Electronic health record utilization data can be used to identify ad hoc coordinated care teams and patterns in a hospital, a team of Vanderbilt University researchers has found.

The researchers, whose work was published recently in the Journal of the American Medical Informatics Association, developed a novel data mining pipeline of EHR utilization records to determine whether the data could identify unrecognized, informal collaborative care teams. They used the pipeline, studying four months’ worth of EHR data, then conducted two evaluations of the system to determine the plausibility of the inferred teams and to determine the collaborative strength and clinical credibility of each inferred team.

They found that the healthcare system was “highly collaborative” and that the utilization records can be translated into knowledge relevant for the definition of collaborative networks. They discovered 34 collaborative care teams, of which 27 were administratively plausible. Moreover, 26 of the 27 plausible teams were “strong” collaborations. They also worked with several oncologists to develop an example of how certain teams worked together.

However, the authors noted that their research was somewhat limited and needed further review. For instance, to make specific team recommendations, more than four months of data needed to be evaluated.

“Inferred collaborative teams are plausible; translating such patterns into optimized collaborative care will require administrative review and integration with management practices,” they said.