Controlling Medicare
  medical necessity write-offs

September 2005
Feature Story

Karen Lusky

In a perfect billing world, clinicians would always check to see whether the lab tests they order for Medicare patients fit the government’s medical necessity requirements to diagnose or treat specified health conditions.

Any time a test didn’t appear to be covered, clinicians would explain to patients why the service would be beneficial even though Medicare would be unlikely to pay for it in their particular situation. Patients would then be given the choice to forego the testing or sign an advance beneficiary notice agreeing to receive it and foot the bill if Medicare wouldn’t.

Clinicians would always attach the ABNs they collected to test requisitions so labs that perform the testing could bill the patients or their secondary insurers for it once Medicare denied the claims.

Dream on, labs say.

In reality, "it’s a rare day when labs receive ABNs from physicians ordering the tests because the physicians really have no [financial] incentive to collect them," says Paul Keoppel, MBA, MT(ASCP), compliance/billing administrator for laboratory services at Intermountain Health Care in Salt Lake City, a nonprofit integrated health system. "And if the physicians don’t collect an ABN, there’s a good chance that the lab will have to write off [the charge] for the test, which the patient gets for free."

In some labs, the incremental losses from medical necessity write-offs go largely unnoticed until they take an obvious toll on the bottom line.

The good news is that labs can quantify and rein in Medicare medical necessity write-offs by implementing Six Sigma, a data-driven measurement and process improvement approach designed to fix systems that fall below specifications, says Keoppel, who shared his own experience in July at the AACC annual conference in Orlando, Fla.

Intermountain Health Care, or IHC, implemented a Six Sigma project in 2002 with the aim of slashing its Medicare medical necessity write-offs for lab services by 75 percent in one year. Sounds like an aggressive goal, Keoppel admits, but not when you consider that IHC had posted $1.3 million in Medicare laboratory write-offs that year because no one had collected ABNs from beneficiaries. The red ink generated by write-offs had been inching up four percent each year since 1999, which was the first year the lab’s computerized billing system had edits to measure the medical necessity adjustments.

"Left unchecked, the lab write-offs would have hit $3.9 million by 2007," Keoppel told conference attendees. And no matter how noble a health care organization’s mission might be, he said, "no money means no mission."

As a result of the Six Sigma project, the lab implemented new standard operating procedures in 2003 for determining medical necessity of testing and collecting ABNs. The changes cut the write-offs from 5.5 percent to 0.65 percent of total Medicare outpatient lab billing overall for IHC’s 21 hospitals. "In 2004, the outpatient lab write-offs were down to $220,000," Keoppel told CAP TODAY.

So how did the lab set about to pull off this dramatic reduction?

Keoppel and his cohorts first had to convince the chief financial officers that obtaining ABNs was not worse than the write-offs. Being able to quantify the growing six-figure losses helped. Keoppel also couched the write-off problem as a compliance risk—failure to obtain ABNs meant the lab was essentially giving Medicare patients free testing.

The icing on the argument for ABNs, however, came when the imaging department began to run into local medical review policies limiting coverage for big-ticket services.

In fact, imaging/radiology accounted for 59 percent of the medical necessity write-off pie in 2002 compared with the lab’s 24 percent share. The cardiac catheterization lab share was nine percent, and a miscellaneous "other" category claimed eight percent. (Imaging/radiology and the cardiac catheterization lab implemented Six Sigma projects concurrent with the lab’s, Keoppel says.)

To tackle the lab write-offs, Keoppel and his project team followed the five steps in Six Sigma, which are sometimes portrayed in a circle.

• Define. The team developed a clear statement of the problem and the goal to reduce lab write-offs for Medicare by 75 percent within a year. It also identified the stakeholders: Medicare beneficiaries (or their family members) who weren’t used to paying for lab tests at IHC, the ordering physicians who often viewed payment for testing as the lab’s problem, and the frontline lab staff since ABN collection usually falls to the lab, Keoppel says.
• Measure. As Keoppel puts it, "It’s hard to improve what you can’t measure." The team defined data parameters it wanted to measure and then collected data to identify problem areas and establish baseline data.
• Analyze. This step involves use of statistical tools to identify the problem’s root causes. The analysis involves "slicing and dicing" the data to look at it in different ways, Keoppel says. For example, the team identified which hospitals, doctors, patient types, tests, and diagnosis codes were connected with the most write-offs.
• Improve. The team designed processes and new standard operating procedures based on an analysis of the problem.
• Control. This step often gets overlooked after an organization makes quality improvements, Keoppel says. But an organization must monitor and control its new procedures because "processes tend to mutate back to old patterns" over time.

