Laboratories face the challenge of providing quality patient care while managing costs and turnaround times (TATs). To this end, we brought the heparin-induced thrombocytopenia (HIT) antibody test in-house with the goal of reducing costs and the time to diagnosis.
To determine the cost-effectiveness and return on investment of our in-house HIT antibody test by comparing it to send-out assays with TATs of 2, 3, or 4 days.
We performed a retrospective chart review of all patients with a HIT antibody assay and analysis of laboratory financial records. Analysis included the percentage of patients receiving alternative treatment, cost of treatment, startup costs of bringing the test in-house, and average TAT of the in-house test.
We found significant reductions in the cost of treatment for patients and the overall cost to the health care system. The in-house assay became cost-effective at between 8 and 20 tests, with a return on investment of up to 298%.
Bringing the HIT antibody assay in-house becomes cost-effective at a very low test volume with excellent return on investment. This novel analysis can provide a framework for other laboratory medicine professionals to analyze the benefits of bringing this and other assays in-house.
Modern clinical laboratories aim to provide excellent quality by producing accurate and cutting-edge results with the least turnaround time (TAT) possible while simultaneously balancing health care costs. To achieve this, laboratories often compare the cost of materials and labor of performing a test in-house versus sending a test out to a reference laboratory. Often, the decision to bring a test in-house requires complex financial analysis to determine the overall benefit. To this end, laboratories weigh the decrease in TAT by having the test immediately available to physicians, and the initial startup costs of bringing a test in-house. However, global assessment of a test, including clinical outcomes, process outcomes, and structural-related outcomes, is often lacking.1
Heparin-induced thrombocytopenia (HIT) is a serious hematologic condition that can lead to significant morbidity and mortality if not promptly diagnosed and treated. It is caused by the development of antibodies against a complex of heparin and platelet factor 4,2–5 leading to activation of platelets and endothelial cells. This in turn leads to a precipitous drop in the platelet count and a simultaneous risk of clinically devastating thrombosis, with an overall mortality rate of up to 20%.6,7
The diagnosis of HIT is typically made with the PF4/polyanion enzyme-linked immunosorbent assay (ELISA), which detects the presence of antibodies against the platelet factor 4.8,9 We note that this assay has high sensitivity, but lower specificity, as not all antibodies discovered will lead to platelet activation.2,6,10,11 The problem of overdiagnosis of HIT has been somewhat mitigated by determining pretest probability through the use of clinical scoring systems such as the 4T score,6,10,12 which is recommended as the first step in a case of suspected HIT. For 4T scores of 4 or higher, current recommendations are to obtain a HIT antibody assay. If the antibody assay result is negative, HIT can be safely deemed unlikely.13 If the antibody assay result is positive, the American Society for Hematology recommends obtaining confirmatory testing.13 The gold standard for diagnosing HIT remains carbon 14–labeled serotonin release assays, which directly detect platelet activation. However, these assays are technically complex, time-consuming, and costly, and are not routinely offered outside reference laboratories.8–10 In 2017, a new fully automated latex immunoturbidimetric assay was announced, marketed under the name of HemosIL (Werfen Industries, Bedford, Massachusetts).12,14,15 This assay detects the inhibition of particle agglutination through competition between patient PF4/heparin antibodies and a monoclonal HIT-mimicking antibody. This assay has a higher positive predictive value (55.6%) than traditional ELISAs (31.6%) and a high negative predictive value (99.7%).12 The HemosIL assay had a sensitivity of up to 100.0% and a specificity of up to 94.0%,12,14 which indicates it outperforms traditional ELISAs with favorable diagnostic accuracy.12,14,15 Moreover, the assay is automated, has a rapid TAT, and can be performed onsite 24 hours a day, making it an ideal choice for an initial diagnostic assay. Thus, latex immunoturbidimetric assay is currently considered a reasonable assay to perform, even in patients with a lower pretest probability of HIT (4T score of 4, probability of HIT ∼14%).
While the diagnosis of HIT is in question, patients may be treated with alternative anticoagulants, such as argatroban. Argatroban is a direct thrombin inhibitor and reduces the risk of thrombosis in patients with HIT.3,7,16 Argatroban is currently approved by the US Food and Drug Administration as an anticoagulant of choice in HIT17 ; however, it can be expensive and have a range of side effects. These alternative therapies may be necessary to prevent a life-threatening complication, but they also come with a high cost to the patient and the health care system. In addition, the use of alternative therapies can prolong hospitalization, further increasing costs.
