College of American Pathologists
Printable Version

  Simplifying send-outs—what experts advise


cap today



February 2007
Feature Story

Karen Lusky

The best advice for hospitals looking to make the most of their relationship with reference labs isn’t so much caveat sender as buyer be prepared. Hospitals have to negotiate with reference labs and manage send-out testing to meet their patients’, clinicians’, and fiscal needs. In that regard, they might benefit from the collective wisdom of those on different sides of the send-out fence.

The most serious mistake Wayne Mercer, PhD, senior operations consultant for Chi Solutions Inc., Ann Arbor, Mich., sees hospital labs make is failing to define expectations up front.

“Some want the reference lab to be a service provider at the lowest reasonable price compared to competitors, [while] others are looking for an outreach partner,” Dr. Mercer observes. The hospital’s needs may also change over time. For example, “The hospital may want to start outreach, and the reference lab it’s teamed up with may be a competitor in its service area,” a scenario that Dr. Mercer has encountered quite often.

“If a reference lab is a competitor, the hospital may not want to include it in a bid process” for send-out testing, he says.

In addition to selecting a reference lab that fits their outreach goals, hospitals have to negotiate a fair price for their send-out testing—a task that offers more leeway than some may realize. “Since the mid-1990s, the market for national reference and esoteric testing has been subjected to sustained and often intense price competition,” says Robert Michel, editor of The Dark Report, which has reported on the trend over the last decade.

Thus, to differentiate themselves, national reference lab companies, as sellers, will often offer “some combination of lower prices and/or value-added services as a way to make the total contract package worth more to the buyer—the hospital laboratory,” observes Michel. “So the question for hospitals is simple: During the RFP process for send-out testing, are you negotiating a final agreement that works to your lab’s maximum advantage?”

When might a hospital consider changing its primary reference lab based on pricing? If a hospital sends out RFPs to several large reference labs and finds it could save $50,000 a year by choosing one, it’s “time to sit up and take notice,” says Debbie Warren, director of business development at Chi Solutions and a former general manager with LabCorp. “The first thing the lab can do is go back to its primary reference lab, if they are satisfied with the overall service, and give the primary lab an opportunity to review the new pricing. But the hospital lab also has to look at the cost of changing its interface to accommodate a new reference lab. How much support is the reference lab offering to provide the hospital to make that happen?”

Warren says, “A hospital can usually negotiate a contract where it gets no price increases for two years and then gets five percent in year three and another five percent in year five.”

Stephen Travassos, director of the lab at Brockton (Mass.) Hospital, which uses Quest Diagnostics for its send-out testing, thinks it’s a good idea for a hospital to send out an RFP when its contract with a reference lab ends. “We’re all expected to do more with less,” he points out.

“When we do an RFP,” says Travassos, “we ask reference labs if we can contact and interview their clients” to ask them about the lab’s turnaround time—and how the lab responds to their problems.

To prevent reference labs from cherry picking their references, Travassos asks them for a list of all their clients in his area. In addition, Brockton Hospital belongs to a buying group, so he asks hospital members of that group for information about the reference labs they use.

Before Brockton decided to renew its contract with Quest, Travassos interviewed “different hospital labs that were using one of the labs we had sent an RFP to and got some negative and positive comments. But Quest wanted to keep us, so their pricing was pretty aggressive.”

Hospital labs can use Medicare as the ceiling for pricing “because once you price above Medicare,” the hospital is losing money, says Larry Small, MT (ASCP), MS, president and principal consultant, L Small Associates, Tierra Verde, Fla. “You use the Part B rates as the yardstick even though it’s Part A (DRG-based) reimbursement,” he adds.

Thus, “when the hospital negotiates, it can say [to the reference lab], “We want everything priced at 50 to 75 percent of Medicare with our ceiling being the Medicare rate…give us your offer,” suggests Small. “The hospital can make an exception that’s a known exception”—for example, a genetic test or one not covered by the reimbursement.

