Collaboration Fuels Best Practices for Health IT Implementation

For successful health IT implementation, organizations should develop central governing bodies that incorporate the perspectives of frontline workers to form best practices, according to new research published online in the journal Organization Science.

Kate Kellogg, an MIT management professor, conducted an 18-month ethnographic study of a primary care unit in an academic hospital that was in the beginning stages of the implementation of a clinical decision support tool into the EHR. 

The unit’s new clinical decision support technology indicated when patients needed vaccinations, diabetes tests, and pap smears during the preclinical patient routine.

There were six medical teams in the primary care unit, each composed of between seven and nine doctors, three or four nurses and medical assistants, and four or five receptionists.

She found that while the health IT implementation required high-level planning at the organizational level, ensuring the tool’s usability primarily fell on the shoulders of frontline workers, such as medical assistants.

Kellog found that medical assistants were tasked with using the tool’s alerts and communicating with patients. Additionally, when the clinical decision support tool’s suggestion did not align with a doctor's judgment, the medical assistants would be tasked with diving deeper into a patient's medical history.

"The promise of these technologies is that they're going to automate a lot of practices and processes, but they don't do that perfectly,” Kellog, the David J. McGrath Jr. Professor of Management and Innovation at the MIT Sloan School of Management, said in a press release.

“There often need to be people who fill the gaps between what the technology can do and what's really required, and oftentimes it's less-skilled workers who are asked to do that,” Kellog continued.

Kellog’s research revealed three problems regarding health IT implementation.

First, she noted “participation problems” which occurred when some employees, such as medical assistants, did not feel comfortable speaking up about workplace issues related to the new technology.

Additionally, Kellog identified "threshold problems," which involved getting enough people on board to use health IT solutions discovered through local experiments with the new technology.

Lastly, Kellog observed "free rider problems," where doctors benefitted from medical assistants doing a wider range of work tasks but did not follow proposed guidelines to free up medical assistant time.

Therefore, the hospital took an additional step for the effective implementation of the clinical decision support tool: forming a centralized working group surrounding the new technology that included high-ranking and low-ranking representatives alike.

Placing a representative of the hospital's medical assistants on a governing body gave the medical assistant the opportunity to speak on behalf of his or her counterparts to improve the tool’s usability, Kellog said.

"What I found was this local adaptation of digital technology needed to be complemented by a central governing body," Kellogg said. "The central group could do things like introduce technical training and a new performance evaluation system for medical assistants, and quickly spread locally developed technology solutions, such as reprogrammed code with revised decision support rules."

Kellog found that the best way to get doctors on board with the central group’s suggestions was to obtain "provisional commitments" from doctors expressing their willingness to try the best practices, rather than demanding that all doctors follow the recommendations.

Kellog called this practice “experimentalist governance.”

"What experimentalist governance is, you allow for all the local experimentation, you come up with solutions, but then you have a central body composed of people from different levels, and you solve participation problems and leverage opportunities that arise during local adaptation," Kellogg said. “If you just let everyone do everything locally, you're going to see resistance to the technology, particularly among frontline employees.”