In defining the problem and its backdrop, the project team noted that Medicare pays for clinical laboratory services if the patient has the specified indications for a test, and that some tests have frequency limits.

Knowing what will be covered has become easier since the Centers for Medicare and Medicaid Services implemented 23 national coverage determinations for testing that encompass 80 percent of testing. The NCDs list all of the Current Procedural Terminology codes and descriptors for the tests, as well as the diagnosis codes that support medical necessity for the CPT codes.

Local Medicare contractors can add to a national coverage determination or impose local coverage decisions, though Keoppel says he’s unaware of any that have done so thus far. Medicare contractors do impose local coverage decisions for testing that the NCDs do not address.

Labs and other providers cannot bill Medicare for tests that don’t have an ICD-9-CM code listed by an NCD or local coverage determination unless they obtain an ABN before providing the services, Keoppel notes. But outpatient labs that make that determination up front and collect ABNs whenever possible can shift financial responsibility for the testing to the beneficiary.

To its surprise, the Six Sigma project team discovered that 93 percent of IHC’s Medicare laboratory outpatients had secondary Medigap-type health plans that covered the testing listed on the ABNs about 80 percent of the time. A lab is also free to bill the Medicare patients its customary charge, which usually includes a markup. IHC is, however, looking at developing a discounted fee schedule for patients who pay out of pocket.

Once the Six Sigma team defined the problem and the improvement goal, it decided to measure the following:

• Write-offs attributed to lack of an ABN.
• How the problem was stratified. The team stratified the write-offs based on where the testing was done, for example, walk-in outpatient testing, other outpatient (where patients register in imaging or cardiology, for example, and then receive lab tests ordered by that department), outreach, emergency department, and ambulatory day surgery.
• Details listing write-offs by hospital per month; doctor; patient type (for example, outpatient, outreach, ER); test; ICD-9-CM code used; and the ratio of write-offs to total Medicare outpatient charges.

The key data elements included the date of service; patient identification; hospital (if more than one—IHC has 21); ordering doctor; patient type (outpatient, emergency room, etc.); test name, CPT, charge number; ICD-9-CM code submitted; and test price (amount written off).

In analyzing the data, the Six Sigma project team found that the so-called Pareto principle (also known as the "80-20" rule) applied in that 80 percent of the problem could be attributed to 20 percent of the ordering physicians. But in IHC’s case, 80 percent of that 20 percent of physicians worked for IHC, which gave the organization clear leverage in getting the physicians to provide accurate diagnostic information for lab testing.

Walk-in outpatient and outreach testing accounted for the bulk of the write-offs. The tests racking up the most write-offs because of inadequate diagnosis codes were the lipid profile and thyroid-stimulating hormone, prostate-specific antigen, and glycohemoglobin tests.

The project team also drilled down to specific ICD-9-CM codes submitted for tests with the most write-offs. For example, when PSA testing was done for diagnostic purposes, the most commonly rejected diagnosis was for benign prostatic hypertrophy. "You can’t report a benign condition when doing PSA to see if someone has cancer," Keoppel says. Medicare would pay for the test, he adds, when the patient had an ICD-9-CM code of 601.9 (prostatitis, unspecified) or 602.9 (unspecified disorder of prostate).

For the Six Sigma team, walk-in outpatient testing was a golden opportunity to collect ABNs before the phlebotomy procedure. At the time, the lab was checking the codes for medical necessity after collecting specimens. And that left the lab behind the eight ball if the diagnosis codes didn’t pass muster with Medicare contractors’ local medical review policies.

To reverse that process, the Six Sigma team decided to turn phlebotomists and registration clerks into certified coders for lab services and give them a coding program to help change physician-supplied narrative diagnoses into an ICD-9-CM code.

To develop coding expertise, the laboratory staff attends a three-day session where they learn the outpatient coding guidelines for ICD-9-CM, and then annual half-day training on the new diagnosis codes and problem areas, says Anita Orenstein, corporate health information management compliance coordinator for IHC. The training also covers compliance dos and don’ts for ABN collection—for example, the lab can’t dispense blanket ABNs to cover any unforeseen Medicare denial. Labs can, however, tender ABNs for any test that has a frequency limit since there’s no way to validate whether the patient has hit the limit.

Following the new standard operating procedures, the phlebotomists and registration clerks trained in coding and ABN collection run the diagnosis codes through a computerized program developed in-house to see if the codes are covered for the test. If the test doesn’t have a code on the national or local coverage determination lists, the phlebotomist or registration clerk collects an ABN from the beneficiary or his or her representative before the lab draw.

