May 2007
The Emory Medical Laboratory, part of Emory Healthcare (Atlanta, GA), was making an effort to evaluate its laboratory send-outs and—as James Ritchie, Ph.D., the lab’s associate director, says—see if we could rein things in." Right around that time, ARUP Laboratories (Salt Lake City, UT), began offering a new program to its customers. The program was called ATOP, which stands for Analyzing Test Ordering Patterns, and is a new component of its suite of integrated services.
|
James Ritchie, Ph.D.
 |
"There is a lot of evidence out there of inefficient use of all kinds of different medical services," says Brian Jackson, M.D., medical director for informatics at ARUP. "A lot of attention has been paid to things like surgical procedures, hospital admissions, and drug prescriptions, but not as much attention historically has been paid to laboratory tests."
|
Brian Jackson, Ph.D.
|
Jackson cites a study published in Health Affairs that analyzed the variations in total expenses among various medical procedures. "Overall there were about 60% higher expenditures at the more expensive centers than at the less expensive, but more interesting to me was that the laboratory costs varied even more than the other areas."
The purpose of ARUP’s ATOP program is to analyze laboratory test ordering patterns to look for inefficient or inappropriate ordering. "We know that lab test ordering is driven by individual physicians, but there’s not a lot of oversight," says Jackson. "Payers don’t provide a lot of oversight of this compared to, for instance, the oversight given to surgical procedures. There’s a lot of variation."
Case Study: Serum Drug Screening
In the course of analyzing Emory’s lab ordering patterns, ARUP discovered that the single largest test request was for serum drug screens. In fact, Emory was ARUP’s largest single client using the test. "We looked at that and thought it was odd," says Ritchie. "We weren’t a forensic drug lab. Why was that happening?"
The answer is slightly embarrassing, although not all that hard to believe in any complex laboratory setting. Emory Healthcare has a large organ transplant program. A major component of that program is kidney transplants. "Often people who don’t have functioning kidneys can’t make urine, so they couldn’t send urine samples," says Ritchie. "In those instances, they called down to the lab and asked if they could send a serum drug screen. I OK’d that."
And what appears to have happened was the protocols for kidney transplants were cut-and-pasted into the protocols for all organ transplants. Every single transplant patient had a serum drug screen instead of the less expensive urine drug screen. Emory responded by changing the protocols.
"We met with the surgeons," says Ritchie. "They actually wanted to keep that option in for patients who could produce urine." This wasn’t terribly efficient, but the lab agreed that the physicians could attempt to get a urine sample, but if they couldn’t, to go ahead and draw a serum drug screen sample. "They were to request a urine form for the patient and give him or her time to provide a sample," explains Ritchie. "If we get a urine sample in the next six hours, we will just cancel the serum and only send the urine. And that’s worked amazingly well."
ARUP’s ATOP program also detected other anomalies in Emory’s test ordering, but the blood serum drug screen was the most significant change. How significant? According to Ritchie, the changes made as a result of the program saved Emory about $75,000 per year.
Nuts & Bolts
So how does the system work? A team of ARUP pathologists and data analysts, as well as University of Utah School of Medicine faculty, analyze a client’s test ordering patterns. They are looking for areas of potential over-, under-, and misuse. Costs and referral test volumes are compared and evaluated against ARUP’s knowledge base of ordering issues, which draws on their database of more than eight years of archived test orders from hundreds of hospitals and laboratories nationwide.
"Simply by screening ordering volumes of different tests and looking at ratios of ordering volumes of different tests, it’s possible to identify areas of inefficient testing," says Johnson. The first area of testing that they looked at was hepatitis C followed by prostate specific antigen (PSA). The program currently looks at about two dozen different disease topics. Typically, they focus on whatever the clients ask them to, rather than perform a complete analysis.
The process takes about a month. Interestingly, ARUP does not charge its largest clients for this service, although there is a fee schedule for other clients. Plans to offer the service outside the ARUP client base are not yet fully formed. "We haven’t answered that question entirely," says Jackson. "We might consider a project with an outside ARUP client. Historically we haven’t. We have all the data on our clients’ esoteric testing. We’ve done some experiments on looking at our clients’ internal ordering data as well, where the clients provided us with a spreadsheet of their test menu and annual volumes, and by doing that, we can do a more complete analysis."
Targets of Opportunity
The ATOP analysis focuses on four primary categories: order volumes compared to order volumes of related tests; order volumes compared to those of other clients; result distributions; and age, sex, and/or result distribution. In addition, they will make client-specific recommendations about patient care, profitability, and efficiency.
One aspect of the analysis might be to review the top 80% of expenses in order to identify opportunities for improvement. In one example, of the 792 individual tests ordered by a laboratory in a 12-month period, only 84 of these tests (or 10.6%) were responsible for 80% of the total laboratory expenses incurred in that period.
ARUP also provides the more specific example of a laboratory that was performing a number of tests for inherited thrombotic disorders. ATOP analysis demonstrated that while the prothrombin mutation, protein C and S deficiencies, and anti-thrombin III deficiency combined accounted for less than 10% of all inherited thrombotic disorders, the tests to diagnose these disorders represented 72% of the assays ordered by the laboratory in the period analyzed. By testing for APC resistance first, the analysis suggested, the laboratory’s clinicians would greatly reduce the number and costs of subsequent tests.
"The role of ATOP analysis is very much a screening tool," says Jackson. "Because it’s based on a limited dataset that we have, we’re not linking that to clinical data or diagnoses or follow-up testing or anything like that. So we can’t say with respect to any single order: that order is appropriate or not." What they can say, for example, is that this hospital is ordering five times more of a certain test than another hospital and there is no clear reason why, or that the ratio of Test A to Test B is much higher than you would expect if they were following national guidelines. "We’re looking at the patterns of screening, and then it’s up to our client to decide which topics to choose to follow up further," he adds.
|
Ritchie, however, has found it so helpful he has asked for regularly scheduled reports. "What the whole process has pointed us toward is tracking our send-outs on a computerized system." Ritchie receives a quarterly report on what his laboratory is sending out and the laboratory’s top 10 tests, which he then passes on to the rest of the staff broken down by lab sections. "That way we can say, ‘Here are targets of opportunity to possibly bring things in-house.’"
|
May 2007 - Table of Contents
Archives