INNOVATIVE CLINICAL LABS are making significant changes to accommodate the shift from volume to values. In one change, they are collaborating with clinicians to use lab test data to improve patient outcomes.
In another, they are collecting the clinical data from these efforts to publish the outcomes in peer-reviewed medical journals to spread the word about their efforts.
The clinical laboratory at Northwell Health in Lake Success, N.Y., is such an innovator. At the Executive War College last month, Tarush Kothari, MD, MPH, Physician Informaticist at Northwell Health Laboratories, explained how collaboration between the lab and clinicians can improve the diagnosis and treatment of patients with acute kidney injury (AKI).
Now in its third year, this clinical collaboration has helped physicians, nurses, and other caregivers diagnose AKI earlier. In turn, patients are getting the right therapy sooner. This effort has improved patient outcomes and reduced healthcare costs.
At the center of these impressive outcomes is a lab test that is ubiquitous, easy to run, and inexpensive: serum creatinine.
In his presentation, Kothari explained the evidence-based criteria for diagnosis and staging of AKI and how laboratories are positioned to drive quality improvement efforts for patients outside the lab.
The findings from the three-year project show that the lab’s efforts resulted in a significantly higher rate of detection of AKI from the baseline year of 2014 through 2016. Clinical data collected during this project showed that using lab test data—including creatinine results—while following the Kidney Disease Improving Global Outcomes (KDIGO) guidelines allowed clinicians to identify more AKI cases than what the hospitals had previously identified using DRG data.
Value of Collaboration
“The result was earlier detection of AKI leading to better patient care, more accurate diagnoses and coding, and reduced costs,” said Kothari, who has submitted the data for publication in a peer-reviewed journal.
“Today, it’s not enough for clinical lab scientists to just sit in our silo and think that we’re doing a great job,” observed Kothari, an Assistant Professor in Pathology and Laboratory Medicine at the Hofstra Northwell School of Medicine. “What this project taught us is that we need to step out of our lab and collaborate with our peers who are doing quality improvement work in clinical settings.
“For this project, we set out to standardize early detection of AKI and reduce variability in diagnosis and management by embedding clinical decision support systems into patient care workflows,” he explained. “We were not just embedding clinical decision support in the laboratory information systems.
“As we built up the Clinical Lab 2.0 team with the goal of improving diagnosis and treatment of AKI, we asked ourselves this question,” added Kothari. “How could we improve clinical and financial outcomes while showing value to all stakeholders, meaning our patients, providers, health systems, and obviously payers?”
Kothari was in the first week of his job at Northwell Health Laboratories on Long Island when he got an email from the chief medical officer at 250-bed Forest Hills Hospital, in Queens, N.Y., one of 22 hospitals in the Northwell system. “The CMO was seeing three patients with radio-contrast-induced AKI every day,” commented Kothari.
“Three cases per day equals about 1,100 cases per year,” he calculated. “If you attribute two excess length-of-stay days to an episode of AKI, the math works out like this: 2,200 excess days per year at about $500 per day in variable costs, or about $1 million in additional cost annually just for this one hospital. And, in severe AKI cases—where the length of stay goes up by three to seven days—means that total costs of care rise by about $4,000 to $10,000 per patient encounter.
“That amount of excess spending on AKI patients represented potentially huge savings—not just at this hospital, but across our entire health system,” recalled Kothari. “That was the impetus that got me started.
“What we knew about AKI is that—when serum creatinine increases by even minute amounts in a short time—mortality rates rise and healthcare costs rise too,” he noted. “We also know that about 20% of medical and surgical patients suffer from AKI in general hospital settings.
“The incidence in critical care settings is actually much higher, about 20% to 30%,” continued Kothari. “Also, AKI encompasses a variety of disease states.
How Northwell’s Lab Team Helped Improve Care of Patient’s with Acute Kidney Injury
TO IMPROVE THE DIAGNOSIS AND TREATMENT OF PATIENTS WITH ACUTE KIDNEY INJURY (AKI) at Northwell Health, the clinical lab team worked with physicians and nurses to agree on care protocols and put new alerts into the Cerner LIS and Epic EHR to support closer interaction. Creatinine results for individual patients were used to indicate the need for clinical teams to provide appropriate care. The project produced faster diagnosis of early stage AKI and improved outcomes.
Who Makes The Diagnosis?
“It is important to note that, although this is a condition nephrologists treat, the doctors who make most of the diagnoses of AKI include general internists, surgeons, and ER physicians,” he said.
“To pick up this diagnosis can be challenging because AKI is usually secondary to a primary diagnosis, such as sepsis, pneumonia, or trauma,” commented Kothari. “It’s easy for doctors to forget about this diagnosis. Yet, AKI is a broad problem in all hospital settings for all subspecialties.
