CEO SUMMARY: Probably no state has seen a faster transition to ACOs, medical homes, and other types of integrated clinical care organizations than Arizona. Recognizing that this change created a new opportunity to add more value with clinical lab testing services, Sonora Quest Laboratories (SQL) developed an enterprise-wide master patient index. This gives SQL the ability to build and maintain a complete longitudinal record of an individual patient’s laboratory test data.
WITH THE AMERICAN HEALTHCARE SYSTEM BEGINNING its transition to integrated care, it is timely for pathologists and lab administrators to think about developing the new capabilities required to succeed in the era of accountable care organizations (ACOs) and patient-centered medical homes (PCMHs).
As hospitals, physicians, and other providers come together in a single organization, the emphasis will be on improving patient outcomes. Remuneration will evolve away from fee-for-service. Both trends portend profound changes in how clinical laboratories deliver lab testing services and get paid for their efforts.
One essential capability that every lab serving an ACO will need to develop is an enterprise-wide master patient index (EMPI). The EMPI underpins all the patient-centric services that tomorrow’s clinical laboratory must support to be successful at meeting the needs of ACOs, PCMHs, and other emerging models of integrated clinical care.
Only a handful of clinical laboratories has created an effective EMPI. That is because the current generation of laboratory information systems (LISs) used throughout the United States were designed to be physician-centric, not patient-centric. It is also because building an EMPI requires extensive capital and special information technology (IT) expertise—resources that many clinical lab organizations do not have available at this time.
One lab that has built an EMPI and now uses it daily in support of patient-centric services is Sonora Quest Laboratories/Laboratory Services of Arizona, Inc. (SQL), based in Tempe, a suburb of Phoenix.
SQL’s EMPI is the essential element that allows SQL to bring together all the data on a specific patient and then use that data to deliver high-value services to the patient, the patient’s physician, the ACO, any health information exchange (HIE) with which SQL has a working interface, and the health insurers that may be part of individual ACOs.
Certainly all pathologists and lab administrators recognize the value of a complete longitudinal record of an individual patient’s laboratory test data. “We saw this need as one way we could deliver a patient-centric service,” stated David N. Moore, Chief Information Officer at SQL. “Once we have the capability to associate different lab test services to a single patient and create that longitudinal record, then we differentiate ourselves in the lab marketplace and we give multiple ACOs a reason to select us as a provider.
“The problem of correctly matching patient data is straightforward, but an EMPI is required to solve it,” explained Moore. “An EMPI is the essential tool.
“A properly-designed EMPI allows a laboratory to have confidence that it has correctly identified a patient,” he noted. “Then the lab can assemble all the clinical and other data associated with that patient.”
Tracking A Single Patient
Every laboratory manager is familiar with the challenge of following a single patient across all sites where care may be provided. “For example, there are times when a patient in an ACO goes to a provider outside of the ACO’s network,” observed Moore. “The ACO may not have all of that patient’s lab test records from that out-of-network service on file.
“If true, it means the ACO does not have a complete medical record for that patient or for any other patient who gets care out of network,” he said. “This inability to collect data from out-of-network providers is a problem for patients and for ACOs.
“In fact, this is the problem ACOs—as integrated clinical care providers—are designed to solve,” noted Moore. “If ACOs cannot collect and store all the data from each patient’s past encounters with the healthcare system, then much time and money was invested in the ACOs’ systems for nothing.
“Further, who knows better than pathologists and lab managers that different information systems used by various ACOs and provider networks are not always compatible,” he emphasized. “At SQL, we saw this as an opportunity to step up with a patient-centric service that was of value to all the participants in an ACO.
Eliminating Fragmented Care
“We did two things,” continued Moore. “First, we built an EMPI. Second, we created a layer of informatics to work in concert with our EMPI. Among other things, these other IT layers helped automate the process of matching different spellings of a patient name to a specific individual.
“Our collaborator in this effort was Atlas Development,” he said. “Our goal was to minimize manual matching of patient names by the staff and use IT to automate those functions as much as possible.
“This capability is now integrated into our workflow,” he commented. “Each time a patient presents at one of our patient service centers to provide a specimen, our staff looks in the EMPI to match that patient.”
With this patient-centric capability in place, SQL was positioned to go a step further in building a complete longitudinal record for each patient in its EMPI. “Now we could solve the problem of incomplete and fragmented patient records that plagues ACOs and hospitals,” observed Moore.
LIS Interfaced to HIEs
“For example, SQL is interfaced to several HIEs,” he stated. “To augment the existing data we have on patients, we can poll HIEs in our region and combine information from the HIEs into to our existing data. This allows us to construct a longitudinal patient record instead of having fragmented data.
“Let’s say a patient in Phoenix who would normally go to his or her in-network doctor has an emergency one weekend while out of the area,” Moore said. “That patient goes to an urgent care center that is out of the network.
“In this instance, the patient’s health insurer in Phoenix would have no idea about that patient encounter,” he explained. “However, because our lab serves both the in-network physicians and the out-of-net- work urgent care centers, SQL will contain the longitudinal record on that patient.
“The point is that we have gone from having a provider-centric LIS and the associated information systems that were cutting edge in the 1980s and 1990s, and—by adding additional layers of functionality—we have made these systems more patient-centric.
