Tracking Transplant Patients
Through advances in science and medicine, organ transplanting has increased in frequency—and developed into an ever-growing business. At
A dozen years later, the hospital is still working toward that goal. “There’s nothing magical about information systems,” says Bryan Barshick, the center’s decision support/transplant manager, who is in charge of the project. “It will give you the bedsheets but it won’t make the bed for you.”
As a May 2007 Baseline cover story on the transplant program at the
While John Hopkins does not yet provide a text-book example of how to implement systems to manage a complex transplant center, it does offer other hospitals a guide to get closer to that goal.
No Room for Error
At John Hopkins, a big challenge has been getting staff and clinicians to abandon their comfort zone with paper-based processes in favor of automated transplant management systems. It’s not an uncommon problem for organizations adopting new technology. The transition is doubly challenging for John Hopkins and other top hospitals with zero tolerance for error in patient care and transplant compatibility. Mistakes can cost patients their lives.
All patients are listed in a national database administered by the United Network for Organ Sharing (UNOS), which matches organs with recipients based, in part, on how long they have been waiting for a transplant. UNOS uses complex algorithms to calculate that waiting time. It collects about 120 types of data and treats each type of organ tissue differently. Its database doesn’t automatically connect with the transplant databases at all hospitals, requiring some transplant centers to fax or e-mail patient data to UNOS.
The systems used to track patients and transplants tripped up Kaiser Permanente and, according to California regulators, placed patients “at risk for…potentially life-threatening delays in care.” Kaiser lost track of patients’ paper records, patients’ complaints and the time they spent on UNOS’ waiting list.
Two years ago, Johns Hopkins replaced the transplant database it had been using, which it declines to name, with a more modern one from TeleResults, a database vendor based in
The hospital is building interfaces between the database and various in-house and third-party laboratories. Radiology will soon get an interface, and Barshick also will try exporting data electronically from TeleResults to UNOS.
IT systems were sparse
In the mid 1990s, when Barshick joined Johns Hopkins as a clinical nurse on the abdominal transplant floor, the options for transplant-specific IT systems were “pretty sparse,” he says. He thought Johns Hopkins’ system did a good job warehousing data and was useful for hospital administrators, but clinical information was still siloed among the various transplant teams. The thoracic team, for example, used a different tool altogether—Microsoft Access.
In 2000, Barshick went back to school and got a master’s degree in nursing informatics. He’d been working with liver donors and saw how information systems could improve the transplant process. Meanwhile, Ghassan Khabbaz, founder and president of TeleResults, was developing the transplant database, a project that took more than eight years.
Khabbaz, an engineer, got the idea for TeleResults from a conversation he had with a doctor while he was working on an unrelated hospital security project. The doctor was searching desperately for a system to keep track of a transplant patient, he says. But Khabbaz’s first effort to create a database failed because, he says, he was overly focused on the doctor’s way of doing things. The challenge, he discovered, was designing a database to serve data to all the different groups working with transplant patients—doctors, nurses, social workers, clinicians, nutritionists and administrators. Johns Hopkins has a dozen such groups.
When Barshick returned to Johns Hopkins in 2002, the hospital created his current position, which includes determining how the transplant center can best use IT systems. That same year the hospital researched TeleResults and other undisclosed IT systems. It committed to TeleResults in 2005.
Barshick’s first challenge was to get data mapped and cleaned. The hospital’s abdominal database had approximately 4,000 data fields, he says, and some of it was dirty—text had slipped into a date field, or a typo had changed a medical record number. After months of work, Barshick’s team and TeleResults preserved 95 percent of the hospital’s data.
Training staff to use TeleResults took about a year, Barshick says. During 2005, he spent entire days in his office with the door closed, flipping through screens of information to understand how TeleResults stored data and where it was stored, so processes could be developed for everyone who needed to work with the database. He met individually with members of each hospital team—social workers, pharmacists, lab technicians—to ask them what they needed from TeleResults to get their jobs done.
“We had to change the way we talked,” he says, because the terminology TeleResults used differed from the hospital’s system. “Active,” for example, meant the patient was alive and being cared for; “active” on the UNOS waiting list meant the patient was still seeking a transplant candidate.
Getting TeleResults to fill out forms was also an issue. One of the most important forms for transplant teams is a flow sheet, a step-by-step description of the treatment plan for each patient. Until the conversion, all the transplant teams used handwritten forms to map and communicate the care plan. TeleResults worked with the hospital to link memos and other unstructured data to electronic flow sheets, so the information is available with the click of a mouse.
Change: A Tough Sell
It’s not easy to sell TeleResults to IT departments, Barshick and Khabbaz say. The hospital has more data on patients now than it ever had, and IT folks tend to see the database as “one more system to maintain.” But they also say TeleResults is better than generic health care IT systems at tracking transplant patients.
Khabbaz sees a bright future for the database at Johns Hopkins. When kidneys become available, qualified patients must be located and notified immediately: Transplants at Johns Hopkins are becoming increasingly elaborate, and hospital mathematicians have developed algorithms to sort and match patients to organs very quickly. In 2003, the hospital performed a three-way transplant, where three patients swapped donors so they could each get a kidney compatible with their blood or tissue type. In 2005, there was a five-way transplant spread across six operating rooms that lasted 10 hours. Four patients traded donors and the fifth got a kidney from an altruistic donor.
When it comes to medical care, transplants are often just the beginning. Transplants are a continuous process of managing disease, Khabbaz says, so he will design TeleResults to enable his customers—which include 20 large hospitals—to keep adding data to the database. Johns Hopkins will eventually use TeleResults to track all its patients who have undergone transplants.