The ABC's of Pediatric Laboratory Medicine-S is for Send Outs
Mark A. Cervinski, Ph.D., DABCC, FACB
Director of Clinical Chemistry and Point-of-Care Testing, Assistant Professor of Pathology, The Geisel School of Medicine at Dartmouth, Hanover NH and Dartmouth-Hitchcock Medical Center, Lebanon NH
Clinical laboratories have been faced with a number of challenges over the past few years, most notably reduced reimbursement for the routine tests ordered on many of our patients. This is most evident in the decreases in the Medicare Part B clinical laboratory fee schedule (1). Coupled with this challenge has been the tremendous growth in referral laboratory tests known colloquially in the laboratory as send-outs or mail-outs. In the U.S. the referral laboratory business is a $70 billion dollar industry (2) that is growing by approximately 8% per year (3). Within this 8% increase, the largest driver has been the growth of expensive "boutique" molecular laboratory testing for constitutional developmental disorders, cancer genetics and neurological conditions. Consequently, the combination of these financial pressures has led many clinical laboratories to evaluate strategies to reduce their financial liability in an effort to stave off layoffs and other drastic financial austerity measures.
One such strategy has been to rely heavily on the Laboratory Medicine/Clinical Pathology faculty to develop a send-out review strategy to lessen the financial burden associated with this testing. We, like many other laboratories, have also noticed the dramatic increase in the growth of these referral tests, and with it, our financial liability. To respond to this growth, we have moved to develop a referral test utilization committee to address this serious threat to our solvency. Not only are these activities important to the continued success of individual laboratories and hospital systems, they should also be a priority for those of us in laboratory medicine to ensure that our patients are receiving testing along the lines of the Clinical Laboratorians three "R's": The Right Test, for the Right Patient, at the Right Time. In this article I will be sharing with you the strategies we have used to develop a test utilization committee from a conceptual vision to a program in its adolescence. Throughout this article I will also reveal the successes that we have managed to achieve and discuss what our next steps are to develop our system into a self-sustaining program that enhances the training our pathology residents receive during their clinical pathology rotations.
Our Impetus and Our Target
As mentioned we have seen our financial liability to referral laboratories increase in the past few years. A rather shocking increase in volume (Fig 1A) and cost (Fig 1B) was noted between 2008 and 2011 wherein our send-out volume increased from 160,703 to 275,110 tests with a concomitant increase in expense from $4,965,955 to $7,928,310 annually. We had noted this trend prior to 2011 and had moved to address the increase proactively by working with the major reference laboratories in the U.S. in a Request For Proposals (RFP) process between the 2009 and 2010 fiscal years to consolidate as much of our business as possible to net a reduction in the cost per reportable result. While this activity did reduce our cost per reportable result, it was not enough to stem the increase. In 2011, the Department of Pathology leadership at Dartmouth-Hitchcock Medical Center convened a meeting to formulate a plan and an aggressive cost reduction target of $2.5 million a year by the end of fiscal year 2015.
While this target was and is aggressive, our main challenge has been how to be fiscally responsible with our limited resources without impacting patient care. To manage both of these goals we've approached this task with a three-pronged strategy. The three strategies are: Make vs. Buy Decisions; laboratory consolidation and price negotiation; and test utilization review.
Make vs. Buy Decisions
Undoubtedly this is the strategy that all laboratorians have employed at one point or another in their careers. In the view of many, testing is best done in a facility that is closely associated with the physician and patient; however the cost of testing does need to be factored in. The most logical place to start a make vs. buy analysis is to generate a list of the instrumentation currently available department wide and the assays that all available platforms can accommodate. Given that information you can compare the assay menu of your department/laboratory to the list of assays currently sent to reference laboratories and highlight those tests that have a combination of cost per test and volume. Most national reference laboratories will provide their laboratory customers with a monthly test utilization report that highlights the number of tests ordered as well as the cost per test. Using this strategy you can identify those targets that can have the greatest impact on your budget.
