Striving For Zero

By Sarah Talpos | 2 August

Each year, Department of Pathology laboratories perform millions of clinical tests. For each test, a care provider places an electronic order, and then waits for the results. It may sound simple, but in a large health system like Michigan Medicine’s, with over two dozen laboratories that receive a wide array of specimens—blood, urine, cervical cells, sputum, stool, moles, and more—there’s always a small chance that a given sample could go missing along the way.

Meet the Team

Amy Mapili acts as the project manager and coordinates with both end users and the technical team, where she ensures that the processes followed within the different laboratories are used to inform the development of the application.
John Hamilton is the technical lead and takes the requirements of business logic, user experience, and technical dependencies and combines the multiple disciplines of the team while ensuring each component is factored into the final vision of the application.
Bill Hubbard and Todd Kandow collaborate as the database architects and developers. By connecting the IBM interface engine to the PathTrack database, they are able to find, store, format, and push the clinical patient data needed to the web server.
Brian Royer is the main developer and uses React.js to deliver high-volume data to a responsive interface using advanced algorithms. He enables PathTrack to seamlessly deliver secure payloads of data and create a robust user experience, all while maintaining real-time feedback to the user.

“We’re striving for zero defects,” says Duane Newton, PhD, associate director of the Division of Clinical Pathology. He contrasts this with other industries, such as the automotive industry, where “low defects” is considered acceptable. Meeting the Department’s goal will be a challenge, he says, but at the end of the day, it’s important to get as close to zero missing specimens as possible.

The Department’s upcoming move to the North Campus Research Complex will add an additional layer of complexity to the movement of specimens. Given this, in 2017, the Division of Quality and Health Improvement (DQHI) spearheaded the Patient Asset Management Initiative in order to develop an approach to precisely track specimens as they travel throughout the health system. Early in the process, the project team reached an important conclusion: The Department would need to design and launch its own web-based application. This new app, called PathTrack, now has the potential to markedly improve patient safety.

“I think there is a perception that we can always collect another specimen: ‘Well, it’s just urine or blood. If it gets lost, we’ll collect another one,’” says Amy Mapili, who serves as the initiative’s project manager. From a patient’s perspective, however, that second sample “can be a big inconvenience.” It might require a return-trip from far away, or an additional needle poke, which is particularly unpleasant for babies and young children. Then, too, some specimens are simply irreplaceable: a lymph node removed during surgery, for example, that pathologists will use to assess the severity of a patient’s cancer.

As a starting point, Mapili and her colleagues conducted focus groups with laboratory staff to document the complexity of the Department’s workflow. “We didn’t want to create a system where someone comes down from above and tells everyone what to do,” says Mapili. “The laboratory staff needed to have a voice.”

Mapili knew from the outset that specimens travel very different routes from collection to delivery. Consider a woman whose cervical cells are collected for a Pap smear at East Ann Arbor Health and Geriatrics Center, off of Plymouth Road. Those cells will travel by courier to University Hospital’s specimen processing center. The cells will then be sent down the hall to the Cytology Laboratory for evaluation. If irregularities are detected, that specimen would then be transferred to the Clinical Microbiology Laboratory, where the cervical cells would undergo testing for HPV. On the other hand, a patient at the Cancer Center will have blood drawn there. The blood sample will then travel via pneumatic tube to the specimen processing center, then onto the Hematology Laboratory. Following specimens from these types of disparate locations would require a flexible and robust tracking system.

During the focus groups, the team also identified “pain points” that currently make laboratory work unnecessarily difficult. For example, scanning: with the laboratory information system (LIS), you can scan a specimen’s barcode, but if the computer’s cursor is in the wrong place, the scans might not be entered into the system. “We wanted to build our tracking system so that when the scanner beeps, you know that it will have done something on the screen,” says John Hamilton, a department business analyst. “It takes away the frustration.”

Hamilton joined the project early last year, at the suggestion of Ulysses Balis, MD, director of the Division of Pathology Informatics. By that point, the PathTrack team had created an extremely detailed and accurate map of user requirements for the new system. Because these requirements were so disparate across the Department, and because they incorporated appliances such as barcode scanners and label printers, the team knew they needed a web-based application. The Department also needed a program that could interface well with LIS, as well as with the clinicians’ system, MiChart. At the time, such an application did not exist. Informatics leadership determined that the best approach was to build one in-house.

Brian Royer explains the architecture and mapping of the PathTrack application with the team.Hamilton was tapped to serve as the lead software architect for engineering the app, and quickly realized that the scope of the project was going to require a team of Informatics experts. “We really needed all the people in Informatics to do the piece they are experts in,” he explains. “The database people would create the database pieces. The web people would build the web pieces. I’d be the person who understands what the finished product should look like, how all the pieces should work together based on what we learned from the focus groups.”

“As a software engineer, the greatest challenge in building an application is creating something that actually answers the need,” says Balis. “Too many times, I’ve seen situations where the application is built ‘correctly,’ but when deployed in the field, it completely fails because the need wasn’t correctly or fully understood.” Thanks to the team’s many hours working with the frontline staff, PathTrack is positioned to work well once it’s rolled out later this year. And because it is a homegrown application, it can be monitored and adjusted if necessary.

“We felt that if done well, this could make people’s jobs easier and improve patient safety,” says Hamilton. He adds, “With the resources and talent we have here, we can take things to the next level in a way that an external vendor can’t.” Perhaps ultimately, other hospitals may even want to purchase the application. The problem of specimen tracking gets raised at professional conferences, says Hamilton. “The need is out there,” and the new PathTrack application can be scaled, making it a potential fit for any institution with specimens that move between sites.

For now, though, everyone involved in the project is looking forward to its launch here at Michigan Medicine, where patient- and family-centered care are at the forefront, even in a department like Pathology, where faculty and staff work mostly behind the scenes. Says Newton, “We may not be sitting in front of the patient discussing their care, but we are caretakers of their specimens, all along the way.”