Inside the Lab Industry: Hospital Systems Use Variety of Methods to Control Test Utilization

Although many students are taught to the test, it is the experience of Johns Hopkins University physician and professor Larry Feldman, M.D., that medical residents learn from the test instead. And while residents gain knowledge from ordering large volumes of assays, Johns Hopkins’s bottom line loses traction in the process. “Residents send tests all of the time. And it was driving people crazy,” Feldman said during a workshop at G2 Intelligence’s annual Lab Institute on how providers reduce utilization while optimizing costs and outcomes. The institute was held in Arlington, Va., Oct. 16-18. Examples of lab utilization management programs were cited from Johns Hopkins and two other renown providers: the Mayo Clinic and the University of Michigan Health System. Feldman said the goal of Johns Hopkins was to have its residents be “Oslerian” in the way they diagnosed a patient, in reference to medical scholar William Osler’s teachings. But the organization also wanted them to be “parsimonious” in test ordering and usage. That’s a tall order for the health care industry. Feldman noted that clinicians are always itching to use the latest technologies to diagnose and treat patients, leading to unwarranted usage, uncoordinated care, and other examples of waste. And most doctors have “no clue” how much their overutilization costs, he observed. Feldman cited one study from Australian researchers who concluded that nearly 68 percent of all lab tests ordered in hospital settings—about two per patient per day—did not contribute to patient care. “Does this happen at Johns Hopkins? Absolutely!” he exclaimed. Baked-in institutional policies—such as washboards reminding clinicians to run CMP, CBC, and ionized calcium tests as a matter of routine—contribute to this issue. Even the system’s electronic medical records (EMRs) can help pile on, with a “repeat” button readily available to order duplicative tests on a daily basis, or glitches that make it difficult to retrieve previous results, prompting retesting. A culture of test permissiveness also leads to what Jeffrey Warren, M.D., director of the division of clinical pathology at the UM Health system, termed “bizarre” requests. Among them was an order for a complete mitochondrial DNA sequencing, apparently in the search for a new form of mitochondrial myopathy. Jim Hernandez, M.D., chair of Mayo Clinic Arizona’s division of laboratory medicine, called this approach to testing “carpet bombing” and dryly illustrated it with a B-52 dropping dozens of pieces of ordnance. At Mayo, Hernandez said the blanket fashion in which tests are being performed tended to actually harm the health of its patients, as unnecessary bleed time tests could lead to iatrogenic anemia. “We are literally phlebotomizing patients,” he said. A New Culture How did Johns Hopkins instill a culture of assay parsimony in its residents—and how did Mayo and Michigan similarly cut test volumes? By giving some or all of the clinicians a clue on costs. Johns Hopkins developed a pilot program wherein the Medicare allowable fee for performing a specific test popped up on the EMR system at the point of order, giving clinicians an idea of how much they actually cost. Mayo began educating its medical staff on test overuse. And UM Health created a testing “formulary” that simply barred clinicians from ordering tests not normally considered medically necessary unless there was a valid reason for doing so. For the UM formulary, which was created in 2008, the use of tests must be evidence-based, with a committee poring over efficacy and utilization data. A similar evidence-based approach was adopted by Mayo, although it does not have a formulary per se. Although Warren noted that most lab tests normally performed have been kept in the formulary, some assays have been dropped, including 57 send-out tests and 11 in-house tests. A myelin-based protein test that cost $61 to perform was dropped after it was determined no more than 60 such assays were being performed systemwide, the majority for outpatient neurology patients. Another, a proprietary celiac disease panel, was dropped altogether. Multiple myeloma fluorescence in situ hybridization tests are limited to one per patient. Circulating tumor cell tests may only be ordered by oncologists. At Mayo, Hernandez estimated that about 30 percent of the tests it performs could be pared back without impacting patient care. He noted that the system was performing 112.3 lab tests per patient discharge during the first quarter of 2013. That’s down from 126 per discharge last year but is similar to the 2011 rate of 112.7. He noted that Massachusetts General Hospital was able to cut tests per discharge from 81 in 2002 to 60 in 2007, meaning Mayo has the equivalent of “a high golf score.” A Mayo team comprised of pathologists, scientists, nurses, finance experts, and administrators examines literature on best practices and outcomes to determine which tests may be unnecessary. Heavy users of specific tests receive “report cards” drawing attention to their ordering practices. Overutilization “alerts” are also inserted into Mayo’s EMR system to remind clinicians of the last time that specific test was ordered. Among the system’s two most overutilized tests: NT-proBNP, which determines some probability for heart attacks or other cardiac events, and magnesium testing. Literature suggested that NT-proBNP had efficacy on patients with congestive heart failure but not on patients who had no cardiac symptoms at all. Mayo’s EMR system is currently tracking repeat orders of the test. And less than 5 percent of magnesium testing orders demonstrated any clinical value at all, even though inpatients were undergoing more than five such assays during an average stay. One patient was tested for magnesium levels 228 times during a single hospital stay. At Johns Hopkins, Feldman noted that the national “Choosing Wisely” initiative to reduce unnecessary testing and procedures has been helpful in communicating to residents the need to curtail unnecessary testing. Altogether, 35 of Johns Hopkins’s most frequently ordered tests and 35 of the most expensive tests (ordered at least 50 times in the past year) had their costs pop up when ordered up by a clinician. As a result of putting those prices in front of the doctors, Johns Hopkins reduced tests per patient day from about 3.7 to 3.4 among the group exposed to costs (the testing rate went up in the control group, which did not see the testing costs). It helped save $433,000 during the first six months of the implementation of the program, according to Feldman. Communication Is Key Communication with clinicians is key in reducing test utilization. UM always notifies medical staff in advance if a test is being evaluated. It also performs follow-up assessments six months after a decision has been made. An appeals process is also in place if there are objections to a decision, although it is rarely used. “We’ve done a good job of prospective communication,” Warren said. He added that the mission of the formulary committee is to improve quality and utilization, not cut costs, and that its tone is along those lines. “If there’s debate, we tend to err on the side of liberalism . . . in making changes,” he said. Although progress in these campaigns has been incremental, that it is occurring at more than one health care system was encouraging to Feldman. “If it’s being picked up by Hopkins—which has a tendency not to change—it is being picked up in other places as well,” he said. Takeaway: Hospitals and hospital systems are slowly but steadily focusing on the overutilization of laboratory tests.