Griffin Health is anchored by the Griffin Hospital in Derby, Connecticut, a 160-person hospital of acute care community serving over 130,000 inhabitants of the Lower Valley of the Naugatuck Valley.
Challenge
Like many healthcare systems, Griffin Health fought to ensure that patients received timely observations when the imaging examination revealed the discovery that it required additional imaging to diagnose the problem or give a clear response to the ordering supplier.
“The results of clear and random radiology often signal the earliest stages of the disease, but in the entire industry about 50% of additional images that have been recommended, is not made,” said Todd Liu, executive vice president and operational director of Griffin Health.
“The causes vary, but include inconsistent transfer, no centralized tracking, impoverished communication and gaps between radiology and basic or specialist teams.
“We had manual systems to notify order providers, but we lacked an effective way of managing and organizing observations, coordination with our patients and monitoring what happened next”, “he contradicted. “It was appropriate Planned check -up test? Did the patient conduct the examination? Was there a diagnosis or a clear response to the reason for the ordering supplier to order an image examination? “
Griffin Health care navigators worked difficult to provide patients with any further imaging and completed tests, but it was not possible to manually manage the load load, and the staff was not always sure that the patient went to meet. It was a safety problem that the healthcare system could no longer afford toleration.
“We needed a solution to this challenge that would help to close this loop, not only at the time of the first report, but until the solution,” said Liu.
APPLICATION
Reflo Health, a seller of further care, turned to Griffin Health with a proposal: if The healthcare system could exploit artificial intelligence to identify radiology reports with the recommended examination of supplementary imaging, automating their escalation in the flows of care coordination and organizing care, Griffin Health can significantly improve compliance with compliance with compliance and safety of patients without overwhelming staff.
“The platform would help identify open recommendations for additional imaging tests in real time, put each patient in a appropriate, predefined care path to observe and piss off our patients and suppliers to act,” Liu explained.
“What made the approach particularly convincing was his flexibility,” he continued. “The system has been designed to connect to our existing processes and increase our internal capacity for tracking and action based on the arrangements – not fully replacing our existing work flows.”
The goal was not only to mark endangered patients, but to automate the range and provide insight into each stage of the process, from identification to closure.
“This helped us ensure the loop closing and there is a mechanism for tracking the supplier’s responsibility, progress and results in one place,” he said.
Fulfillment of the challenge
Griffin Health set Inflo technology into existing care navigation processes. The platform began with Analysis of radiology reports, looking for a language or patterns that indicated a potentially significant discovery, such as a lung lump. After identifying, the platform organized patients in work flows based on the type of consequence and time for the next step in care.
“The platform of an automated supplier, to ensure the order of the recommended next step in EHR, monitored patients’ involvement and determined whether and when observation took place,” Liu explained. “Thanks to this centralized information in one place, our care coordinators can begin to exploit the system for monitoring and identifying any omitted imaging tests, identifying barriers and contacting patients directly if necessary.
“The system gave them a central navigation desktop, which followed the status of each case: preliminary determination, recommended observations, whether this meeting was made and whether the patient finally conducted a control test,” he added.
RESULTS
First of all, the exploit of technology increased the closing rate by 50%.
“In the case of accidental arrangements – which include nodules and other arrangements that have not been clearly evaluated – we were aware of 50% of the improvement in closing indicators,” Liu informed. “This means that more and more patients receive the needed care, and less falls through cracks. This change is huge from risk and patient safety perspectives.”
Secondly, the observation end indicator increased by 17%.
“Our indicator of completion of observation in patients marked increased by 17%,” he said. “This is a significant jump in real patients who were seen, diagnosed and treated earlier. In some cases this is the difference between early catching the disease and the lack of an intervention window.”
Thirdly, patients were included in the screening program for lung cancer based on Recommendations of artificial intelligence.
“Health Inflo has helped us identify and enroll 18 patients to our screening of lung cancer – patients that we would not involve at this critical moment,” he explained. “This is a real life potentially extended or saved, because the AI system appeared in a timely manner, and our teams were prepared for action.”
Tips for others
Explain what problem you are trying to solve – and what success looks like, he advised Liu.
“For us, the priority was to close the loop on control care,” he explained. “We had to follow the arrangements; but more importantly, we needed a system that created useful information. Regardless of the technology chosen, make sure that it not only identifies problems or produce more data – it must make people easier.
“Secondly, do not underestimate the importance of cooperation with suppliers and members of the care team to carefully integrate new tools with existing work flows, if possible,” he continued. “The best artificial intelligence in the world will not matter if your teams do not use it – or if they are not integrated with the way suppliers and employees do their work.”
In the case of Griffin Health, success depends on the choice of a platform that could work with care coordinators and clinicians.
“Good technology should always support basic work processes and flows, and not decide how suppliers and staff provide care for patients,” he concluded.
Watch now: the main AI officers require a deep understanding of OPS technology and clinical technology and clinical