One of the motivations to bring mass spectrometry (MS) assays in-house is to reduce send out costs. Examples include vitamin D testing, urine drug analysis, and tests for other high-volume analytes. The expenses associated with setting up an MS system and assays can be offset easily by the immediate savings recognized from bringing this high-volume workload in-house. These are relatively straightforward business cases with readily quantifiable economic benefits attributed directly to labs, and they are low-hanging fruit when it comes to integrating new clinical applications of MS.
But once these straightforward cases have been implemented, labs often face challenges expanding MS applications, despite the technology’s capability to uniquely address crucial clinical needs. Laboratory directors quite understandably can find it frustrating when there is a clear clinical case for bringing a test in-house using MS, but the business case is not compelling.
Most often the problem boils down to limited financial return on a cost per reportable basis for low-volume testing. While there is truth to this argument when narrowly focusing on laboratory costs in isolation, many times big picture considerations tell a different story: If an analysis does not consider institutional-level costs, significant opportunities may be missed.
In this article, we share two examples that demonstrate how clinical laboratories led efforts to establish MS assays that were not supported by traditional department-centric financial models. The first example is from the University of Rochester Medical Center for a busulfan assay, and the second is from the Ohio State University Wexner Medical Center for neonatal drug testing. Both cases demonstrate the value of breaking outside the walls of clinical labs to bring multiple stakeholders into the discussion.
Many benefits accrue when clinical and administrative partners fully appreciate the value that MS platforms deliver to an overall health system—not just the laboratory. We hope, through these cases, to shed some light on less traditional pathways for justifying MS testing, with the goal of seeing it expanded further to the benefit of patient care.
Busulfan Testing to Improve Management of Stem Cell Transplant Patients
Busulfan is a chemotherapeutic agent used as a component in a myeloablative preconditioning regime before hematological stem cell transplantation. Therapeutic drug monitoring (TDM) and timely plasma concentration reporting is critical for determining proper dosage. Busulfan has a narrow therapeutic index, and its pharmacokinetic (PK) profile, liver metabolism, and clearance differ greatly from patient to patient, so the same dosage regimen could lead to very different clinical outcomes in different patients. Toxic drug levels in blood are associated with harmful side effects and even death, whereas low exposures put patients at risk for ineffective therapy potentially causing graft rejection or disease relapse.
For several reasons, busulfan testing had been a send-out at the University of Rochester Medical Center. First and foremost, low test volumes—less than 30 patients per year—didn’t support a good business case to implement the test in-house, in part because the lab would need additional capital equipment.
The prospect of in-house busulfan testing posed other challenges as well. Busulfan in patient samples is not very stable, presenting technical difficulties in developing a reliable assay. In addition, no traditional proficiency testing (PT) for busulfan is available from organizations like the College of American Pathologists. That meant that to fulfill regulatory requirements the lab would have to establish unique assay-based PT programs.
The equation changed, however, when the pharmacy and bone marrow transplant group approached the lab a couple of years ago to discuss the feasibility of moving this test in-house. Leaders from the lab and these two hospital units formed a brainstorming group to evaluate the question from a more holistic point-of-view. From the lab perspective alone, all of the challenges mentioned above did not support the assay as an in-house test on a cost per reportable basis. However, with further input from multiple groups, the lab discovered that in-house testing for busulfan would help not only the pharmacy and the bone marrow transplant groups but also offer significant benefits to the overall hospital.
The first major benefit was that in-house testing enabled clinicians to administer the drug in an outpatient rather than an inpatient setting. Patients received four doses of busulfan as the preconditioning regimen. The first dose provided samples for PK estimates, which involve calculating the area under the receiver operating characteristic curve to guide subsequent dose adjustments. Four samples were taken after administration of the first dose. When the test was being sent-out, in order to have time to adjust the third and fourth dose, our lab had to send out the PK samples overnight to the reference lab. To meet the reference lab sample processing deadlines, the last PK sample had to be drawn before 11 a.m.—9 hours after the start of the infusion. Consequently, the busulfan dose had to be started around 2 a.m. Due to this time constraint, clinicians had to administer busulfan on an inpatient basis since the outpatient infusion center is not open 24 hours.