Can hospitals make money off their send-out testing? Chi’s Warren believes the potential exists. She observes that most large reference labs give hospitals a price break on their most commonly ordered tests. But hospitals are losing revenue on high-priced tests, such as HIV resistance testing and cytogenetics, she says.

Warren thus advises hospital labs to look at any test over $100 and renegotiate those with the reference lab. She has seen hospitals cut their negotiated prices $50 or more on such tests. “Even though the hospital may be ordering only five such tests a month, when you lose $50 or more per test after reimbursement, it adds up,” she says. “The reference labs are trying to market these tests as nondiscountable or book price, but hospitals can still negotiate pricing, especially based on average reimbursement.”

Warren notes that some reference labs will offer to take the risk for high-priced tests, but that deal may not be in the hospital’s best interests. “It’s more likely the reference lab is offering to do so because there’s money to be made on that test. Be sure to ask for the price you need based on reimbursement and see if it makes sense for the hospital to bill for that test.”

Of course, a hospital lab ultimately gets what it pays for. “The eternal truth,” says The Dark Report’s Michel, “is that the cheapest price is not always the lowest price. If your business doesn’t provide [the reference lab] adequate operating profit margins, then [the] reference laboratory will not be able to afford the expenses of scheduling courier services as frequently as one might like,” he says. The reference lab “may not be able to put as much money into writing interfaces or providing additional service support.”

Hospitals have to decide which tests they should send out versus do in-house—the proverbial and constantly evolving “make versus buy” decision that can make or break the hospital’s bottom line and service quality.

Sometimes hospitals have to lose money by bringing a low-volume test in-house to meet demands for turnaround time, says Ira Allen, MD, medical director of the clinical laboratory, Valley Medical Center, Renton, Wash. In such cases, “you bite the bullet and try to get as good a price as you can on reagents, and set up the testing to keep costs as low as possible. But you have to stay on top of the financials.”

Brockton Hospital brought fetal fibronectin testing in-house, even though doing so was expensive. But the hospital lab had no choice, says Travassos, given that the test is a standard of care for obstetrics and the doctors wanted a faster turnaround time. By bringing the test in-house, the hospital lab reduced the TAT from four hours to one hour.

Chi’s Warren suggests that “if doctors are insisting on a same-day TAT for a test that is more costly to perform in the hospital lab than at the reference lab,” work with them to determine if they could possibly receive the test results the next day without compromising patient care.

Hospitals and reference labs should partner on their test offerings, says Ronald Blum, PhD, vice president of marketing, Specialty Laboratories, Valencia, Calif. For example, “The hospital may be able to perform an antibody test that can detect HCV, but the physician may want to followup with related tests, often molecular in nature: HCV RNA detection, quantitation of viral load, and genotype.” Specialty just introduced the HCV liver fibrosis GenotypR test, which identifies the patient’s risk for progressing to cirrhosis, he notes.

To determine which tests to bring in-house, hospital labs should look at their test utilization to identify the tests—and test volumes—they send out each month. If the hospital has the test volume, instrumentation, and capacity, including staffing and expertise, it may make sense to bring a test in-house, says Dr. Blum. He cautions that sometimes “it can appear to make sense to bring in a test but the test is actually very seasonal,” such as influenza testing in the winter.

Most of the large reference labs offer consultation to help their clients decide what tests to send out versus do in-house. For example, ARUP’s Analyzing Test Ordering Patterns recently helped one of the lab’s larger clients identify an opportunity to switch drug screen testing from whole blood to urine, reports Ronald Weiss, MD, MBA, president and chief operating officer, ARUP Laboratories, Salt Lake City. That discovery paved the way for the client to do the test in-house as part of its urine drug screening testing. By making that change, the client ended up saving approximately $50,000 per year in send-out costs, Dr. Weiss says. And the hospital moved the testing “closer to their patients, thereby improving overall TAT and also reducing the opportunities for preanalytical problems.”