Once targets for internalization are identified, it is important to get an accurate determination of your estimated cost per test in order to decide if it is financially feasible to internalize the test. Because of the economies of scale, reference laboratories can command a much lower cost per reportable result than a typical hospital laboratory. At times, some targets for internalization are not financially viable due to higher reagent cost. An illustration of two Make vs. Buy calculations are included in Figure 2. Hypothetical test "A" has a monthly volume of 2000 orders at a purchased cost of $20 per reportable result. Test "A" is an available assay on one of our analyzers and our vendor has quoted us a reagent cost of $500 per a kit containing enough reagents to perform 100 tests. In working up a cost per test it is important to remember to include the number of quality control samples you will be running as well as a certain re-run rate to account for samples that will need to be diluted or re-run for various reasons. When we factor in all of the disposable reagents that will go into analyzing this test in-house our cost per reportable result for only the reagent and associated disposables is $6.53 per test. While this cost is significantly lower than the reference laboratory price, it is also important to work with your administration to determine the amount of technical labor and overhead per sample as well as this will add to the cost. As you can see in the example our final cost per test is $7.00 and with the test volume we could save approximately $312,000 annually with this plan.
Not every test you identify as a potential target will work out. For test "B" on the same instrument platform, we can see that the cost per reportable calculation was not favorable, and internalizing this test would result in a net loss of $5,040 annually. When considering the internalization of a test it is also crucial that the stability of the reagent is considered as well. If the reagent once opened is only stable for a short time there may be significant reagent wastage. It is important to point out that both of these examples assumed that I already had the appropriate instrumentation on site and that I would be using existing staff to implement the new testing. If new instrumentation needs to be purchased/leased or new staff would be needed, it is implicitly understood that the costs of the new instrumentation, service contracts, fit-up/construction costs, and salaries for technologist positions need to be included.
Using this strategy as a department, we identified a small number of tests that were internalized at the midpoint of FY 2012. Two of these tests alone, 25-hydroxyvitamin D and the Chromosome SNP array, accounted for an approximate $800,000 annual reduction in our send-out costs. This reduction in send-out volume and cost staved off a further increase in FY 2012 (Fig 1A and 1B) and for the first time in four years we did not experience a significant increase over the previous years' volume and cost. The full effect of these internalization efforts and additional tests accounting for an additional $200,000 is clearly seen between FY years 2012 and 2013.
Laboratory Consolidation and Price Negotiations
We, like many other laboratories, send out to a significant number of reference laboratories. At the outset of our strategy to reduce costs, we had our general chemistry/microbiology testing split between two large national reference laboratories. This not only added to our cost per test as we weren't commanding enough volume at either to benefit from a volume based price reduction, but it also led to a significant amount of confusion as to where the requested assays should be directed. As expected we undertook a request for proposal (RFP) process in order to compare several national reference laboratories. In the RFP process, it is not sufficient for the laboratory medicine faculty to be merely involved in the process. In order to assure that the quality of testing, turnaround time and other value added features between laboratories is considered, the laboratory medicine staff needs to be key players in the RFP process.
Prior to the first presentations by the various vendors we elected to explore, the committee consisting of both laboratory medicine faculty and departmental administration developed a scoring tool with which to evaluate the vendors. This scoring tool took into consideration many factors in addition to the cost per reportable test (Table 1). As part of this process, the group placed a weighting factor to each category that reflected our departmental mission and cost reduction goals. The RFP process is a large and time consuming process that will involve many individuals. To manage this group and coordinate the schedules of all involved, it is wise to enlist the help of a project manager to keep the group on task and working towards a common goal. Changing reference laboratories is a time consuming process that can involve a significant amount of work following the RFP process. This in and of itself will have a cost associated with it, and as such the RFP process is not done frequently.
Our RFP process was completed at the end of FY 2010 and this process was an important component in our strategy to reduce our send-out expense. While our total send-out volume and expense increased in FY 2011 we were able to realize a drop in our cost per test ratio (Fig 1A and 1B). This reduction in cost per test was maintained until FY 2013 when our cost per test ratio again climbed above the pre-RFP levels. This increase was largely due to the internalization of the high volume, low cost vitamin D testing.