Our institution brought the HemosIL assay in-house in 2018, with the goal of reducing costs, hospital length of stay, and the time to diagnosis. Our hematology laboratory supports approximately 1000 beds across 4 campuses, so the impetus to provide accurate and cost-effective laboratory diagnosis is high. In this study, we present a global analysis of the cost of the HIT antibody assay, considering not only materials and labor, but also the cost of treatment with heparin alternatives. Our analysis includes the percentage of patients treated with alternative anticoagulants, the cost of this treatment, the startup costs of bringing the test in-house, and the average TAT of the in-house test. We also evaluate the monetary benefits to both the patients and the institution by performing the test in-house by calculating the cost-effectiveness and return on investment (ROI). In our analysis, the in-house HIT antibody test significantly reduced the use of costly alternative therapies, ultimately leading to cost savings for both the patient and the health care system.
METHODS
Patient Charts/Treatment Data
To obtain the most accurate estimate possible of the cost of treatment, we obtained institutional review board (IRB) approval to access the charts of all patients who had a HemosIL assay at our institution during a 6-month period, without regard to demographics or illness severity. This totaled 99 instances of a HemosIL assay being performed, 2 of which were removed from analysis as duplicates. Of the remaining 97 individual patients, 90 had negative results, and 7 had positive results. The 4T scores were missing from 6 negative charts, so the total population included in that analysis is 91. It is important to note that at our institution, it is suggested but not required that a patient have a 4T score of 4 or greater before an order can be placed. Thus, most patients who receive a HemosIL assay have at least an intermediate probability (∼14%) of having HIT. This work was approved by our IRB, protocol No. 22-0304.
Statistical Analysis
Basic statistical analysis was performed in Microsoft Excel (Redmond, Washington) using built-in functions, and included average, median, and other simple equations.
Financial Calculations
Financial data were calculated by using 2 decimal places in most instances. Final costs, for example, final cost of the in-house test after materials and labor, were rounded to the nearest dollar. Decimals of 0.5 or greater were rounded up. Decimals of 0.4 or less were rounded down.
Cost of Medications
The cost of medications was supplied by our pharmacy and was accurate for our institution at the time this manuscript was in preparation.
Cost of Reagents
The cost of reagents and the estimated number of tests each unit of reagent could support was supplied by a representative of Werfen Industries and was accurate for our institution at the time this manuscript was prepared. We routinely use all reagents before their expiration date. Additional costs from wastage were not included in this study.
RESULTS
4T Scores in Our Study Population
At our institution, it is suggested but not required that 4T scores be 4 or greater to order a HIT antibody assay. Of the 97 charts included in the financial analysis, 91 had 4T score data available. In this population there was a range of 4T scores, with 81% of patients having a 4T score of 4 or greater, and 19% of patients having a 4T score below 4 (Figure 1, A). Patients who tested positive for the HIT antibody assay had scores of 3, 4, 6, or 8 (Figure 1, B). Remarkably, of the 2 patients with a 4T score of 3 who tested positive on the HIT antibody test, 1 was confirmed positive on serotonin-release assay. This highlights the role of clinical presentation and judgment in the diagnosis of HIT.
The 4T scores of the study population and HIT-Ab–positive patients. A, Graphical presentation of the distribution of 4T scores in the study population. Scores ranged from 1 through 6, and 8. A score of 7 was not represented. B, Graphical presentation of the 4T scores of patients who tested positive on our HIT-Ab assay. Patients who tested positive had 4T scores of 3 (2 patients), 4 (2 patients), 6 (2 patients), or 8 (1 patient). Zero patients who tested positive had scores of 1, 2, or 5. Abbreviation: HIT-Ab, heparin-induced thrombocytopenia antibody.
The 4T scores of the study population and HIT-Ab–positive patients. A, Graphical presentation of the distribution of 4T scores in the study population. Scores ranged from 1 through 6, and 8. A score of 7 was not represented. B, Graphical presentation of the 4T scores of patients who tested positive on our HIT-Ab assay. Patients who tested positive had 4T scores of 3 (2 patients), 4 (2 patients), 6 (2 patients), or 8 (1 patient). Zero patients who tested positive had scores of 1, 2, or 5. Abbreviation: HIT-Ab, heparin-induced thrombocytopenia antibody.
Calculating the Average Cost of Treatment With Argatroban for In-house Tests, Both Negative and Positive
Estimating the cost of a HIT antibody test involves calculation not only of materials and labor, but also of the costs of treatment for HIT while the result of the test is pending. To obtain an accurate estimation of the cost of an in-house test, we obtained IRB approval to access the charts of all patients—regardless of demographics or illness severity—who had a HIT test ordered and posted to their chart. We noted the presence or absence of heparin alternatives in their medication administration registry and found that most patients who were treated received argatroban, with a small minority receiving bivalirudin, while 2 patients were treated with both argatroban and bivalirudin, separately and in succession (Figure 2, A). In one instance, a patient with a HIT positive test was treated with fondaparinux (Figure 2, B). For ease of analysis, we focused our financial investigation on the cost of treatment with argatroban.