Explaining the ABN to patients requires finesse lest they get the idea that their physicians ordered a test they didn’t need. So the Six Sigma team developed scripts for staff to use to explain to beneficiaries why Medicare probably won’t pay for a test.

While the ABN process sounds time-consuming, it moves along quickly, Keoppel says. "When the patient is familiar with the ABN, the process takes 40 seconds," he says. And if the frontline staff person has to explain the ABN to the patient for the first time, the process requires, on average, two minutes, which Keoppel views as time well spent compared with the expense of a write-off.

After the patient signs an ABN, the frontline lab staff may, if time permits, call the physician to ask for more specific information about a diagnosis code, Orenstein says. Lab personnel use a script prepared by the Six Sigma team for that task as well. The script standardizes the process and helps ensure that lab personnel don’t inadvertently engage in the verboten practice of "code steering" by prompting the physician to provide a covered code for the service.

("The lab staff can, however, use a cheat sheet internally to determine if a physician-supplied code is covered for a test," billing and compliance consultant Christopher Young, principal of Laboratory Management Support Services, Phoenix, told CAP TODAY.)

In cases where the physician supplies additional codes or narrative diagnoses that meet medical necessity requirements for the test, then the staff person discards the ABN.

Once the lab obtains a signed ABN from the beneficiary for specified tests, the frontline lab staff person adds a "32" occurrence code to the account which goes on the UB-92 (universal billing form) sent to the hospital lab’s fiscal intermediary. (Inpatient providers use the UB-92 in lieu of the 1500 form to file Part B lab claims.) "An occurrence code 32 tells the fiscal intermediary that the lab has an ABN on file, which is comparable to the GA modifier on a 1500 form," Keoppel says.

With the ABN on file, IHC then bills Medicare for the testing. If the fiscal intermediary denies payment based on national or local coverage determinations, the lab is free to bill the patient’s secondary insurer or the patient on a private-pay basis. (All claims with a signed ABN are submitted to Medicare and Medicare makes the final determination of medical necessity. If the laboratory collected an ABN in error, the CMS will evaluate the claim and pay it if the diagnosis is medically necessary.)

Getting a handle on write-offs for outreach testing was trickier because the lab never sees the so-called nonpatients to secure an ABN.

The Six Sigma team considered how best to work around that obstacle. The lab could call patients to obtain an ABN before doing the testing, but that involved the hassle of having the patient send the signed form by mail, Keoppel says. The lab could also simply refuse to do testing if the physician hadn’t collected an ABN up front, but, Keoppel says, the team concluded: "If a patient gets stuck, it implies you’re going to do something with the sample."

A third option was to "share the pain" with physicians by charging them a penalty when they didn’t obtain an ABN, Keoppel says. But alas, the physicians would just find another lab.

So the project team decided to use for outreach testing its existing process for determining medical necessity. The outreach lab staff codes the specimens in batches as they arrive in the evening. They then run the codes through the computerized edits to see if the diagnoses are covered for the specific test. If not, the testing goes on an "exception report" that comes out each morning.

In fact, all Medicare testing that fails medical necessity edits and lacks an ABN appears on the exception report, says Craig Stacey MBA, MT(ASCP), laboratory compliance and billing director at IHC Urban Central Region. The billing clerk uses the list to call the ordering physicians and ask them why they ordered the testing.

"The physicians will sometimes say the patient wanted the test so he or she ordered it and didn’t obtain an ABN, in which case the lab just writes off the test," Stacey says.

Any time a physician does supply more diagnostic information for the testing, the lab staff documents the conversation, including the date and time, and who supplied the information. "The lab needs significant documentation when you change the original diagnosis codes or narrative that appeared on the requisition," Stacey says.

The Six Sigma team devised new procedures to track outreach write-offs by physicians and tests so as to target educational efforts. Says Keoppel: "We’ve had success in that regard with IHC-employed physicians ... but we still have trouble with the nonemployed, unaffiliated physicians. That’s where we face the same challenges as a Quest or LabCorp-type operation. Education helps, but some physicians don’t care who pays for the test."

The lab does station its own phlebotomists in high-volume, unaffiliated practices, which has the side benefit of reducing write-offs once the phlebotomists are trained to check for medical necessity and obtain ABNs correctly, Keoppel says. The lab hasn’t thus far tracked the precise impact that placing the trained phlebotomists in high-volume offices has had.

Keoppel says IHC will implement early next year a computerized physician order entry system from 4Medica in all physician offices using its labs, which should improve compliance with the ABN process.