To help hospitals decide whether to let US Labs, Irvine, Calif., handle their immunohistochemistry tests, Dennis O’Malley, MD, medical director for the anatomic pathology lab, suggests they identify the 10 least-often performed immunostains they do in-house. Then they should decide if they can use up the antibody before it expires. “The answer is usually no—and that’s money down the drain,” he says. US Labs then offers to do those rare esoteric tests for the hospital, which “saves time, reagents, employees—and we do it all the time,” says Dr. O’Malley.

US Labs also does virtual image analysis for prognostic markers and virtual flow cytometry. “The methods allow us to do the technical components of those tests, but the pathologists interpret them,” says Dr. O’Malley. “We put the image on our Web site. The pathologist logs on to a secure link, interprets and manipulates the information, then integrates that into their own lab system,” he says. “They bill the professional component themselves.”

Pathologists Associated, the lab for Ball Memorial Hospital, Muncie, Ind., chose to use US Labs for flow cytometry because it was losing “a lot of money” on doing flow in-house, says Janet Roepke, MD, PhD. “But outsourcing flow cytometry completely was a problem because you want to correlate it with morphology and clinical history,” says Dr. Roepke, who works for Pathologists Associated.

Not only that, but given that Ball Memorial is a medium-sized hospital with a fair volume of flow cytometry, with a cancer center serving much of east central Indiana and a pathology residency teaching program, it was important to interpret the flow cytometry in-house, adds Dr. Roepke.

“US Labs has revolutionized the way we do flow,” she says. “They provided us with some markers that we had never been able to do here because they were more difficult to perform, such as TdT (a nuclear marker) and cytoplasmic markers such as cytoplasmic CD3 and cytoplasmic kappa and lambda.”

US Labs has also improved Ball Memorial’s TAT for the flow cytometry, which is hard to believe given the distance between Indiana and the California-based lab. But, as Dr. Roepke explains, “We only had one tech on one shift fully dedicated to flow cytometry,” while US Labs works 24/7.

There’s another issue hospitals have to address: information technology. “The IT portion is a huge deal,” says Chi’s Warren.

Connectivity is a big issue, she adds. The hospital has to have a bidirectional interface with the reference lab, says Warren. And before signing a contract, the hospital should determine how much offline entry will be required. “There are some interfaces where there are multiple tests that cannot be ordered or results received seamlessly through the interface. This can require an FTE to manage offline maintenance.”

“So just because the reference lab says it can interface, dig a little deeper and look at what types of tests the hospital will be ordering and what percent will have to be supported offline,” Warren advises. And keep in mind that “what a large academic reference lab can do can be different from what a national commercial lab can do,” she adds, which is “also dependent on the hospital’s system.”

Warren knows of one hospital lab that switched from a national commercial lab to an academic facility lab “and the maintenance time the lab was spending on its interface dropped by 80 percent.” By maintenance, she’s referring to the time the lab spends doing offline entry and having to reformat results coming from its reference lab.

Chi’s Dr. Mercer notes that a hospital often signs “a boilerplate contract [with reference labs] whereas it should be specific about what it expects in terms of interface support.” Also, in renewing contracts with a lab information system vendor, the hospital should get more specific about the level of support it expects, he says. “For example, if the hospital puts out an RFP, it could require any bidder to provide support to do database conversion.”

One scenario that hospital labs should try to avoid is having only one interface to a primary lab, says Harold Kisner, PhD, president of NorthEast Clinical & Anatomic Pathology Solutions LLC (DBA NECAPS), North Hampton, NH. That hospital-reference lab interface becomes “an umbilical cord that creates a moat around the hospital,” he cautions. And “for the hospital to change reference labs, it has to sever, then rebuild the umbilical cord [which is] not an insignificant challenge.” So once the hospital is “inside the moat,” it may start accepting service glitches more readily, because it recognizes the “friction cost” of changing labs, says Dr. Kisner.

“One way to handle that is to have two reference lab affiliations, each with an interface, so the hospital doesn’t put all of its eggs in one basket,” advises Dr. Kisner. That way, the hospital could easily decide which tests to send to which lab. “If the hospital is big enough to pull that off, it’s one way to keep its reference labs on their toes.” He notes that a “reference lab may try to strong-arm the hospital into having a sole relationship with it as a primary reference lab [by] making pricing discounts contingent on that being the case.” But “a lot of labs forget they are the buyer and need to negotiate what is in their best interest,” he says.