“The LIS itself is still provider-centric and we have left that untouched,” he said. “What makes our entire IT system patient-centric are the new informatics systems we layered on top of the LIS.
“This sounds simple when you say it that way, but it’s a significant shift in focus and in output,” declared Moore. “In fact, we haven’t changed anything inside the LIS other than to feed it the additional patient-centric data that we want it to have.
“That feed into our LIS complements all the information we gather from accessioning specimens when they arrive at the lab,” he said. “A program within the new informatics solutions analyzes the acces- sions and the demographics associated with them. It then matches the incoming requisitions to other patient data that it has from past patient encounters.
“Now the question to ask is what level of confidence do we have with our system’s ability to correctly match the data we give it with the data it stores from past encounters?” noted Moore. “We studied this closely and determined that about 96% of all orders are matched correctly the first time. Then our health information management systems team validates the suggested matches and gets the rate of overall validation above 98%.”
Due to continuous improvement programs, the error rate is declining. “Each time we deal with the various errors the system encounters, we implement a fix. We believe we will be able to approach a Six Sigma level of accuracy in our automated processes for patient identification.
“From our analysis of incoming requisitions, we know that, on average, about 3% to 4% of requisitions contain incorrect patient information,” he explained. “We also know that about 90% of all accessions verify correctly when they arrive. By using the algorithms Atlas Development has developed for us, our automated informatics solution can identify and correct errors about 99% of the time. That’s impressive.
“Best of all, this additional layer of information systems doesn’t affect lab operations at all,” emphasized Moore. “In fact, we are installing an LIS from NeTLIMS that has the ability to implement specific quality control measures that we will select. Our existing system with the additional layers for the EMPI simply allows us to speed up the registration part of patient throughput.
Identifying Next Steps
“Looking ahead to the next few years, SQL plans to introduce a patient loyalty program,” he said. “A patient who signs up will get a card to swipe upon entering any of our patient service centers.
“The card would tell the system the name of the patient and perhaps would allow the patient to get expedited service, and other benefits,” Moore explained. “With SQL’s EMPI and the ability to be patient-centric, we want to use the patient loyalty program as a way to create strong relationships with our patients.”
Arizona’s Providers and Insurers Moving Fast With ACOs, HIEs, and Integrated Clinical Care
ARIZONA IS AHEAD OF MOST OF THE UNITED STATES in its efforts to develop integrated clinical care organizations. Not only are there multiple accountable care organizations (ACOs) currently in full operation in Arizona, but one health information exchange (HIE) in the state already has hospitals representing 47% of the state’s beds feeding data into the HIE.
This raises the stakes for Sonora Quest Laboratories (SQL) as one of the state’s largest providers of clinical laboratory testing. It can gain a competitive advantage—and deliver more value to providers—if it can follow a single patient from one doctor to another and across different care settings. Such a capability, anchored by an enterprise-wide master patient index (EMPI), allows it to maintain a complete longitudinal record of that patient’s laboratory test data.
Because the traditional LIS is mostly a physician-centric system, SQL has adopted a strategy of layering different informatics solutions on top of its LIS in order to provide patient-centric services to providers, ACOs, health insurers, and others. (See TDR, December 2, 2013.) This gives SQL the capability to establish records keyed to the patient’s name and identifiable information rather than to a physician’s name.
Practical Value For Labs
This capability has practical value in the daily interactions SQL has with providers. “Take the example of a call to our patient service team with a question about a test or about a test result,” noted David N. Moore, Chief Information Officer at SQL. “Once we establish the identity of that patient, we can see at a glance every interaction that patient has had with our lab and our health system.
“This real-time access to the patient’s full record enables us to support clinical care and compliance with medical protocols,” added Moore. “Maybe the patient calling us needs fol- low-up testing as a result of a test done earlier. Possibly the patient needs a screening test of some kind. Our service team can now use that phone call to schedule a visit.”
In this regard, SQL is out in front with the deployment of its patient-centric features. Moore explained why existing laboratory information systems cannot be patient-centric in the manner required to appropriately service patients in ACOs, medical homes and similar integrated care organizations.
“Today, in most larger labs and with older LISs that are not patient-centric, all the data is linked to the ordering physician’s name,” observed Moore. “The patient’s name was not even a variable that the LIS addressed.
“To run a test, the lab needed to know only the patient’s date of birth, gender, and the client physician’s name who was ordering the test,” he said. “If the patient’s name on the requisition was unreadable as written, the lab could still run the test and call the doctor’s office the next day.
“But what happened when a test was done incorrectly?” asked Moore. “If the lab didn’t have the patient’s name, staff would have no idea how to contact that patient to re-run a test. That would be a problem for the lab and for the health insurance company.
“How about instances where your lab’s patients in New York go away every winter to vacation in Florida or Arizona and have lab work done that doesn’t get into the system your lab has established?” he asked. “Even some of the big national labs have this problem because they cannot track a patient who gets a test in one state and then goes somewhere else for a vacation or for part of the year and has other lab testing done. There is no longitudinal record of all of that patient’s encounters.
“Here at Sonora Quest, we rectified this problem,” concluded Moore. “It required us to invest time and resources to layer in the informatics capabilities required to offer patient-centric services that deliver more value.”