“The literature shows that AKI represents about 5% of total hospital costs and annual healthcare costs attributable to hospital-acquired AKI exceed about $10 billion in the United States,” he said. “When AKI progresses in severity from Stage 1 to Stage 2 to Stage 3, the mortality, length of stay, and costs all worsen,” he added.
“The diagnosis of AKI relies on the incremental rise in inpatient serum creatinine compared with a minimum baseline value within a fixed time period,” noted Kothari. “According to the KDIGO standard, the diagnosis of AKI requires one of these two criteria: 0.3 mg/dl rise above baseline within 48 hours or 1.5 to 1.9 times baseline within seven days.
A Challenging Diagnosis
“The staging of AKI also depends on the relative rise of creatinine,” he said. “So if the creatinine rises by 0.3 mg/dl from baseline or by 1.5 to 1.9 times baseline, that is Stage 1. Greater than 2.0 to 2.9 times is Stage 2, and greater than 3.0 times baseline is Stage 3.
“Making this diagnosis is challenging and is often missed because of two key factors,” Kothari said. “One is the definition of baseline. In cases where the patient’s baseline value was not known, we decided to use the minimum inpatient value as the baseline, as per KDIGO criteria.
“The second important factor for diagnosis is the time frame,” he stated. “The rise has to happen by a certain amount in a set duration. So a rise of 0.3 mg/dl can happen only within 48 hours. A rise of greater than 1.5 times baseline should happen within seven days.
One problem the lab faced when seeking to identify patients with AKI was how to reconcile the creatinine data collected at the point of care with the hospital’s DRG data. “When we compared the lab data showing the incidence of AKI with the data the hospital had for AKI incidence based on DRGs, we observed a significant increase in the documented rate of AKI from about 5% in 2014 to more than 12% in 2016,” Kothari explained.
“When we looked at the absolute numbers of cases from our baseline of 2014, we captured 8,000 more episodes of AKI in 2015 and about 10,000 more episodes of AKI in 2016,” he added.
“Then, even rough math will show that if the hospital gets paid about several hundred more dollars for a secondary diagnosis of AKI, then our clinical collaboration could increase reimbursement significantly in 2015 and in 2016.
“One conclusion we can draw from this project is there was a significant difference between the lab-detected AKI episodes and coded DRG AKI episodes,” emphasized Kothari. “This gap narrowed in 2015 and in 2016 because of the attention we gave to this condition. Plus, we captured disease severity more accurately.
“One factor that was essential to our success was collaborating with our clinical documentation team and with our physician colleagues,” he said.
“At this stage, our work is not done,” he added. “We are now linking our lab data sets to other data sources—such as pharmacy and hospital cost data—to refine our intervention so that we can measure reductions in the cost of care for patients with AKI objectively. There is much more value that can be realized by doing so.”
Contact Tarush Kothari, MD, MPH, at 516-719-1528 or firstname.lastname@example.org.
Northwell’s Laboratory Had Role in Developing Decision Support to Help Physicians with AKI
IN ANY HOSPITAL, GETTING PHYSICIANS to understand the value of lab data at the point of care can be challenging. Identifying patients with acute kidney injury (AKI) is one example because guidelines show that the diagnosis depends on the increase in serum creatinine over a certain time.
“This is where clinical decision support comes in because this is a busy hospital,” stated Tarush Kothari, MD, MPH, Physician Informaticist at Northwell Health Laboratories. “Busy clinicians do not have time to apply the Kidney Disease Improving Global Outcomes (KDIGO) criteria consistently and prospectively in real time. Plus, there is a lack of effective clinical decision support tools for AKI within our EHR. Even if you build alerts in the EHR, they may not be integrated appropriately into the clinical workflow.”
The solution Northwell implemented was to apply the KDIGO guidelines into the laboratory information system to flag patients who meet the AKI criteria. “The goal was not to miss a single AKI patient,” explained Kothari.
“We let the physicians decide how they wanted to act on the lab AKI alert because ultimately this is a clinico-pathologic diagnosis,” he explained.
“When we implemented this alert at Forest HillsHospital,ourexpectationwasthatwewould see about 10 to 15 alerts per day,” Kothari said. “But we actually saw about 40 AKI alerts per day in a 250-bed hospital. This corresponded to roughly 20 patients, or a 10% to 12% incidence rate in a busy community hospital.
“Next, before we rolled out this alert across seven other hospitals in the Northwell Health system, we validated the algorithm,” he continued. “Then we educated physicians and nurses about how it worked and why it’s important to identify AKI early.
“These alerts are meaningless unless we supplement them with physician education and awareness,” noted Kothari. “That is why our CMO carried out a major awareness and physician-education campaign to ensure that all the key physician champions and everyone on staff were educated about implementation of this alert.
”In essence, this was a multifactorial informatics intervention guided by lab data,” Kothari concluded.