Moving the assay in-house eliminated these timing and transportation challenges. The dosing regimen and lab sampling followed the same schedule, but doses could be started in the early morning in the outpatient setting during normal business hours while still allowing the pharmacy team to adjust doses three and four in a timely manner.
Changing from an inpatient to an outpatient setting provided significant value to both patients and the hospital. Importantly, it saved patients 4-5 days of being hospitalized. This not only reduced patients’ costs but also enabled the hospital to be more efficient in utilizing its inpatient space and in taking on other critical patients. Furthermore, research has shown that patients have better clinical outcomes when treated in outpatient settings. Outpatient care decreases the anxiety of staying in a hospital, minimizes patient inconveniences, and reduces the prospect of nosocomial infections.
In offering busulfan in an outpatient setting the hospital also was able to bill the drug under the 340b schedule under Medicare rather than an inpatient schedule. Though a seemingly small change, this resulted in significant drug cost savings for both patients and the hospital.
After collaborating with all related stakeholders and evaluating the benefits holistically as a healthcare team, the lab was able to obtain the instruments and other capital equipment needed—as well as proper staffing—to support busulfan in-house testing. More importantly, this collaborative effort provided both economic and quality-of-care improvements for the hospital and for patients. The hospital would not have realized these gains if the lab had evaluated busulfan testing solely from its own financial perspective.
Meeting Clinical Needs for Neonatal Drug Testing
Neonatal drug testing has been an overlooked service for clinical laboratories. But a rise in substance abuse during pregnancy as part of the overall opioid misuse crisis makes close attention to this area more urgent. Clinical laboratories now play a key role in identifying newborns who may have been exposed to drugs in utero. The following case from the clinical lab at the Ohio State University illustrates how using MS helped reshape the workflow to better serve this patient population.
Clinical laboratories find neonatal drug testing challenging for a variety of reasons. For example, the specimens available from newborns are generally different than those used in adult populations. This presents preanalytic challenges in addition to the usual analytic considerations: Neonatal urine, meconium, or umbilical cord have very specific collection requirements to ensure optimal results. These specimens often require additional validation of the collection by the laboratory as well as extraction of the target compounds from the specimen matrix.
Interpretation of neonatal drug test results is also more complex for many healthcare providers, including not only physicians but also nursing, labor and delivery, pediatrics, and social work staff. Often these results are integrated with two other sets of data, namely maternal urine drug screen results and historical accounts of previous maternal drug use. Integrating these discrete data streams into a single data set that represents cumulative exposure during pregnancy is not straightforward. Limitations related to the specimens used, extraction efficiency, and assay sensitivity can also contribute to discrepancies.
At Ohio State University, the labor and delivery nurse manager recognized that the hospital’s workflow was ineffective for identifying and testing newborns at risk for drug exposure during pregnancy. Calls for a better workflow came from multiple providers who spanned a variety of care teams, from nurses in labor and delivery to providers in the neonatal intensive care unit and social services. These staff members were concerned about poor handoff of information related to knowing when testing should be ordered, as well as slow test results turnaround time by the lab, delays in discharging patients waiting for results, and results that were inconsistent or inadequate to detect expected drugs.
To deal with these problems, the hospital established an interdisciplinary team with representatives from pathology, labor and delivery, obstetrics and gynecology, pediatrics, and social services to critically evaluate the hospital’s process and to improve the continuity of care. As part of this process, the team evaluated switching from meconium to umbilical cord tissue as the default specimen for evaluating drug exposures. This specimen type offered several advantages for nursing, such as being able to collect specimens at birth and being easier to collect than meconium. The group simultaneously identified both of these issues as problems in the current workflow that delayed results and possible interventions.
To tackle this dilemma, the clinical laboratory first set out to validate a new specimen type staff had no prior experience dealing with. The conventional approach to sample preparation for MS typically involves just a single extraction step. However, the laboratory quickly learned that the consistency of umbilical cord tissue required additional considerations to go from specimen to MS instrument-ready samples.