Treatment categories in patients with a HIT antibody test. A, Graphical presentation of the categories of treatment in HIT antibody-negative patients. Most patients, 77 of 90 (85.6%), did not receive heparin alternatives. Ten of 90 HIT antibody-negative patients (11.1%) received argatroban, 2 of 90 (2.2%) received bivalirudin, and 1 of 90 (1.1%) received argatroban and bivalirudin sequentially. B, Graphical presentation of the categories of treatment in HIT antibody-positive patients. Most patients, 5 of 7 (71.4%), received argatroban. One of 7 (14.3%) received bivalirudin, and 1 of 7 (14.3%) received fondaparinux. Abbreviation: HIT, heparin-induced thrombocytopenia.
Treatment categories in patients with a HIT antibody test. A, Graphical presentation of the categories of treatment in HIT antibody-negative patients. Most patients, 77 of 90 (85.6%), did not receive heparin alternatives. Ten of 90 HIT antibody-negative patients (11.1%) received argatroban, 2 of 90 (2.2%) received bivalirudin, and 1 of 90 (1.1%) received argatroban and bivalirudin sequentially. B, Graphical presentation of the categories of treatment in HIT antibody-positive patients. Most patients, 5 of 7 (71.4%), received argatroban. One of 7 (14.3%) received bivalirudin, and 1 of 7 (14.3%) received fondaparinux. Abbreviation: HIT, heparin-induced thrombocytopenia.
The total cost of treatment was calculated by total medication dose in milligrams released to the patient and pharmacy data on the hospital price of argatroban for either a 250-mg bag or a 50-mg bag. Using this information, the average cost of argatroban treatment per patient with either a negative or positive HIT antibody test result was calculated to be $197 and $3348, respectively (Figure 3). For further analysis of the cost of the in-house test, we included only the cost of treatment for a negative HIT test result.
Cost of in-house versus send-out heparin-induced thrombocytopenia (HIT) antibody tests with negative findings. Analysis of the total cost of a HIT antibody test with negative findings, including materials and labor (cost of test), as well as cost of treatment with argatroban (cost of treatment). The cost of argatroban treatment was estimated by using a dose of 2 μg/kg/h for a hypothetical 80-kg patient. Abbreviation: TAT, turnaround time.
Figure 4. Cost-effectiveness analysis of in-house versus send-out for heparin-induced thrombocytopenia antibody tests with negative findings. Line graph of the cost in US dollars per number of tests. In-house testing is shown in black, send-out testing with TATs of 48, 72, and 96 hours are shown in green, blue, and orange, respectively. The point at which in-house testing becomes cost-effective is 8 tests for TAT of 96 hours, 10 tests for TAT of 72 hours, and 20 tests for TAT of 48 hours. Abbreviation: TAT, turnaround time.
Cost of in-house versus send-out heparin-induced thrombocytopenia (HIT) antibody tests with negative findings. Analysis of the total cost of a HIT antibody test with negative findings, including materials and labor (cost of test), as well as cost of treatment with argatroban (cost of treatment). The cost of argatroban treatment was estimated by using a dose of 2 μg/kg/h for a hypothetical 80-kg patient. Abbreviation: TAT, turnaround time.
Figure 4. Cost-effectiveness analysis of in-house versus send-out for heparin-induced thrombocytopenia antibody tests with negative findings. Line graph of the cost in US dollars per number of tests. In-house testing is shown in black, send-out testing with TATs of 48, 72, and 96 hours are shown in green, blue, and orange, respectively. The point at which in-house testing becomes cost-effective is 8 tests for TAT of 96 hours, 10 tests for TAT of 72 hours, and 20 tests for TAT of 48 hours. Abbreviation: TAT, turnaround time.
Estimating the Cost of Treatment With Argatroban for a Send-Out Test, TAT for Send-Out Tests, and the Cost of Send-Out Tests
Argatroban is typically given as a continuous medication and titrated to achieve an activated partial thromboplastin time in the target range. Dosing can vary widely by comorbidities, anywhere from 0.1 μg/kg/min to 10 μg/kg/min.18 For HIT, dosing recommendations are 2 μg/kg/min18 ; we took a hypothetical dose of 2 μg/kg/min for a hypothetical 80-kg patient, which was calculated to be 9.6 mg/h. From this, we estimated a cost of argatroban treatment of $17.94/h, which was multiplied by 48, 72, and 96 hours to estimate the cost of treatment for a send-out test with a TAT of 2 days ($861), 3 days ($1292), and 4 days ($1722), respectively (Figure 3).