Test Utilization Review
One potent tool in reducing the financial liability associated with reference laboratory testing that is not utilized at most hospitals and academic medical centers is using the laboratory medicine faculty for test review and consultation. In many laboratories, including ours, there was little if any review structure in place. Large panels of serum antibodies or genetic markers were frequently ordered by our physicians when alternative strategies could have been employed to save the facility and our patients a considerable amount of money. Knowing that this was an area that we needed to come up to speed on quickly, we circulated a large database of tests to all laboratory medicine faculty for review. The goal of this review was to identify candidate tests that were likely misordered, over ordered, or of limited clinical utility.
This review identified many test panels, such as celiac disease testing, that were sent to highly specialized laboratories that could be replaced with sequential testing profiles from our main reference laboratory for a significant financial savings. In addressing changes in reference laboratory testing such as this, we elicited the help of clinical colleagues. For celiac testing we reviewed the panel approach vs. the sequential testing profiles with Gastroenterology and presented them with the diagnostic and financial evidence. Once we had reached an agreement with our clinical colleagues, it was a matter of removing the large panel from our test catalog and following up with providers who continued to order the large antibody panel.
Using this same strategy, we also noted a number of lower volume tests of limited clinical utility such as the use of adenosine deaminase on pericardial fluid for the diagnosis of extraplumonary tuberculosis. We also noted that a number of providers were ordering both a TIBC as well as a transferrin simultaneously. Our first step in addressing tests such as these was to remove them from the Computerized Physician Order Interface (CPOE). While we occasionally still get miscellaneous laboratory test requests for these tests, simply removing the offending test can have a profound effect on ordering patterns.
Unsurprisingly, we also identified a few tests that had a high likelihood of being ordered in error such as 1,25-dihydroxyvitamin D in place of 25-hydroxyvitamin D or Parathyroid Hormone related protein (PTHrP) in place of Parathyroid Hormone (PTH). In fact, in the case of 1,25 dihydroxyvitamin D, we noted an increase in the number of requests that coincided with the advent of CPOE at our institution (Figure 3). Knowing that direct physician education would likely have limited effect on ordering patterns as we continually have new resident physicians each year, we chose a separate approach. Tests that had a high potential of being ordered in error such as PTHrP and 1,25-dihydroxyvitamin D, as well as all miscellaneous laboratory test requests, were collated into a spreadsheet by our laboratory information system program three times a day Monday through Friday. This spreadsheet is then reviewed by a combination of laboratory staff, pathology residents and pathology faculty.
As part of the review process for commonly mistaken orders such as 1,25 dihydroxyvitamin D, the laboratory staff will cull out certain requests such as those on known renal failure/transplant patients. The remaining requests are reviewed by the pathology resident and/or faculty. Then if it is determined that the physician likely meant to order the 25-hydroxyvitamin D, an email template directing physician to the correct order in CPOE is sent. This activity combined with creative test naming schemes, such as renaming 1,25-dihydroxyvitamin D to 1,25-dihydroxycholecalciferol, successfully reduced the number of incorrect tests being performed on patients for whom this testing was unnecessary (Figure 2).
Review of Molecular Testing
While these small steps can have profound effects on patient care and cost, there is a whole other area of testing that many laboratorians have not yet begun to tackle: the requests for large gene panels. In addressing this challenge we sought to hire a content expert to help in the review of these requests as suggested by others (4). Although these requests are not as frequent as others the cases are often complex and the interpretation of clinical symptoms and presentation can be lost in clinical notes. To aid us in this complex area we were fortunate to work with a Medical Geneticist with a background in connective tissue disorders. When a request for a large gene panel is received, the case is summarized and sent to a team of laboratory medicine faculty including the Medical Geneticist. Following a review of the case, if a conversation between the laboratory and the ordering physician is necessary the conversation is typically led by the Medical Geneticist. In our experience the majority of these requests upon review are either cancelled or reordered in a tiered or sequential manner with the highest prevalence genes targeted first.
At times the requests are quite simple, such as a request for whole gene sequencing when there is a known familial mutation. Requests such as this do not require consultation via the Medical Geneticist, and the physician is typically consulted by either a Pathology Resident or other Laboratory Medicine faculty to amend the order to query the gene for the known familial mutation.