Putting all your eggs in one basket is one thing, but having a dozen or more baskets filled with small numbers of send-out tests can turn into a logistical challenge.

Chi’s Warren suggests hospitals assess how many reference labs they are using, noting that the number can creep up due to physicians’ requests to have certain labs perform tests. Warren worked with one hospital lab that was using 15 different reference labs.

“The hospital has to make a concerted effort to see how many of its reference labs it can eliminate,” says Warren. For example, the hospital can “check with the primary reference lab to see if it offers the same or a comparable test that a doctor wants sent elsewhere. Sometimes it’s just a matter of educating the physician that the main reference lab offers the test. Or if the doctor insists the test be done by another lab, the primary reference lab may be able to send that test to that lab if it is not an issue with cost or TAT.”

US Labs’ Dr. O’Malley says that if his lab can’t do the test a client wants, it finds out who does it “either within our larger system of associated labs or outside that system.” In one instance, a client called him about a “difficult diagnostic challenge in a case, and, as a result, we created a new diagnostic test for that particular situation involving a cancer diagnosis.”

Hospitals have to find ways to evaluate the quality of their reference labs’ performance—and their role in creating any glitches—which gets tricky when in the thick of daily operations.

Valley Medical Center monitors turnaround times for its sendouts, says Sharon Hu, MT (ASCP), MS, director of laboratory services. “We have an overdue pending log for our reference tests, which is an indirect way of monitoring TATs,” she explains. “Say the reference lab says the TAT for a test is 72 hours. In the computer, we set up a log that shows overdue testing, which allows staff to identify tests that are always falling outside the expected TAT.”

Hu says, “The No. 1 complaint about reference testing is that the hospital lab isn’t notified in a timely fashion that a test was canceled, due to, for example, insufficient specimen or the specimen being compromised in some way.” Frederick Meier, MD, lead author of a 1993 CAP Q-Probe, “Reference Test Service Quality Data Analysis and Critique,” agrees that’s the “No. 1 dissatisfier with reference labs.” But the most frequent problem documented in the Q-Probe was the wrong test being ordered, he says.

Dr. Meier says that test ordering is a problem on the hospital side when the physician orders a test “within a genre of tests and delegates the task of identifying the appropriate, specific test order to his/her nurse, resident, or ward clerk.” To address that problem, the hospital lab needs to invest in personnel who interact with reference labs to ensure they have people who know what’s going on, says Dr. Meier. For example, some hospitals have send-out specialists, a tactic that Brockton Hospital’s Travassos recommends.

For their part, reference labs can help by making clear “to everyone ordering the test what the test is they are trying to order,” says Dr. Meier. He suggests reference labs might offer Web-based information for physicians to read up on various tests.

ARUP has recently done that with its ARUP Consult, a Web-based public domain tool ( that helps clinicians determine how to use and interpret many of its esoteric tests. “The software database is updated every two months with additional entries,” says Dr. Weiss, “and can be downloaded to most PDAs and smart phones so the physician can carry this tool to the point of care.” ARUP eventually wants to populate the tool with its entire test mix, including routine lab tests.

The CAP Q-Probe identifies a checklist of communication events between send-out and reference labs. Forty-nine percent of the communication events involved postanalytical issues occurring on the reference lab’s watch. Those issues ran the gamut from turnaround delays to computer or printer problems to uninterpretable and/or erroneous reports to specimens being lost or destroyed to the wrong testing being performed.

“If people can agree that the problems we found in our study are the problems [they are encountering], they could track them,” says Dr. Meier.

Dr. Weiss says that preventing problems that occur with send-out testing “requires well-designed systems within and between laboratories. Both sides of this interaction are responsible to see that these systems work well—open communication is vital to a successful, long-term relationship.”

Karen Lusky is a writer in Brentwood, Tenn.