TATs at reference laboratories for a HIT antibody test are typically listed as 24 hours. Our analysis indicates that our average TAT for send-out HIT tests, before bringing the HemosIL assay in-house, was approximately 72 hours. However, TAT for a send-out test might be significantly lower or higher depending on shipping times and other factors. For analysis, we chose 3 estimated TATs covering a broad range of potential circumstances: 48, 72, and 96 hours, or 2, 3, and 4 days, respectively.
The cost of a send-out test per our contract with our reference laboratory was $54 dollars per test for a HIT antibody ELISA assay (Figure 3). We did not assess confirmatory functional tests in this study.
Estimating the Cost of the In-house Test
To calculate the average cost of an in-house patient test, we obtained the cost to our institution of all Werfen HIT antibody assay reagents, as well as the expected number of assays per reagent unit, from our Werfen representative. A total price per test for each reagent was calculated (see Table 1). The total cost including labor of a control test was found to be $69.93 and the total cost of a patient test was found to be $61.25.
At our institution, patient tests are typically run individually and on demand and include both a positive and negative control. Thus, the price of a patient test would be $61.25 plus 2 × $69.93, or $201.11, which was rounded to the nearest dollar for further analysis ($201) (Figure 3). However, there are circumstances in which this would be an overestimation. For example, if more than 1 HIT antibody assay is ordered within a short time frame, tests would be run together with the same controls, and the total cost of testing per patient sample would be less.
Calculating the Average Turnaround Times for In-house Tests
To analyze the average TAT of HIT testing at our institution, we obtained the TAT of all 97 nonduplicate HIT antibody tests performed in a 6-month period. The range of TATs was 18 minutes to 172 minutes, with 1 extreme outlier of 748 minutes, which was excluded from analysis. The average TAT was found to be 62 minutes, with a median of 53 minutes.
Estimating Start-up Costs on a Preexisting Machine
HIT testing was brought in-house on a preexisting platform, the ACL TOP 750 from Werfen. Thus, our startup costs were limited to the controls and reagents necessary to validate this test on our existing equipment. We estimated we needed 100 patient samples and 50 control samples to validate the test. Using the costs listed in Table 1, we estimated the price of a control assay to be $69.93 and that of a patient assay to be $61.25. The total estimated startup cost was calculated to be $9622.
Cost-effectiveness and Return on Investment Analysis
To determine the cost-effectiveness of in-house testing, we created a line plot comparing the number of tests against cost in US dollars for the in-house test versus a send-out test with TAT of 48, 72, and 96 hours (Figure 4). The y-intercept of the in-house test plot was adjusted to reflect start-up costs of $9622 (Figure 4). The points at which the lines intercept represent the number of tests at which the in-house test becomes cost-effective. In-house testing becomes cost-effective when compared to a send-out test with TAT of 48 hours at 20 tests, compared to 10 tests for a send-out test with TAT of 72 hours, and 8 tests for a send-out test with TAT of 96 hours (Figure 4).
ROI analysis is used to determine the ROI, relative to the investment's initial cost.19,20 For our analysis, ROI was calculated by using the total savings in costs at 100 and 200 assays. The total estimated cost of in-house tests was subtracted from the total estimated cost of send-out tests, divided by the total cost of in-house testing, then multiplied by 100. This analysis revealed significant positive ROI for the in-house test versus send-out tests at volumes of both 100 and 200 total assays (Table 2). Our institution averages approximately 200 HIT antibody assays per year. According to this analysis, bringing the HIT antibody assay in-house has saved our institution up to $265 000 with an ROI of up to 298% annually.
DISCUSSION
In this study, we performed cost-effectiveness and ROI analysis of an in-house HemosIL HIT antibody assay.19,20 For comparison, we used a send-out test with TAT of 48 hours (2 days), 72 hours (3 days), and 96 hours (4 days). We found in-house HIT antibody assays to be cost-effective at low test volumes in each analysis, with a positive ROI in each case.
