our efforts: Are inter-hospital patient transfer networks moving patients systematically towards better care? What other factors drive inter-hospital patient mobility? Finally, what are the implication of PTN for patient-level outcomes? We address these questions using statewide data that we have collected on every patient transfer event ever observed between every hospital in Massachusetts during the period 2007-2015. The data at our disposal include comprehensive demographic, clinical, and outcome data for stroke patients in the state. In the empirical part of our study we focus on patient-transfer networks arising from inter-hospital transfer of ischemic stroke patients – the most common form of stroke for which patient transfer is common as much as it is critical. State-of-the art relational event models allow us to specify and estimate stochastic models for time-stamped sequences of patient transfer events that incorporate information on the sender and receiver hospitals, on the patients being transferred, and on the complex dependencies induced by network-like mechanisms like, for example, reciprocity and transitive closure. Relational event models will also assist in specifying and testing how different principles of inter-organizational bonding affect patient transfer and, ultimately, patient outcomes. The study promises to contribute new evidence on the concrete benefits that relational coordination among health care organizations may bring to patients.

Additional information

Start date
End date
3 Months
Funding sources
Swiss National Science Foundation / Scientific Exchanges