The lab found it helpful to separate into discrete components all the processes that typically take place in a single step. This included thinking about tissue disruption, drug isolation, and sample cleanup as three individual steps. The first step of mechanical disruption of the tissue ultimately required the purchase of an unbudgeted piece of equipment. Collectively, the team petitioned for funding directly to executive leadership and were successful in moving the project forward.
While the laboratory continued research and development to establish this assay, the remaining clinical teams worked on redesigning the workflow around this test. The goal was to develop a total testing process that began with properly identifying patients who need testing and ended with acknowledging the umbilical cord toxicology results—all within 24 hours.
The team ran into several roadblocks at this stage: suboptimal communication between care providers, delays in ordering and missed collections, extended laboratory turnaround times, and patient discharges prior to results being available. The team dealt with these issues by implementing new processes or streamlining existing ones. A focus on better communication between obstetrics and gynecology and labor and delivery decreased missed testing for at-risk babies, while collecting samples at delivery helped avoid missed collections. It was also important to clarify which physicians should place the testing orders. This reduced the number of patients who should have been tested but were missed.
With improvements underway, it was then up to the laboratory to reduce turnaround times from 20-24 hours to achieve the target goal. Months of development and a method comparison with meconium finally achieved an assay that met current performance expectations but with a much-reduced turnaround time of approximately 9 hours. After implementing this assay, the lab continued to closely monitor test performance to reveal areas for improvement.
Notably, one of the most engaged stakeholders using neonatal drug test results is social services, which must follow up on positive results. This group has the challenging job of integrating multiple streams of data, including mothers’ disclosed drug use, prescribed medications, maternal urine drug testing results, and the umbilical cord toxicology results. Discrepancies between maternal and infant results or the absence of expected results make decisions related to social interventions very difficult.
Detecting buprenorphine in umbilical cord tissues remains challenging in some specimens. The laboratory and social services rely on close communication between staff in both departments to troubleshoot unexpected outcomes. This project continues to bring together individuals from multiple departments to work toward a common goal of improving patient care.
Crucially, the conventional quality indicator of turnaround time was insufficient in this case to capture the complete impact of the redesigned workflow and test development. Developing this new clinical MS assay impacted multiple departments and care teams, so the lab put careful consideration into measuring success.
In addition to monitoring collect-to-receive and receive-to-result times for the umbilical cord toxicology, the team also monitored birth-to-order and birth-to-collect times. These quality indicators better evaluated preanalytical delays and coordination between care teams. The team also collected clinical outcome data that compared opiate-positive results with a clinical diagnosis of neonatal abstinence syndrome. Together, the preanalytic, analytic, and postanalytic metrics provide a comprehensive view of the redesigned clinical workflow that integrated the performance of the new MS test.
Continuing to Add Value
MS has gained recognition as a technology that can uniquely support critical patient needs. Yet challenges arise when clinical laboratories attempt to expand beyond the first line, obvious tests to implement on these instruments.
The two successful cases in this article examine how to approach implementing the next tier of MS testing, even when at first glance, it wasn’t viable. Both demonstrate that success was only possible by working with multiple clinical teams in healthcare systems. We hope these cases provide insights into the justification process when labs are considering creative applications of MS to serve essential patient needs.
We believe that with a collaborative, interdepartmental approach—one that evaluates the costs and benefits beyond the walls of clinical laboratories, we will see more applications of MS that add value to healthcare systems.
Yan Victoria Zhang, PhD, DABCC, FADLM, is associate professor, director of the regional toxicology lab, and vice chair of clinical enterprise strategy at the University of Rochester Medical Center in Rochester, NY. EMAIL: [email protected]
Steven Cotten, PhD, DABCC, FADLM, is assistant professor-clinical in the department of pathology and laboratory medicine at the University of North Carolina at Chapel Hill in Chapel Hill, North Carolina. EMAIL: [email protected]