It is well known in the laboratory medicine community that bringing assays in-house reduces cost, and considering the expense of heparin alternatives, it may be common sense that in-house HIT antibody assays would be cost-effective. Indeed, several studies have analyzed the cost-effectiveness of HIT antibody testing in prior years. In 2008, Goldman et al21 published a cost analysis of ELISA HIT antibody testing, which concluded that in-house testing would likely be beneficial but did not directly analyze the cost of in-house versus send-out testing. In 2016, Caton et al22 published an analysis of on-demand HIT antibody testing. This analysis included the HemosIL assay but was based on structured interviews and modeling of a hypothetical cohort of patients, and not real-world financial data. Also in 2016, Mariampillai et al23 assessed the total cost of HIT testing and replacement therapy in their institution but did not provide a comparison of in-house and send-out testing, or a cost-effectiveness or ROI analysis.23 In contrast, the analysis presented here specifically determines the benefit of bringing the HemosIL assay in-house with direct comparison to send-out tests using real-world financial data. This study provides the novel finding of in-house HemosIL HIT antibody assays becoming cost-effective between 8 and 20 tests, with a potential ROI between 85% and 298%. Thus, bringing the HemosIL assay in-house created direct financial benefit to our institution, as well as our patients, by reducing unnecessary expensive treatments.
Currently, the prevailing funding model for laboratory medicine is based on a cost-per-test approach, which typically considers only the costs of reagents, upkeep, and labor. However, this does not consider the global value of a laboratory test, which includes “providing information to clinicians that enable them to make better decisions in the care of patients.”24 Thus, the value of a laboratory test can include reduced length of stay in the hospital, patient satisfaction, life-years gained, and even quality of life—all of which can be difficult to measure.1,19 For laboratory medicine, it can be challenging to justify investment in a new test given the myriad of quantitative and qualitative impacts it produces, and the number of stakeholders that must be considered.1,19,20,24 In this study, we perform a straightforward cost-effectiveness and ROI analysis, borrowing from financial principles in the current literature,1,19,20,24 as well as widely accepted business principles. The analysis presented here can be used as a blueprint for other laboratories considering bringing new technology in-house.
Per our analysis, the average TAT of the in-house assay was 62 minutes, with a median of 53 minutes. One of the strengths of having an in-house assay with rapid TAT is that it allows teams to comfortably pause anticoagulation until results can be achieved. We noted that most patients with a negative finding for HIT antibody assay received no heparin alternatives, and heparin therapy was resumed once HIT had been effectively ruled out. In the case of patients who were treated with heparin alternatives, their treatment was minimal. In contrast, a send-out test with TAT of even 48 hours makes it necessary to treat with heparin alternatives over a prolonged period, which can substantially drive up costs and prolong hospitalization. We did not include data on time to discharge in this study, although future studies may focus on the effect of in-house versus send-out assays on hospital length of stay.
Our assessment found argatroban to be the most frequently used heparin alternative at our institution, with a minority of patients receiving bivalirudin, and a small number of patients receiving argatroban and bivalirudin separately and in succession. Only 1 patient was noted to receive fondaparinux. Given the relatively short TAT of in-house testing, we were initially surprised to note any treatment with heparin alternatives in patients with a negative finding for HIT antibody assay, as the short TAT should allow for a comfortable pause in anticoagulation. Our assessment is that this is due to lack of communication between the laboratory and clinical teams, or a lack of awareness among clinical teams that the HIT antibody test has an approximate TAT of 1 hour.
This study largely relies on estimates, which were made as accurately as possible by drawing from real-world laboratory financial data and patient data. However, in application we acknowledge that variables may exist that render the values cited in this study either overestimations or underestimations. For instance, we focused our analysis on the cost of treatment with argatroban, as it was the most used at our institution; however, other institutions may carry a wider formulary or preferentially use different heparin alternatives. We also note that costs of medications and materials were gathered from financial records at our unique institution and may vary in other locations and markets. Finally, this analysis was a focused analysis of the Werfen ACL Top 750 system using Werfen HemosIL reagents. There may be differences in the cost-effectiveness and ROI found with other systems by other companies. However, given the high cost of heparin alternatives and the relatively low cost of HIT antibody testing, analysis of other systems will likely also find that in-house assays become cost-effective at a low assay volume.
In the United States, health care costs have been steadily rising during the past 20 years. Hospital laboratories are called on to provide excellent quality of care while balancing cost and providing the best value to both the patients and hospital systems they serve. When considering the cost of a laboratory test, a global analysis must be undertaken. While the costs of materials and labor must be tabulated, consideration must also be given to the costs of delayed diagnosis and treatment, which create direct financial costs to patients, as well as unmeasurable deficits in quality of life. This latter value is difficult to quantify; however, including the cost of alternative therapies in analysis helps provide a broader assessment of the cost of an assay to both the patient and the hospital system as a whole.
References
Author notes
Golding is currently located at the University of Kentucky, Lexington
Competing Interests
The authors have no relevant financial interest in the products or companies described in this article.