In addition to the RFP process previously mentioned, we've also been actively monitoring the cost of our esoteric molecular tests. Traditionally we've not directed our physicians to particular laboratories for molecular testing. However it became apparent that significant disparities exist in pricing between laboratories. In order to address this challenge as part of the test utilization process that will be described below, we also query the Gene Tests online data base for locations that perform the requested test (www.genetests.org). Once the various laboratories are identified, the cost of the test is determined either through their online test catalog or via a phone call with a laboratory representative. This data is currently being collected and our preferred laboratory for various esoteric tests will eventually be included in our test catalog/laboratory test formulary.
The test review strategy is the most labor intensive activity of all three strategies but it does hold immense potential in reducing unnecessary testing. While our test utilization review only fully came into being in the latter half of FY 2012, we noted significant reductions in our send- out volume and expense. Approximately half of the $2 million dollar reduction in our send-out expenditures between FY 2012 and FY 2013 (Figure 1A and 1B) can be accounted for by internalization of certain assays. The remainder, approximately $1 million dollars, is due partly due to direct intervention in cases of large gene panels and a change in our physicians ordering patterns once the magnitude of the cost of testing was realized.
We've approached a reduction of our reference laboratory volume in three ways. We've internalized high volume and/or high cost testing, we've negotiated test pricing and we've initiated an active review process. While the active review of physician orders requires a significant investment of time, it does fulfill the mantra of many in laboratory medicine: the Right Test, for the Right Patient at the Right Time. These strategies cannot be done by one person working alone. A team of committed individuals is needed to drive the process from various positions. We would not have been able to achieve the early successes we have it had not been for the work done by departmental administration, laboratory staff, departmental faculty and especially without support from the department chair and hospital administration.
Our next steps along this process are to create an interdepartmental test utilization committee to discuss new testing strategies and to develop approved form letters and procedures to codify the test review process. These activities will be necessary as we move from a fee for service era into accountable care and for the review of requests for Whole Exome and Whole Genome testing.
The significant increase in reference laboratory costs for many hospitals is growing to such a level that it may soon threaten our ability to provide care to all of our patients. To that end we live by one last mantra: No Margin, No Mission. It is imperative that this challenge is faced, and while we are a not-for-profit facility, we are seeking new ways of delivering effective and efficient care to our patients so that we may continue to provide care for those who need it most.
Figure 1: Send-out volume and expense trend. Test volumes (A) and cost (B) increased from FY 2008 to FY 2011. Test utilization plateaued and decreased in FY 2012 and FY2013 as a result of our test utilization strategies.
Figure 2: Comparison of two assays being evaluated for internalization. Test "A" on the left demonstrates an assay that would be ideal for internalization while Test "B" on the right represents an assay that would cost more to perform internally.
Figure 3: 1,25 Vitamin D utilization trend. We noted that following the advent of a new CPOE program (red arrow) that our requests for 1,25-dihydroxyvitamin D (Red Squares) and the % 1,25-dihydroxyvitamin D (Green Triangles) increased disproportionally to our 25-hydroxyvitamin D volumes. Following the institution of a review process (blue arrow) the volume of 1,25-dihydroxyvitamin D tests dropped considerably.
Reference Lab Selection Criteria
Test Menu/Breadth of Testing/Quality
Value Added Activities
Table 1 Reference Lab Selection Criteria
*These weighting factors are for demonstration purposes only. Each laboratory should choose weighting factors based on their goals.
- CMS Fee Schedule Document. Accessed October 19, 2013, http://www.cms.gov/Outreach-and-Education/Medicare-Learning-Network-MLN/MLNProducts/downloads/clinical_lab_fee_schedule_fact_sheet.pdf
- Paxton A, How Labs are Taming Test Utilization. CAP Today, June 2013
- Paxton A, The Sense and Cents of Send-Outs. CAP Today, February 2012
- Dickerson JA, Cole B, Conta JH, et al. 2013. Improving the value of costly genetic reference laboratory testing with active utilization management. Arch Path Lab Med. In Press.
Other Suggested Reading:
Kim JY et al. Utilization management in a large urban academic medical center. AJCP. 2011;135:108-118
Dickerson J, et al. Ten Ways to Improve the Quality of Send-out Testing. Clin Lab News. 2012;38(4): 12-13. www.aacc.org/publications/cln/2012/april/Pages/SendOutTesting.aspx#