It happened about two weeks ago, when the coronavirus outbreak had started to reach emergency status in the Seattle area. UW Medicine — a large healthcare system that includes three hospitals, a network of practitioners, clinics and a school of medicine — had determined a series of protocols to help expedite IT-related preparation for this unprecedented event. Without hesitation, leadership launched an online resource through which best practices could be shared with healthcare facilities around the world.
“We had this ‘a-ha’ moment where we realized that we’re ground zero,” said Eric Neil, CIO with UW Medicine IT Services. But instead of compiling a presentation based on lessons learned that could be shared months down the road, his team wanted to get in front of it. “We thought, let’s share the information now.”
They did, and not just through the website. Recently, Neil spoke with healthsystemCIO.com, along with Kelly Summers, Chief Nursing Informatics Officer, and Stephanie Klainer, Program Manager, IT Services Emergency Management. During a conference call, UW Medicine’s leaders described the strategy and tactics that have been utilized to maintain continuity.
Part 1
According to Klainer, the IT response is organized around three primary areas: people, EHR/clinical, and technology. But it starts with establishing — or, in some cases, activating — an incident command structure. In UW Medicine’s case, it meant standing up a series of individual sites when the reported up to an enterprise-level command center.
“We have a storied history of being able to leverage this capability,” she added. Once the enterprise center was live, decisions had to be made quickly in terms of how IT services would handle this particular situation.
Support of Clinical/EHR Response
With any type of disaster — particularly one that involves an influx of patients — adjustments must be made to support clinicians and patient care, including changes to EHRs and other systems. Some change requirements included new orders and alerts, precautions for blood processing, and standard phrasing needed for documentation. Because of the unique nature of this pandemic, requests have increased significantly, along with expectations for quicker turnaround times, according to Summers.
“We have individuals available 24/7/365 to make changes to any of our clinical and business systems on an expedited change request process,” she said. “We’ve gone into an emergency change request control model for anything related to this particular COVID response.”
For organizations dealing with these challenges, Summers and colleagues provided the following recommendations:
- Establish a process for keeping your COVID-19 documentation updated 24/7. Include IT representation in your incident command, so they can quickly operationalize new guidance by incorporating it into your EHRs and other systems.
- Monitor EHR vendor sites for emergent system updates.
- Establish a process for expedited change requests to the EHR.
- Plan to have certain IT employees staffed around the clock who can evaluate and implement these change requests.
- Communications to the workforce for these EHR updates are managed through our Crisis Communications Team in the Enterprise Command Center. Be sure you have a process that works for your organization.
Summers also advised involving vendors in the process as much as possible. “We’ve made numerous changes to our clinical documentation and results reporting structure in collaboration with our EHR vendors to make sure all of our clinicians have the tools that they need to take care of patients,” she added. “We don’t want them to have to worry about the documentation tools supporting them.”
Another critical piece in supporting clinical practitioners is ensuring updated metrics are accessible in a centralized location. At UW Medicine, IT Services created a dashboard for incident command leadership that included vital metrics such as number of tests by result and by facility per day, lab turnaround time, number of current and positive pending tests among inpatients, and more. As the virus spread, information was updated, which proved critical as the organization prepared for an influx of patients requiring ICU care. To more effectively track hospitalizations, IT Services built a table within the enterprise data warehouse which provides testing information from the lab and connects it with ADT data.
The final element in supporting clinical work is maintaining system stability. This means IT Services Leadership must continually evaluate the timing for applying patches or fixes, and must be willing to reprioritize other critical IT work, UW Medicine leaders stated.
Technical Support
The second thread of an effective IT response plan is technical support, which is always critical, but becomes even more important as care extends beyond the traditional four walls.
“We’re doing everything we can to make it easy to access care,” said Neil, even if it means directing patients away from hospitals and clinics and setting up drive-through testing sites. UW Medicine was an early adopter of the mobile clinic concept, which he believes is a credit to the organization’s strong focus on agility and innovation.
“Our mindset was going in one direction, and then almost overnight, things changed. Now we’re working with hundreds of thousands of patients to push information to them and to teach them how to follow a new screening procedure.”
Of course, it also requires having a solid infrastructure in place. According to Summers, a key factor in UW Medicine’s ability to convert an employee parking garage into a testing site was its strong wireless network, which extended outside the hospital. “We were able to create in our scheduling and EHR system a remote location for one of our primary care clinics, and handle the workflow just as if the patient walked into the clinic. We even used the same workflows.”
Depending on the results of an entrant screening process, a patient may be evaluated by a clinician and sent to a triage nurse. An appointment is then made, and the patient proceeds to the clinic just as he or she would any other appointment. After the patient’s ID is confirmed, a care provider performs a swab, collects the specimen, and provides any information the patient might need. The best part? Patients never have to leave the car, noted Summers.
UW Medicine also set up a secondary clinic in a sports medicine trailer; this one, however, fell outside of the network’s range, and so IT Services created a “network in a box” that can be spun up at any location. “We’re able to maintain all of these visits within our EHR,” she said. “It’s a seamless care delivery model.”
Interestingly, it wasn’t a viral outbreak — or any medical emergency, in fact — that drove UW Medicine to shore up its network, but rather, power outages that frequently impact the Pacific Northwest. A few years ago, the organization purchased cellular hotspots, which have come in handy quite a bit in recent weeks.
“If a clinic loses connectivity, the IT department can rapidly go to that clinic and set up hotspots and connect the desktops and printers so they can resume normal operations,” said Neil. “They’re not a long-term solution, but they’re great in the interim.”
And it didn’t stop there. Searching for a more robust connecting, the team utilized a microwave radio relay system in which microwaves are transmitted on a line of sight path between points using directional antennas. The solution has worked so well that it has been incorporated into the organization’s disaster recovery strategy, he noted.
“The ability to stand up a clinic using tents and trailers in a parking lot is not something we had really anticipated,” Neil added. “We will, going forward.”
In the next installment of this two-part series, UW Medicine IT Services leaders will discuss the third — and perhaps, most important — thread of disaster preparation and recovery: the human element. To access UW Medicine’s online resource site, click here. To follow the organization on Twitter, click here.
Part 2 Coming Soon…
Interview Sponsored by:
Share Your Thoughts
You must be logged in to post a comment.