"The [electronic ordering] system’s primary goal is to get quicker orders from physicians and to tie the orders into the office management system so the lab can get accurate billing data," Keoppel explains. But the software program also has a coding tool and can check codes against medical necessity requirements. "The system prompts the physician that the diagnosis code doesn’t meet medical necessity and asks for additional diagnostic information," Keoppel says. And it automatically generates an ABN for tests that don’t meet medical necessity requirements or that have a frequency limit.

To maintain the gains realized by the Six Sigma project’s new standard operating procedures for outpatient and outreach testing, the lab monitors write-offs monthly and sends "drill-down reports" to hospitals that identify the write-offs by patient type, patient name, doctors, tests, and ICD-9-CM codes. IHC-employed physicians receive feedback when their write-offs climb too high. IHC also holds people accountable for achieving reductions. "We give them annual goals and let them know monthly how they are doing," Keoppel says.

The Intermountain Health Care lab plans to tackle ambulatory same-day surgery and the emergency department next.

In ambulatory surgery, write-offs for partial thromboplastin time and prothrombin time INR testing have become a major problem since the CMS eliminated a V code two years ago for preoperative screening that covered those two tests, Keoppel says. "If the lab could resolve inappropriate coding and ordering for those two tests, it’d vanquish about 80 percent of the write-offs in the day surgery area for Medicare," he says.

The problem with prothrombin times isn’t unique to IHC, notes consultant Christopher Young, who sees write-offs for these tests in all settings with a hospital-based laboratory. "In the ’old days,’ physicians did the tests to make sure a patient wasn’t going to bleed abnormally in surgery," he says. "But now the reimbursement for the testing isn’t there unless the testing meets more narrow medical necessity requirements."

As a first step toward reducing write-offs in same-day surgery, the pathologist is going to educate surgeons and anesthesiologists about medical necessity requirements for preoperative prothrombin time testing.

For example, neither of the tests detects a surgical patient’s bleeding risk caused by inherited or acquired platelet disorders, including antiplatelet medications such as aspirin, says Sterling Bennett, MD, director of clinical pathology for IHC-owned LDS Hospital in Salt Lake City, who assisted with the Six Sigma project. "So we think surgeons sometimes rely on testing that doesn’t fully answer the question of whether the patient may have a coagulopathy because PT and PTT tests don’t detect platelet disorders."

Dr. Bennett says the decision to do preoperative coagulation testing should be based on the patient’s personal and family history of bleeding problems or thrombotic conditions and on risk factors, including medications. "Unless the patient has signs or symptoms of a current coagulopathy, the most reliable predictor of excessive surgical bleeding is the patient’s response to previous hemostatic challenges, including accidents, dental extractions, surgeries, and other invasive procedures, and history of menorrhagia," he says.

And the most common condition associated with excessive or prolonged bleeding is von Willebrand disease, which has a worldwide prevalence of about one percent, Dr. Bennett says. "PT and PTT are not good screening tests for this condition," he says. "Patient history is the best screening ’test’ and von Willebrand factor testing is required to establish the diagnosis."

PTT and heparin assays are used to assess bleeding risk for people on heparin. "But in many cases, [the physician] will stop the patient’s heparin prior to the surgery," depending on the physician’s cost-benefit analysis and type of surgical procedure, says Dr. Bennett. "And PTT is not a reliable test for monitoring low-dose heparin therapy because it is frequently not prolonged even during therapy."

If the educational effort doesn’t do the trick, the lab may move to plan B, which is to give same-day surgery patients an ABN before doing the testing since the lab staff collects the specimens, Keoppel says.

As for reducing lab write-offs in the ER, education will be the lab’s only recourse. The Emergency Medical Treatment and Active Labor Act requires a hospital emergency department to evaluate and stabilize patients before asking about insurance information, Keoppel notes. That means the ER staff can’t give a patient an ABN until the person is stabilized.

And the initial workup to evaluate the patient usually involves laboratory testing that doesn’t meet Medicare medical necessity criteria about 15 percent of the time at IHC, based on 2004 figures. Says Dr. Bennett: "Testing in the emergency department setting is directed toward immediate decisionmaking rather than screening or risk assessment tests such as lipids, PSA, etc."

Does a lab have to use Six Sigma to reduce Medicare writeoffs?

Not necessarily, Young says. "But the only really effective way to reduce them systematically is to use some sort of quality improvement process involving data collection, such as Six Sigma or a strategic planning approach," he says. The problem has to be defined and data have to be collected to find out who is making what kind of errors—and then the people making them have to be taught how to put an end to the problems. "’One shot’ deals where you get mad one day and write to clients to ride hard on them don’t work."